Internet Marketing Essentials A Comprehensive Digital Marketing Textbook

1. 1Define the one conversion the page exists for. The source defines a conversion as a target action you want visitors to take (a purchase in eCommerce), and the conversion rate as the percentage of visitors who take it. Write the action and the baseline rate before you change anything.
2. 2Start from the ad or link that sends the traffic. Copy its offer and unique value proposition (UVP) word for word. The source rule is that the destination URL must be a landing page relevant to the offer and content in the ad, never a generic home page.
3. 3Match the first screen to that promise. If the ad says 'car insurance quote,' the visitor lands where they can get that quote. If the ad says 'Save 20% This Weekend,' the page says so. Consistent messaging from ad to page is the first optimization.
4. 4Remove anything that competes with the one action. The source notes Google's Panda update penalized pages with a heavy ad footprint (space given to ads instead of content), a useful warning against cluttering a page. Strip navigation, offers and links that do not serve the conversion.
5. 5Fix load time before cosmetics. Searchers expect content almost immediately, and a few seconds frustrates them. The source's tactics: dedicated hosting, compressed images, removing bloated code or plugins, hosting images off-site.
6. 6Treat bounce rate as your relevance alarm. A fast 'back' click signals the page did not provide what was sought. Check title and meta description accuracy, then check that the content is well written and relevant to the term or ad that brought the visitor.
7. 7Build the page elements named in the definition (colors, trust symbols, buttons, images, forms) as a first version, and treat each as a hypothesis. The source offers no rules for them. It does say that early choices are gut instinct and 'you never know until you test.'
8. 8Test one element at a time against the baseline conversion rate. The source's example: a display URL expected to lose beat expectations, and the plain root domain lifted CTR by 30%. Keep the winner, then test the next element. This is continual improvement, not a one-time launch.

The source's car insurance case shows the core mistake and its fix. An ad promises a car insurance quote. Sending the clicker to the insurer's home page, which talks about every type of insurance the company sells, breaks the promise. The correct destination is a landing page where the visitor can get a car insurance quote matching the UVP in the ad. Apply the same test to the source's ChairMarketeer ad ('Folding Chairs, All Colors On Sale. Buy Chairs This Weekend & Save 20%!', pointing to ChairMarketeer.com/Weekend-Special). The page behind that URL should open on folding chairs in all colors, the 20% weekend savings, and one clear way to buy. It should not open on a full catalog. Then measure the conversion rate and bounce rate. If bounces are high, check message match and load time first. Only then test elements such as the button or the form, one change at a time, because the source's CTR test shows that intuition about what wins is often wrong.

• If the ad promises a specific offer (a quote, a percentage off, a product), then the landing page's first screen delivers exactly that offer. Never route to the home page.
• If bounce rate runs higher than expected, then first check that the title, meta description and ad promise match the page content, because a bounce tells search engines the page lacked the sought information.
• If the page loads slowly, then fix speed (compress images, remove bloated code or plugins, consider dedicated or off-site image hosting) before testing design elements.
• If you are choosing between two designs on instinct, then run a test, because gut instinct on the display-URL example was wrong by a 30% CTR margin.
• If the page carries heavy advertising, duplicate content or thin content, then cut or rewrite it, because Panda penalized those characteristics.

• Sending paid traffic to the home page. The tell: the visitor must hunt for the offer promised in the ad, and bounces are high.
• Ad and page say different things (different offer, different wording). The tell: strong CTR but weak conversion rate.
• Slow page. The tell: visitors leave before the page renders and bounce rate climbs, even though the copy is good.
• Cluttered page with ads, extra links or competing calls to action. The tell: visitors click around instead of completing the one action.
• Redesigning by gut feel without a baseline or test. The tell: nobody can say whether the new version converts better, or the change that was 'obviously better' loses.

• ▸Desired conversion: state the single action (purchase, quote, signup) the page exists for.
• ▸Baseline conversion rate: conversions divided by visitors before any change.
• ▸Source ad or link: paste the headline, UVP and offer that drive the click.
• ▸Destination URL: confirm it is a page built for that offer, not the home page.
• ▸First-screen match: does the first screen state the same offer as the ad? Yes/No, and the gap.
• ▸Load time and fixes: note the current time and the tactic applied (compress images, remove plugins, host change).
• ▸Bounce rate: record the current rate and the likely cause (mismatch, speed, weak content).
• ▸Distractions removed: list the ads, links or clutter cut from the page.
• ▸Element under test and result: name the one element changed (color, trust symbol, button, image, form), the variant, and the conversion rate outcome.

1. 1Define email sign-up as a conversion before you build anything. The source treats it as a secondary (alternate) conversion next to the primary one: a purchase for retail, a lead for lead generation, a booking for Expedia. Write down both so you can judge email by what it moves toward the primary conversion.
2. 2Assign email a job by site type. Retail: send product emails to entice more frequent purchases. Lead generation: send periodic information about the company and services to nurture leads through a long sales cycle. Travel/search (Expedia): send special deals on flights, hotels, or vacation packages to win additional bookings.
3. 3Put capture points where intent already exists. Newegg offers two: an 'EMAIL DEALS' tab reachable from the homepage, and account registration during checkout. Add a sign-up at each high-intent step on your own site, and keep the list permission-based (only people who chose to sign up).
4. 4Match the content to the reason the person signed up. A visitor who joined for deals gets deals. A lead-form contact in a long sales cycle gets periodic information, not a hard sell. Segment on what the subscriber asked for, since that is the only signal the source supports.
5. 5Make every email a delivery vehicle for website traffic. The source defines email traffic as recipients clicking a link in the email that takes them to the website. Give each email one clear link to a page that matches the email's promise, so the click lands on the conversion you defined in step 1.
6. 6Measure with the analytics you already have. Track email as its own traffic source, then read visits, page bounce rate (the share of visits beginning on that page that view only one page), and conversion on the landing page. Compare the primary conversion (purchase, lead, booking) against sign-ups for the list.
7. 7Treat spam avoidance as a standing check. The source lists 'Avoiding Spam Filters' as a section of its email chapter, but the supplied text gives no filter rules. Keep to permission-based sign-ups, and consult the full chapter for specific filter criteria before scaling volume.

Take the two sites the source names. Newegg is retail, so its primary conversion is a purchase. It also captures email sign-ups from an 'EMAIL DEALS' tab and registrations at checkout, because emailing customers about products is a profitable way to make them buy more often. Expedia earns money when a user books a hotel or flight, so a booking is its primary conversion. Because emails with special deals on flights, hotels, or vacation packages can produce additional bookings, getting users to join the regular email list is its secondary conversion. In both cases the email's job is to bring a past or prospective buyer back to the site through a link, where the primary conversion happens. To judge either program, you would segment visits from email in analytics, check the bounce rate of the page the email links to, and count how many of those visits end in a purchase or booking. The source does not supply benchmark numbers, so set your own baseline from your first sends.

• If the site is lead generation with a long sales cycle, then send periodic informational emails to nurture rather than promotional blasts.
• If the site is retail or travel, then lead with deals and product emails, because the source ties these to more frequent purchases and additional bookings.
• If a visitor is already at a high-intent step (checkout, registration, homepage deals tab), then place the sign-up there first, as Newegg does.
• If email visits show a high bounce rate on the landing page, then fix the match between the email's promise and the page before changing the list or send volume.
• If you cannot show that a subscriber opted in, then do not mail them; the list must be permission-based.

• Counting list size as success. The tell: subscribers grow but purchases, leads, or bookings from email visits stay flat. Fix: judge email against the primary conversion.
• Sending the same content to everyone. The tell: deal-seekers and nurture-stage leads get identical emails, and clicks are low. Fix: tie content to the reason for sign-up.
• Mismatched click destination. The tell: email visits land on a page unrelated to the email and bounce after one pageview. Fix: link to a page that delivers the email's promise.
• Letting email traffic blend into other sources. The tell: you cannot say how many visits or conversions came from email. Fix: segment email as its own traffic source in analytics.
• Ignoring spam filters until delivery drops. The tell: sends go out but visits from email fall off. Fix: keep the list permission-based and review the avoid-spam-filters guidance before scaling.

• ▸Primary conversion: state the purchase, lead, or booking that email must ultimately support.
• ▸Email sign-up as secondary conversion: where on the site is it offered (homepage tab, checkout, lead form)?
• ▸Email's job: product emails for repeat purchases, nurture information for long sales cycles, or special deals for bookings.
• ▸Subscriber segments: list each reason people signed up and the content each segment gets.
• ▸Permission proof: describe how every address on the list opted in.
• ▸Click destination: name the landing page each email links to and confirm it matches the email's promise.
• ▸Analytics tracking: confirm email visits are segmented as their own traffic source.
• ▸Metrics to review: visits from email, page bounce rate of the landing page, conversions on the primary goal.
• ▸Spam-filter check: note the date you last reviewed filter guidance and any delivery issues found.

1. 1Budget for the work before you open an account. The source's Myth #1 says social media is not free: the platforms don't charge for many activities, but employees must be paid to manage them. Write down who posts, who responds, and how many hours per week, and treat it as a continuing allocation.
2. 2Define one objective before choosing content. The chapter order is plan, then objective, then content, then distribution, then measurement. Pick the outcome from the construct's three aims (build brand equity, induce purchases, or engage followers) and write it as a single sentence you can later measure.
3. 3Tie the objective to the conversion your site type actually needs. A retailer's conversion is a purchase and a lead-generation site's is a lead form. Only then pick a social metric (Likes, retweets, pins, posted pictures) as a leading indicator of that conversion. Those interactions are the engagement behaviors the source says platforms themselves track.
4. 4Match each platform to content built for it. The source covers Facebook (Chapter 10) and Twitter, Pinterest, YouTube and LinkedIn (Chapter 11), each with its own section. Don't paste one asset across all five. Decide for each platform what the native format is (status, tweet, pin, video, professional post) and produce content in that format.
5. 5Set a posting cadence you can sustain, and don't open a platform you can't feed. The source warns that a Facebook page with no activity in the past year suggests the company is out of business, so an abandoned page does active harm.
6. 6Distribute and promote deliberately. Content does not spread on its own, so assign each piece a distribution path: your own followers, links from your site, and any paid promotion. Treat virality as a possible bonus. Myth #2 says good social media need not go viral, so don't make it the plan.
7. 7Keep messaging consistent with your ads, landing pages and email. The guide has a Consistent Messaging section for online advertising, so a social post that promises something the landing page doesn't deliver wastes the click.
8. 8Measure against the Step 2 objective using analytics: referral traffic from the platforms, pageviews, visits, bounce rate, and conversions. Review on a fixed schedule, then keep, change or drop each content type and platform. The source notes that social signals influence search rankings, so also watch whether social activity accompanies better search performance.

The source's own cautionary case is a company whose Facebook page has had no activity in a year, which gives customers the impression the business has closed. Apply the process to that company. Step 1: it admits the page needs a paid owner and a weekly time block, since the page is not free. Step 2: it sets the objective as engaging existing followers, measured by Likes and comments per post, and it explicitly drops 'go viral' as a goal. Step 3: it posts native Facebook content on a cadence it can keep and links each post to a matching landing page with the same offer wording. Step 4: it reviews referral visits and bounce rate from Facebook each month. A high bounce rate means the post's promise and the page's content don't match, and the fix is to change one of them. If engagement is low but referral visits convert, the company keeps the content and works on reach.

• If you cannot staff regular posting, then don't launch or keep the profile, because an inactive page signals the business may be closed.
• If a stakeholder demands that a post 'go viral', then reset the goal to the stated objective (brand equity, purchases or engagement), because virality is not required for good social media.
• If a post drives visits but they bounce after one pageview, then fix the message match between the post and its landing page before posting more.
• If one asset is about to be pushed unchanged to all five platforms, then adapt it to each platform's native format first.
• If the objective is purchases or leads, then judge success on conversions and referral traffic, not Likes alone.

• Treating social media as free and assigning no owner. The tell is a profile whose last post is months old, which makes customers think the business is gone.
• Chasing virality as the definition of success. The tell is a plan that reports shares and ignores whether visitors convert or the stated objective moved.
• Skipping the objective and posting whatever is available. The tell is that no one can say which metric a given post was meant to move.
• Cross-posting identical content to Facebook, Twitter, Pinterest, YouTube and LinkedIn. The tell is uniformly weak engagement and content that looks out of place on most platforms.
• Social messages that don't match the landing page or ads. The tell is high click-through from social but a high bounce rate and few conversions on the destination page.

• ▸Objective (one sentence): state whether the aim is brand equity, purchases, or follower engagement.
• ▸Site conversion it supports: name the purchase, lead, or sign-up this activity should feed.
• ▸Platforms chosen (of Facebook, Twitter, Pinterest, YouTube, LinkedIn): list only those you can staff.
• ▸Native content format per platform: note the post type you will produce for each one.
• ▸Owner and weekly hours: name the person responsible and the time allocated.
• ▸Posting cadence: state the frequency you can sustain without gaps.
• ▸Distribution and promotion path: list where each piece is shared (followers, your site, paid promotion).
• ▸Landing page and message match: record the page each post links to and confirm the wording matches.
• ▸Success metrics and review date: list referral visits, bounce rate, engagement, and conversions, and set the next review date.

1. 1Search the brand name on Google and record every result on the first page by position, URL, source type and tone (positive, neutral, negative). This is your baseline. The restaurant case shows why: ranking #1 told the owner little until she saw that positions 2 through 5 were negative reviews.
2. 2List the places where people talk about the brand: review sites (Yelp and UrbanSpoon in the example), directories, social media sites, Wikipedia, blogs and forums. Check each one on a recurring schedule, because consumers talk about brands online whether or not the brand is listening.
3. 3Sort every negative item into two piles: a real problem (product, service, employee conduct) or a one-off or outdated complaint. Send the first pile to operations as feedback for products, services and employee training. ORM does not excuse bad business practices.
4. 4Respond to negative content professionally and where the complaint appears, aiming to turn the complaint into satisfaction and loyalty. Address the specific issue, state what has been or will be fixed, and do not argue. Negative content can keep doing damage after the underlying issue is solved, so the response is part of the repair.
5. 5Build positive content so it fills the first page. Count how many of the ten slots are positive or brand-controlled, and create assets (site pages, social profiles, directory listings, a LinkedIn company profile and similar) until positive items occupy the whole page. The owned site is only one slot.
6. 6Optimize each new positive asset with on-site and off-site SEO, the same work that took the restaurant's site to #1. Brand-controlled pages need to rank to displace negative links.
7. 7Keep every social profile active. A Facebook page with no activity in a year can make customers think the business has closed, and that damages reputation by itself. Budget staff time for this continually.
8. 8Re-run the first-page audit on a schedule and compare it to your baseline. Track whether negative results are moving down, positive slots are increasing, and new negative mentions are appearing.

A local restaurant hires a web developer, who builds an optimized site and does off-site SEO. Within weeks the site ranks #1 for the restaurant's name. The owner then notices that Google results 2 through 5 are negative reviews from sites like Yelp and UrbanSpoon, so every searcher gets a negative impression despite the #1 ranking. Step one is the audit: record positions 1 through 5 as one owned site and four negative review pages, then fill in positions 6 through 10. Step two is to read the reviews for real operational problems and pass those to the owner for fixes in product, service and staff training. Step three is to respond professionally to the reviews on the review sites themselves. Step four is to create and optimize more positive, brand-related content (additional profiles, listings and pages) so that it competes for the first page. The chapter's stated target is that the developer can fill the entire first page with positive information, which means more business. The measure of success is how many of the ten slots are positive, not the rank of the restaurant's own site.

• If your own site ranks #1 but other top-five results are negative, then treat the reputation work as unfinished and build positive content to displace them rather than celebrating the ranking.
• If a complaint points to a real flaw in product, service or staff behavior, then fix the cause and respond. Do not use ORM to mask a legitimate issue.
• If a negative item appears on a site you do not control, then respond professionally there and add positive content that competes with it in search, rather than ignoring it.
• If you cannot staff a social profile, then do not leave it dormant. An inactive page signals the business may be closed.
• If the first page still has a negative result after your content push, then identify which positive asset can take that slot and optimize it, since the goal is the entire page.

• Treating ORM as a way to avoid responsibility. The tell is that the same complaints keep appearing in reviews while the effort goes only to suppression.
• Stopping at a #1 ranking for the brand name. The tell is a first page where slots 2 through 5 are negative reviews.
• Monitoring only the sites you already use. The tell is that you find out about a damaging thread, blog post or forum discussion from a customer. Mentions can appear on review sites, directories, social media, Wikipedia, blogs and forums.
• Leaving social profiles unattended after creating them. The tell is a last post more than a year old, which suggests the business is defunct.
• Creating positive content without optimizing it. The tell is new assets that exist but do not appear on the first page, so the negative results stay put.

• ▸Brand search term: the exact name searched on Google.
• ▸Position 1–10 log: for each slot, the URL, source type (own site, review site, social, directory, blog, forum, Wikipedia) and tone (positive, neutral, negative).
• ▸Positive slot count: how many of the ten slots are positive or brand-controlled; the target is all ten.
• ▸Negative item detail: for each negative result, the platform, the specific complaint and the date.
• ▸Root-cause flag: does the complaint reveal a real product, service or training issue (yes/no), and who owns the fix?
• ▸Response drafted: a professional reply aimed at turning the complaint into satisfaction and loyalty, plus the date posted.
• ▸New positive asset planned: the profile, page or listing to create or optimize to take over a negative slot.
• ▸Social activity check: the last post date on each profile, and who is responsible for keeping it active.
• ▸Next audit date: when you will re-run the search and compare against this baseline.

1. 1Define the mobile conversion first. The source defines a conversion as a target action you want visitors to take. On a phone that action is often a call, a map visit, or a store visit rather than a purchase form. Write down the one action you want a phone searcher to take.
2. 2Segment mobile traffic from desktop in your analytics and by traffic type: paid search, unpaid search, referral, direct, email. The source notes that more searches happen on mobile devices each day, so a blended average hides mobile behavior. Establish a separate mobile baseline for conversion rate.
3. 3Opt in to the mobile-relevant ad extensions one at a time. Call extensions let mobile searchers call the store directly. Location extensions show the store address with a link to Google Maps. The source says advertisers cannot control when extensions appear but can opt in to each individually, so enable them and check which ones actually show.
4. 4Set up a click-to-call ad where a phone call is the goal. The searcher taps a button on the ad and calls the business, skipping the website. This ad type charges per phone call, so budget per call and not per click.
5. 5Cover map-based intent. Connect the business's local listing through Google+ to the AdWords account so text ads can appear when someone searches Google Maps, for example for the closest dentist. Make the ad and offer distinctive enough to earn the extra trip.
6. 6Use responsive design and simplified content for visitors who do reach the site. Make the page usable on a small screen, with a single clear action and nothing that competes with it. Chapter 13, Mobile Web Design, covers this, but the supplied passages give no further specifics.
7. 7Write mobile-specific keywords and ad copy. Keep titles informative, put keywords in the title so they bold in the SERP, and name the brand if it is well known. Use dynamic keyword insertion when many keyword variants make one fixed headline impractical.
8. 8Review results against the mobile baseline. Compare click-through rate and conversion rate for each extension and ad type. Keep what raises them, and watch for new extensions so you can be among the first to test them.

Tanya's Massage appears in the source as a mobile click-to-call ad. A searcher on a phone sees the ad in a Google search and taps a button on it to call the business at once. The website is skipped entirely, and the searcher goes straight from search to phone call. Google charges per phone call, not per click. For Tanya, the conversion is a call, so she should track calls as conversions and judge the campaign by cost per call. A second case from the source: someone searching Google Maps for the dentist closest to home sees a text ad tied to the dentist's local listing. A location extension with a Google Maps link serves the same nearby intent. In both cases the ad does the mobile adaptation, by shortening the path from search to action.

• If the target action on a phone is a call (a service business, a local shop), then use call extensions or click-to-call ads and measure cost per call.
• If the searcher's intent is a nearby physical location, then add location extensions and connect the local listing to AdWords so you appear in Google Maps.
• If many keyword variants cannot fit in one mobile ad, then use dynamic keyword insertion with a sensible default word.
• If mobile conversion rate trails desktop, then examine the mobile page experience (responsive layout, simplified content, a single clear action) before raising bids.
• If a new ad extension becomes available, then test it early, since extensions can lift click-through rate.

• Treating mobile as shrunken desktop. The tell is that mobile traffic arrives but converts far below desktop, with the same long page and the same multi-step conversion.
• Sending click-to-call searchers to a website anyway, or ignoring calls as a conversion. The tell is that you cannot say how many calls the ads produced or what each cost.
• Assuming opted-in extensions always show. The source says advertisers cannot control when they appear. The tell is an extension enabled for weeks with no impressions reviewed.
• Reporting blended device data. The tell is a healthy overall conversion rate that hides a weak mobile segment.
• Using desktop keywords and copy unchanged for mobile. The tell is ads that never reflect nearby or immediate-need phone searches such as the dentist closest to home.

• ▸Mobile target action: the single conversion you want from a phone (call, directions, purchase, lead).
• ▸Mobile baseline: current mobile traffic and conversion rate, kept separate from desktop.
• ▸Traffic source split: mobile visits by paid search, unpaid search, referral, direct, email.
• ▸Extensions opted in: call, location, offer, or other, and whether each has actually appeared.
• ▸Click-to-call setup: yes or no, cost per call, and number of calls tracked.
• ▸Local listing status: whether it is connected to AdWords for Google Maps ads.
• ▸Mobile page check: responsive layout, simplified content, one clear action (pass or fail with notes).
• ▸Mobile keywords and copy: the phone-intent phrases used and whether the title carries the keyword.
• ▸Next test: one new extension or change to try, with the metric (CTR or conversion rate) that will judge it.

1. 1Define ranking as a two-part score before you touch anything. Google finds every page relevant to the search phrase, then rank-orders those pages by relevance and by PageRank (popularity). Diagnose each page against both halves separately.
2. 2Pick one target keyword phrase per page and confirm the page actually contains it. A page that lacks the word gets a zero relevance score for that search, however popular it is.
3. 3Place the phrase where Google weights it most. Word instances are scored on font size (larger weighs more) and position on the page (closer to the top weighs more). Put the phrase in the page's URL and title tag (the phrase at the top of the browser that labels the page), and set it prominently near the top of the page.
4. 4Audit the anchor text of links pointing to the page. Descriptive anchor text such as "Monarch butterflies" supports relevance; "click here" does not. When you request or earn links, ask for anchor text that names the topic.
5. 5Build popularity through off-site SEO: attract backlinks from other websites. Treat links as the popularity input, and earn them naturally instead of manufacturing them.
6. 6Run a Panda quality check on the page: heavy ad footprint (lots of space for ads instead of content), duplicate content found elsewhere, thin content, stale content with no steady additions, and typos or poor grammar. Fix each one that applies.
7. 7Screen every tactic for white-hat or black-hat. Auto-generated pages pointing at your page, hacked sites, hidden backlinks in code, and purchased links are black-hat methods that Google updates are built to penalize. Drop any tactic that fails the screen.
8. 8Judge results by the framework of traffic, conversions, and revenue. Ranking matters because a well-ranked common keyword can bring hundreds or thousands of visitors per month. Check whether those visits convert.

Take the passage's own butterfly example. A page about monarch butterflies is searched for "butterfly." Step one is relevance: if the page never uses the word, it scores zero relevance for that search, so it cannot rank no matter how many links it has. Once the word appears, Google weighs where it sits. A phrase in a large font near the top of the page, in the title tag, and in the URL counts for more than one buried in small text at the bottom. Next, Google looks at the anchor text of links pointing to the page. A link reading "Monarch butterflies" adds relevance, while a link reading "click here" adds none. Popularity then determines the page's position among all the relevant pages, because Google rank-orders them by relevance combined with PageRank. Last, a quality screen applies. If the page is mostly ad space, copies content from another site, or is full of typos, Panda-style penalties push it down. Google reportedly uses over 200 factors and does not disclose all of them or their weights, so you work from the known levers above.

• If the target phrase does not appear on the page, then fix on-site relevance first. Zero relevance means no ranking, and no amount of link building offsets it.
• If the phrase is present and well placed but the page still ranks below competitors, then shift effort to off-site SEO. Popularity (backlinks) is the other half of the ranking score.
• If someone offers fast link volume (bulk auto-generated pages, hidden links, bought links), then refuse it. Those are black-hat techniques, and Google adjusts its algorithm to penalize artificially inflated popularity.
• If a page loses rankings, then check it against the Panda list: ad-heavy layout, duplicate content, thin content, stale content, grammar errors.
• If you cannot find the weight of an individual factor, then do not wait for it. Google does not publish the full factor list or the relative importance of each factor, so cover relevance, popularity, and quality together.

• Chasing links while ignoring relevance. The tell: the page has backlinks, yet the target phrase is missing from its title tag, URL, and top-of-page content.
• Accepting vague anchor text. The tell: inbound links read "click here" instead of naming the topic, so they add little keyword relevance.
• Inflating popularity with black-hat tactics (purchased links, auto-generated linking pages). The tell: a link profile that looks manufactured, which Google updates are designed to discount or punish.
• Publishing thin, duplicated, ad-heavy, or typo-filled pages. The tell: little original content, text identical to other sites, ad space crowding out content, and visible grammar errors, all of which Panda penalizes.
• Treating ranking as a one-time setup. The tell: no new content is added for long periods, even though steady additions are a positive quality signal and the algorithm keeps changing.

• ▸Target keyword phrase: write the one phrase this page should rank for.
• ▸Page URL and whether the phrase appears in it: yes or no.
• ▸Title tag: paste it and note whether it contains the phrase.
• ▸Phrase position and prominence: describe where it first appears on the page and in what font size.
• ▸Inbound anchor text: list the anchor text of links pointing to the page and mark each descriptive or generic ("click here").
• ▸Backlink sources: list the sites linking to the page and how each link was earned (natural or bought/automated).
• ▸Panda check: mark yes or no for heavy ad footprint, duplicate content, thin content, stale content, and grammar errors.
• ▸Black-hat risk: note any tactic in use that is manipulative (auto-generated pages, hidden links, purchased links).
• ▸Outcome measure: record traffic from the organic result, conversions, and revenue for this page.

1. 1Start from the principle the source establishes: a search engine does not read a page like a human. It records word usage, so the ad group, the ad, and the landing page must each repeat the searcher's words. Write down the exact search phrase you want to win (the source's example is "buy Nike running shoes") before touching anything else.
2. 2Build one tightly themed ad group per search phrase. Note: the source passages cover organic ranking, and the ad-specific Quality Score formula is not in them. The organic logic is the usable evidence: relevance is a measured match between the phrase and the words on the page, and a page that never uses a term gets no credit for it.
3. 3Put the exact phrase, in order, in the places Google is documented to weigh: the page's title tag, URL, and headline/top-of-page text. Larger fonts and positions near the top get greater weight, and an exact-order phrase counts as a perfect match. Adjacent words scattered around the page are worth less.
4. 4Write the ad text to use the same phrase as the keyword and the landing page, so the searcher sees one consistent message from query to ad to page. The source's logic is that relevance is measured by the phrase matching the content, not by general topic closeness.
5. 5Audit the landing page for the quality signals Panda penalizes: heavy ad footprint (more space for ads than content), thin content, duplicate content copied from elsewhere, poor grammar and typos, and stale content. Fix these before raising bids, because a bid cannot repair a page the platform scores as unappealing.
6. 6Check the user-behavior signal. If visitors click through and immediately hit Back, Google reads that as the page failing to give relevant content. Compare click volume against time on page or bounce data for each ad group and treat high immediate-exit groups as relevance failures.
7. 7Judge on-page quality by whether the page offers real content to someone who arrives. In the source's example, Page 2 had only a headline phrase and a butterfly image, and Page 1 had actual content. An exact-match headline on an empty page is the failure pattern.
8. 8Rerun the audit after each edit, one change at a time, so you can tie a change in position or cost-per-click to a specific relevance or quality fix. Since position and cost-per-click come from relevance combined with bid, a relevance gain lets you hold or lower bids instead of raising them.

The source's Monarch butterfly page is a clean relevance test. The page has 61 words. "Monarch" appears twice and "migration" three times, but "butterfly" never appears. Under the original algorithm, a search for "butterfly" scored zero relevance, even though the page is about butterflies. A search for "Monarch migration" scored relatively high, and higher still if the exact phrase appeared in order. Apply that to an ad group targeting "Monarch migration": the ad and landing-page title should carry "Monarch migration" in that order, near the top and in large type. An ad group targeting "butterfly" pointing at this page would be a mismatch, since the page never uses the word. The source notes Google now knows a Monarch is a butterfly, so the zero would no longer apply, but exact wording still earns the strongest match. The second check is quality. A page holding only the phrase and an image (the source's Page 2) matches the words but offers no content, so it should not be paired with a high bid. The source says only the contrast between the two pages, not any ad-score numbers, so judge by that contrast.

• If the searched phrase does not appear in the landing page's title, URL, or top text, then fix the page or move the keyword to a page that does use it, before changing the bid.
• If an ad group mixes several distinct search phrases, then split it so each ad and page can match its phrase exactly.
• If the landing page is thin, duplicated from elsewhere, full of typos, or dominated by ad space, then repair the content first. These are the characteristics Panda penalizes, and bidding higher does not offset them.
• If clicks arrive but visitors leave immediately, then treat it as a relevance mismatch between ad promise and page content, not a traffic problem.
• If a relevance fix improves position, then test lowering the bid before raising it, since position and cost-per-click depend on relevance combined with bid.

• Treating Quality Score as a bid problem. The tell: you keep raising bids on an ad group whose landing page never uses the searched words.
• Keyword stuffing without content. The tell: the page repeats the target phrase, like the source's Page 2, but has no substance for a visitor, and immediate Back-button exits show up.
• Using one generic landing page for every ad group. The tell: the title and headline match none of the specific phrases driving the clicks.
• Ignoring on-page quality signals. The tell: relevance wording is right but the page has thin copy, copied text, typos, or more ad space than content.
• Applying the organic descriptions as the exact ad formula. The tell: you cite specific weights or thresholds that the source does not provide. Use the principles (phrase match, placement, page quality) and verify specifics inside your platform's own reporting.

• ▸Target search phrase: write the exact phrase the ad group is built to win.
• ▸Ad group and keyword(s): list the keywords, confirming they all express the same phrase.
• ▸Ad headline/text wording: record whether the exact phrase appears, and in what order.
• ▸Landing page URL and title tag: note whether each contains the phrase.
• ▸Top-of-page headline and font size: confirm the phrase appears near the top in prominent type.
• ▸Content check: note whether the page has real content, or is thin, duplicated, or full of errors.
• ▸Ad footprint check: record whether ads take up more space than content on the page.
• ▸Visitor behavior: record clicks versus immediate exits and any pattern of Back-button behavior.
• ▸Action and retest date: write the single change you make and the date you will re-measure position and cost-per-click.

1. 1Define engagement per site type before you measure anything. The passages give three working definitions: for media sites it means more pageviews per visit or more videos watched; for social platforms it means users who interact with content ('Like' a status, retweet, pin) or create content (post pictures, create a pin); for retail and other sites, secondary conversions such as account creation and email signup mark deeper interest. Write one sentence naming your site's engagement behavior and the primary conversion it feeds.
2. 2Pull the baseline engagement metrics from your analytics package: Average Page Depth (pageviews ÷ visits), Average Visit Duration (the sum of time on page across a visit, excluding the final 30 minutes of inactivity), Site Bounce Rate (the share of visits with only one pageview), and Time on Page. Record them at site level, then separately for the pages that receive entries.
3. 3Break the metrics down by page. Use Page Bounce Rate (the share of visits beginning on that page that end after one pageview), Entrance Rate, and Exit Rate to find which pages draw visitors in and which end visits. Entry is often not the home page, so audit the real entry pages first.
4. 4Set up engagement goals so the analytics package counts them. Create a pages-per-visit goal (for example, the percentage of visitors who view more than five pages), a time-on-site goal (for example, the percentage who stay longer than two minutes), and event goals for watching a video, downloading a white paper, or clicking social media icons via event tracking.
5. 5Add destination goals (a Thank You page after an order or lead form) so engagement can be tied to Conversion Rate. Then compare the conversion rate of visits that meet an engagement goal with the rate of visits that do not. This tests whether engagement mediates conversion for your site rather than assuming it does.
6. 6Place engagement prompts where entry and exit data show visitors leaving: recommendations, links to further pages, video, and email signup opportunities. Change one page element at a time and compare Page Bounce Rate, pages per visit, and goal completions before and after.
7. 7Where you buy traffic or send email, track engagement by source. Compare ad CTR and email opens and clicks against the post-click metrics above, so you know which sources send visitors who engage and which send visitors who bounce.
8. 8Review the engagement goals on a fixed schedule. Keep only the ones that move conversions or revenue, and drop engagement counts that do not.

Take a media site, which the passages identify as the type whose conversion is hardest to define because it earns money by exposing users to advertisements and can expose them to more ads if it keeps them engaged. Step 1: the team defines engagement as visiting more pages in a visit and watching videos. Step 2: it reads Average Page Depth (pageviews ÷ visits) from analytics and finds the Site Bounce Rate, the percentage of visits with a single pageview. Step 3: it sets a pages-per-visit goal, 'more than five pages,' a time-on-site goal, 'longer than two minutes,' and an event goal for video plays. Step 4: it checks Page Bounce Rate for each entry page, since the first pageview is often not the home page. Pages with high bounce and high exit rates become the first candidates for added links and embedded video. Step 5: after each change it watches the percentage of visitors who meet the five-page and two-minute goals. The passages give no numbers for this case, so the thresholds are the ones they use as examples.

• If you run a media site, then treat pageviews per visit and videos watched as the primary engagement outcomes, because more exposures to ads mean more revenue.
• If a page has a high Page Bounce Rate and receives many entrances (high Entrance Rate), then fix that page first, since it is where most visits begin and end after one pageview.
• If your site is a social platform or has user-generated content, then count interactions (likes, retweets, pins) and content creation as engagement, not only visits.
• If engagement metrics rise but Conversion Rate or revenue does not, then redefine the engagement goal or the audience sending the traffic; the metric is not mediating conversion.
• If a visit ends after 30 minutes of inactivity, then do not count those 30 minutes toward visit duration; analytics already excludes them, so do not adjust for them.

• Reading Site Bounce Rate alone: a sitewide figure hides which entry pages cause it. The tell is that you cannot name a single page to fix.
• Assuming the home page is where visits start: the passages say the entry is often not the home page. The tell is that optimization effort goes to the home page while bounce is high on deep landing pages.
• Treating engagement as a goal in itself: pageviews and time on site rise while conversions and revenue stay flat. The tell is no comparison between engaged and unengaged visits' conversion rates.
• Leaving engagement goals unconfigured: analytics does not track events, video plays, or pages-per-visit thresholds automatically. The tell is that only default metrics exist and no goal reports appear.
• Counting duplicates as volume: repeat hits to the same page inflate traffic. The tell is that Unique Page Views and Unique Visitors are never consulted.

• ▸Site type: media, retail, lead generation, social, affiliate or marketplace; this decides which behaviors count as engagement.
• ▸Primary conversion: the single action engagement should feed (purchase, lead form, ad click, account creation).
• ▸Engagement behaviors: list the specific actions counted (pages viewed, videos watched, likes, shares, email opens, downloads).
• ▸Baseline metrics: current Average Page Depth, Average Visit Duration, Site Bounce Rate.
• ▸Goal definitions: pages-per-visit threshold, time-on-site threshold, and event goals with the exact event to trigger.
• ▸Problem pages: entry pages with high Page Bounce Rate or Exit Rate, plus their Entrance Rate.
• ▸Change to test: the one page element altered, the page it is on, and the date.
• ▸Result: change in goal completions, bounce rate, and Conversion Rate after the change.
• ▸Keep or drop: decision on whether this engagement goal predicts conversion or revenue.

1. 1Define the equity objective before any activity. State how higher brand equity changes customer behavior: purchase likelihood, purchase frequency, or purchase amount. Follower or fan counts are not an objective. Millions of fans matter only if they affect the bottom line.
2. 2Decide what the positive association is built on. It can be product-centric (a car's horsepower and gas mileage) or emotional (Coca-Cola almost never advertises product superiority and builds positive emotional connections instead). Pick one and write it down.
3. 3Fix the technical trust signals first. Clear black-hat practices and comply with Google Webmaster Guidelines. If penalized, correct the problem immediately and file a formal reconsideration request documenting the steps taken. Keep the XML sitemap free of broken links and errors, because a dirty sitemap tells the search engine the site is badly managed or not trustworthy.
4. 4Audit message consistency across every touchpoint, including ads, landing pages, and offline. Check three things on each: the same value propositions, similar calls to action, and the same design elements (color scheme, characters, fonts, imagery). Flag any landing page that seems unrelated to the ad that brought the visitor there.
5. 5Build the landing page for confidence. Match its color scheme to the ad campaign theme and message. Place trust symbols deliberately. For lead generation, show the phone number prominently above the fold with a call to action. Remove any element that does not directly affect the visitor's decision to reach the goal.
6. 6Make the SERP listing accurate and brand-bearing. Put the brand in the title tag, with the keyword first and the brand after a pipe or colon. Put the brand first if it is recognizable, as Overstock does on its homepage. Keep titles under about 67 characters and meta descriptions under about 155. Write the description so it accurately describes the page.
7. 7Earn popularity signals from outside the site. Other webpages linking to yours signal popularity and likely quality, just as popularity in a social circle is determined by what others think of you.
8. 8Measure and test. Track ad traffic in analytics, calculate ROAS, and A/B test ads and landing pages (value propositions, calls to action, buttons, images, trust symbols). Treat trust and equity changes as hypotheses and let the numbers decide.

The Simplicity Software A/B test shows how a trust-and-conversion change gets verified. Ad A (gray button, black text) got 20,000 impressions, 160 clicks (0.8% CTR), and 4 conversions (2.5% conversion rate). Ad B changed only the button to green with white text. It got the same 20,000 impressions, 185 clicks (0.925% CTR), and 5 conversions (2.7% conversion rate). Because ads run at random across the network and the company tracks the traffic in analytics, it can calculate revenue and ROAS directly instead of taking effectiveness on faith, as a TV commercial must. The same discipline applies to trust elements. Change one element at a time, such as the trust symbol, call to action, or phone number placement. Compare the results, and keep the version the numbers favor. The passages report only these figures. They do not say whether the lift is statistically reliable, so treat a small gap on few conversions as provisional.

• If a landing page's look, offer, or call to action differs from the ad that drove the click, then align value proposition, call to action, and design elements before spending more on traffic.
• If a site has been penalized for black-hat tactics, then fix the violations immediately and submit a documented reconsideration request. Treat the punishment as temporary and avoidable.
• If an element on a landing page does not directly affect the visitor's decision to reach the goal, then remove it or A/B test whether it helps or hurts.
• If the product is high-priced or service-based and needs more information than one page holds, then keep the navigation bar. Otherwise test removing it to keep visitors on task.
• If the brand is already recognizable, then lead the title tag with the brand name. If not, lead with the keyword phrase and follow with the brand.

• Chasing audience size instead of behavior. The tell: reports celebrate fans or followers but cannot say how purchase likelihood, frequency, or amount changed.
• Inconsistent messaging between ads and site. The tell: visitors land on a page that seems unrelated to the ad, with different offers, colors, or calls to action, and bounce.
• Neglecting technical hygiene. The tell: the XML sitemap contains broken links or errors, signaling a poorly managed or untrustworthy site to the search engine.
• Cluttering the landing page. The tell: navigation, graphics, and extras compete with the single goal, and conversion falls when they are present.
• Truncated or inaccurate SERP listings. The tell: titles over about 67 characters or descriptions over about 155 get cut off, so the brand and the page description are incomplete.

• ▸Equity objective: state whether it targets purchase likelihood, frequency, or amount, and why.
• ▸Core association: product-centric or emotional, and the one-line message.
• ▸Value proposition used across channels: write the exact wording and confirm it matches on ad, site, and offline.
• ▸Call to action used across channels: list the wording per touchpoint and note any mismatch.
• ▸Design elements: record color scheme, fonts, characters, and imagery for each touchpoint and note differences.
• ▸Trust symbols on landing page: list each symbol and its placement.
• ▸Technical hygiene check: note penalty status, sitemap errors, and fixes made.
• ▸SERP listing: record title tag (under about 67 characters), meta description (under about 155), and brand placement.
• ▸Test log: record the element tested, variants, impressions, clicks, conversions, conversion rate, and the winner.

1. 1Classify every visit into the three source types: direct (URL typed into the browser), referral (a click from another site), and search (a click from a search engine results page). Keep search separate from referral even though search is technically a form of referral; the passages treat them separately because they differ in many ways.
2. 2Split referral and search into paid and unpaid. Paid referral is banner, text, or rich media ads you pay to display on other sites. Unpaid referral is any link you did not pay for. Paid search is ads on the SERP that accompany the organic results. Unpaid search is clicks on organic results. This gives five buckets: direct, paid referral, unpaid referral, paid search, unpaid search.
3. 3Diagnose direct traffic honestly. It comes from offline advertising (a billboard or TV spot showing the URL) or from habit (the espn.com daily visit). If you cannot afford offline advertising and your content does not merit repeat visits, expect little direct traffic and do not plan around it.
4. 4Match each bucket to its acquisition work: unpaid search through on-site SEO, how search engines work, and off-site SEO (Chapters 2-4); paid search through paid search marketing (Chapter 5); paid referral through online advertising (Chapter 6); unpaid referral through earning external links (Chapter 4).
5. 5Choose target search phrases by balancing relevance, search demand, and competition. Estimate monthly volume with a tool such as Google's Keyword Planner. Estimate competition by searching the phrase and checking which sites appear and whether their title tags are perfectly optimized (the phrase in exact word order).
6. 6If the site is small, target long tail phrases (e.g., gold Rolex watch) instead of fat head terms (watches). Long tail terms have lower volume but are often more relevant and easier to rank for. One page can rank for several similar long tail terms, and combined they can add up to significant traffic and worthwhile revenue.
7. 7Earn unpaid referral traffic by getting links from external sites. These links help ranking and also send visitors directly, so each earned link pays twice.
8. 8Judge traffic against the other two objectives, conversions and revenue. Traffic is only the first of the three site objectives, so check that each source's visitors go on to convert.

The passages give the Belkin WeMo home automation line as a case for tracing sources. A visitor who clicks the Belkin WeMo ad on the right side of cnn.com's homepage counts as paid referral traffic, because Belkin paid to display the ad on another site. A reader of the gizmodo.com article about WeMo who clicks one of the three links in its first paragraph to the WeMo section of Belkin's website counts as unpaid referral traffic, because the links were not paid for. A user who searches internet marketing and clicks an ad at the top of the SERP is paid search traffic for the advertiser. If the same user clicks an organic result instead, that site gets unpaid search traffic. A person who saw a billboard with the URL and later typed it into the browser is direct traffic. One site can therefore hold all five source types at once, and each source needs its own tactic.

• If your site is small or new, then target long tail phrases with lower competition rather than fat head terms dominated by sites like Amazon.com or Overstock.com.
• If a phrase has no one searching it, then drop it, however relevant it is; relevance must be balanced with search demand.
• If you cannot fund offline advertising and your content does not merit repeat visits, then do not count on direct traffic; put effort into search and referral.
• If you want visitors from a specific placement quickly and can pay, then buy paid referral or paid search; if you want durable traffic, then build unpaid search through SEO and earned links.
• If a link earned for ranking also sits on a page people read, then count it as a referral traffic source as well as an SEO asset.

• Lumping all visits into one number: you cannot tell what is working. The tell is that you can state total visitors but not the paid versus unpaid split by source.
• Counting paid and organic search as one: ad clicks and organic clicks are different categories. The tell is that a spike in 'search' traffic stops when ad spend stops.
• Chasing fat head keywords with a small site: the tell is that you target a term like watches, competitors are behemoths, and you never rank.
• Targeting phrases with no search demand: the tell is a perfectly optimized page for a phrase the Keyword Planner shows nobody searches.
• Expecting direct traffic without offline ads or habit-forming content: the tell is that a plan lists direct traffic as a growth channel while the site has no billboard or TV presence and little repeat-visit content.

• ▸Direct: visits and likely driver (offline ad or habit) - record what produces typed-URL visits.
• ▸Paid referral: visits and placements - list sites where you pay for banner, text, or rich media ads.
• ▸Unpaid referral: visits and linking sites - list external pages that link to you without payment.
• ▸Paid search: visits and ad-triggering phrases - clicks on ads accompanying organic results.
• ▸Unpaid search: visits and ranking phrases - clicks on organic SERP results.
• ▸Target phrase candidates: list long tail and fat head terms under consideration.
• ▸Monthly search volume per phrase: record the Keyword Planner estimate.
• ▸Competition check: note who ranks and whether their title tags exactly match the phrase.
• ▸Downstream conversion by source: record the conversion rate for each source to judge visitor quality.

1. 1Name the primary conversion in one sentence before you measure anything. Match it to the site type: retail = a purchase; lead generation = a lead (a call or a filled-out lead form); search engine like Expedia = a booking; Google/Bing = an ad click-through; paywalled media like Netflix or Hulu-Plus = a paid account sign-up; social platform = new account creation; affiliate = a click-through; marketplace = both a seller posting and a buyer purchase.
2. 2List secondary conversions separately and never blend them into the primary rate. Examples from the source: account creation on a retail site (newegg.com offers it on the homepage and at checkout), email-list sign-up for a lead generation site or Expedia, purchase as a secondary conversion for an affiliate, account creation for a marketplace.
3. 3Write the formula in plain terms and keep it fixed: conversion rate = visitors who completed the target action ÷ total visitors. Decide whether the denominator is all visitors or only a segment, and keep it the same each time you report.
4. 4Split visitors by traffic source before judging the rate. The source defines five buckets: direct, paid referral (ads on other sites), unpaid referral (links), unpaid search (organic), and paid search (SERP ads). Calculate the rate for each, because one blended number hides which traffic converts.
5. 5For media sites with no clean conversion, pick the engagement action that produces ad exposure and measure it consistently. Options include pageviews per visit or videos watched. Treat this as a proxy and label it as one.
6. 6Check upstream signals when a rate looks weak. A bounce (a quick 'back' to the SERP) means the page did not deliver what the searcher sought. Slow load times frustrate users. An inaccurate title tag or meta description attracts the wrong visitors. Fix the mismatch before changing the offer.
7. 7Pair the rate with revenue per conversion. A site with good traffic and a good conversion rate can still improve by raising average order value, so track the rate alongside AOV (e.g., steer buyers toward a higher-priced version or add-ons).
8. 8For lead generation, remember the conversion is not the sale. Count the lead (call or form) as the conversion. Then track nurturing and closing separately, since closing happens outside the website.

A plumber builds a lead generation website. He does not sell anything online, so a purchase is the wrong conversion. His primary conversion is a lead: any phone call from the site, or a lead form where a visitor leaves contact information. His secondary conversion is an email-list sign-up, which lets him nurture people with periodic information when the sales cycle is long. His conversion rate is visitors who called or submitted the form ÷ total visitors, reported separately for organic search, paid search, and referral visitors. Revenue is a different measure. It depends on how many leads he closes into jobs, which happens outside the site. A software seller shows the revenue side: with a $99 standard and a $149 deluxe package, each visitor who converts to the deluxe version yields $50 more, so steering buyers to deluxe lifts revenue without changing the conversion rate.

• If the site does not sell directly (plumber, lead generation), then count a call or form submission as the conversion, not a purchase.
• If a site has more than one valuable action (marketplace buyers and sellers; retail purchase and account creation), then record each as its own conversion and rank primary versus secondary.
• If traffic is high and the conversion rate is acceptable, then work on average order value (higher-priced items, add-ons) rather than only chasing the rate.
• If organic visitors bounce quickly, then check title tag, meta description, content relevance, and load time before redesigning the conversion step.
• If the site is media without a paywall, then define engagement (pageviews or videos) as the conversion and state that it is a proxy; if it has a paywall, use paid sign-ups.

• Defining the conversion as a purchase for a non-retail site. The tell: a lead generation site reports a near-zero rate while the phone keeps ringing.
• Blending primary and secondary conversions into one number. The tell: email sign-ups raise the 'rate' while purchases or leads stay flat.
• Reading one blended rate across all traffic. The tell: a rate that never changes even though paid, organic, and referral traffic behave differently.
• Optimizing the rate while ignoring revenue per conversion. The tell: conversions rise but revenue does not, because buyers pick the cheaper option and no add-ons.
• Blaming the page for a mismatch that starts at the SERP. The tell: high bounces, a misleading or truncated title/meta description (about 67 and 155 characters), or slow load times.

• ▸Site type: retail, lead generation, search engine, media, social, affiliate, or marketplace.
• ▸Primary conversion: the one target action, in one sentence.
• ▸Secondary conversion(s): other valuable actions, such as account creation or email sign-up, tracked separately.
• ▸Traffic segment: direct, paid referral, unpaid referral, unpaid search, or paid search.
• ▸Visitors in segment: the denominator for the period.
• ▸Converters in segment: visitors who completed the primary conversion.
• ▸Conversion rate: converters ÷ visitors, per segment.
• ▸Revenue per conversion / average order value: record what each conversion yields and the tactic to raise it.
• ▸Upstream checks: bounce behavior, load time, and title/meta description accuracy for this segment.

1. 1Define the revenue event for your site type before measuring anything. Retail: each purchase. Lead generation: a closed deal, which happens offline. Media: advertiser payment for exposure to engaged users. Search engine: a booking (Expedia) or an ad click (Google, Bing). Affiliate: a purchase after the click. Marketplace: seller fees.
2. 2Break revenue into its two drivers: conversion volume and average order value (the money each conversion yields). Record both for the same period so you can see which one moves revenue.
3. 3Check the upstream objectives first. Revenue only follows traffic and conversions, so if traffic or conversion is weak, fix those before working on revenue per conversion.
4. 4Pick the average-order-value lever that fits your model. Retail: push higher-priced items or add-on purchases. Search engine: order results to favor the more lucrative bookings or ads. Media: learn your users so you can charge advertisers more, and learn which ad formats get more clicks. Social media: use what users divulge to help advertisers target likely customers. Affiliate: link to higher-priced items. Marketplace: sell listing upgrades and supplemental services on top of standard fees.
5. 5Apply the lever with other factors held constant. The affiliate rule only holds if the higher-priced products produce the same click-through and conversion rates, so confirm those rates did not fall.
6. 6For a lead generation site, do not try to optimize revenue per conversion on the site. The more lucrative deal is closed outside the website, so concentrate on traffic and conversion and measure what happens to the leads after they leave.
7. 7Compare revenue before and after each change, not traffic or conversion count alone. Revenue is the primary measure of success because the goal of the site is typically to make money.
8. 8Repeat on one lever at a time, so any revenue change can be traced to a single action.

A software company sells a standard package for $99 and a deluxe version for $149. The passages say it should do everything in its power to push visitors toward the deluxe version. Take a site where every conversion is a purchase. Each deluxe sale yields $50 more than a standard sale ($149 − $99), with no change in traffic or conversion rate. Shifting buyers from standard to deluxe raises average order value, and so revenue, from the same visitors. The same logic applies to Amazon, which invests heavily in product recommendation research to encourage add-on items with every purchase. In each case, revenue grows because each conversion yields more, not because conversions grow.

• If traffic is low or the conversion rate is poor, then fix those before optimizing average order value; revenue depends on both.
• If traffic and conversion rate are already good, then focus on increasing average order value, which is the remaining lever and deserves its own focus.
• If you run a lead generation site, then spend effort on traffic and conversion, because maximizing revenue means closing a more lucrative deal outside the website.
• If you run an affiliate site, then link to higher-priced items only when click-through and conversion rates hold steady.
• If you run a media or social platform, then raise revenue by learning more about users and which ad formats work, so advertisers will pay more for targeted audiences.

• Treating traffic or conversion counts as the success measure. The tell: reports celebrate visits and sign-ups while revenue stays flat.
• Ignoring average order value on a site with good traffic and conversion. The tell: strong volume, but every conversion yields the same small amount and no upsell or add-on exists.
• Trying to raise revenue per conversion on a lead generation site's pages. The tell: effort goes into on-site pricing tactics while lead quality and offline closing go unexamined.
• Pushing higher-priced affiliate links without checking rates. The tell: click-through or conversion falls after the swap, cancelling the price gain.
• Assuming money always defines success. The tell: you apply revenue targets to a site whose goal is not profit; the passages say the goal is typically, but not always, to make money.

• ▸Site type: retail, lead generation, search engine, media, social media, affiliate, or marketplace.
• ▸Revenue event: the specific action that produces money for your site type.
• ▸Traffic (period): visitors in the measurement period.
• ▸Conversions (period): count of the primary conversion in the same period.
• ▸Average order value: revenue divided by conversions, or the amount each conversion yields.
• ▸Total revenue (period): what the site actually earned.
• ▸Chosen AOV lever: the one action you will test (add-on, higher-priced item, ad format, upgrade fee, and so on).
• ▸Rate check: click-through and conversion rates before and after the change.
• ▸Result: revenue after the change versus before, and your decision to keep or revert.

1. 1Pick the target phrase before you touch the page. A page never ranks #1 in general; it ranks for a particular search. Score each candidate phrase on relevance, traffic, and competition. For an existing site, add a fourth factor: current ranking. A #1 rank on a phrase nobody searches is worthless.
2. 2Apply the relevance filter first. If the phrase does not describe what the page offers, drop it. The source's example is that a watch seller should not chase 'running shoes'. A phrase that fails here is out, however much traffic it has.
3. 3Assign one primary phrase (plus close variants) to each page, and choose it by the type of site. A new site picks phrases it hopes to earn a first-page ranking for eventually. An existing site picks phrases where it wants to improve its current position. The target is the first page of results, roughly the top 10 links.
4. 4Place the phrase deliberately in the elements the construct names: the URL, the title tag, the headers, the body copy, the image alt text, and the anchor text of links pointing to the page. Write each placement so it reads naturally to a person. Never hide keywords, for example in white text; the source names that as black-hat.
5. 5Audit the page for quality signals, because the algorithm now looks directly at the page as well as at links. Check three things. Is the page mostly content rather than advertising space (Panda penalizes a heavy ad footprint)? Is the content unique, not identical to content found elsewhere? Is it substantial, not thin? Fix any page that fails.
6. 6Keep link anchor text varied and natural. Do not make every link to a page use the target phrase. The source's benchmark: a natural fly-fishing profile might be about 80% brand/site-name anchors and 20% keyword anchors, while 95% keyword anchors suggests deliberate manipulation. The source states this for backlinks. Treat the same restraint as sensible for your own link text, though that is an inference.
7. 7Review site architecture and internal linking as part of the same pass. The source lists good site architecture and internal linking as white-hat practices and ends the chapter on technical issues that can impede rankings. Look for technical problems on any page that should rank but does not.
8. 8Run a user-first test before publishing. Every SEO change must not harm the user experience or try to deceive crawlers. If a change helps the engine but makes the page worse for a reader, revert it.

Take a fly-fishing retailer choosing phrases. It scores each candidate on relevance, traffic, and competition, and, because the site already exists, on current ranking. It drops any phrase unrelated to its products, in the same way a watch seller drops 'running shoes'. It keeps 'fly-fishing' variants that match what it sells and that people actually search. Next it assigns one phrase per page and places it in the URL, the title tag, a header, the body copy, and image alt text. Then it audits quality: no heavy ad footprint, no duplicate text copied from elsewhere, and no thin pages. Finally it checks link text. The source says that if 95% of the backlinks to a fly-fishing site use keyword anchors, a search engine is likely to assume deliberate manipulation, whereas a natural profile is closer to 80% brand-name anchors and 20% keyword anchors. So the retailer keeps its own link wording varied and does not push keywords into every anchor.

• If a candidate phrase fails the relevance test, then drop it regardless of traffic; relevance comes first.
• If a phrase has strong traffic but heavy competition, then weigh it against current ranking: for an existing site, favor phrases where you already have some position to improve, and for a new site, treat it as a long-term first-page goal.
• If an SEO edit would hurt readability or mislead a crawler, then reject it; the user comes first and deception is black-hat.
• If a page has little content, mostly ads, or text duplicated elsewhere, then fix quality before adding more keyword placement; Panda targets exactly these traits.
• If keyword anchors dominate the link text pointing to a page, then diversify toward brand and natural phrasing; a lopsided profile reads as manipulation.

• Targeting a phrase for its traffic without checking relevance. The tell: visitors arrive but the page has nothing to offer them, and the phrase would not describe your products.
• Treating 'rank #1' as a goal without naming the search. The tell: no keyword list exists, or the stated goal is ranking in general rather than for specific phrases.
• Stuffing or hiding keywords (e.g., white text) to force relevance. The tell: copy that reads awkwardly aloud, or text visible only in the code.
• Optimizing relevance while ignoring quality. The tell: well-placed keywords on pages that are thin, heavy with ads, or duplicate content found elsewhere, and rankings that stay poor.
• Over-optimizing anchor text. The tell: nearly every link to a page uses the exact target phrase, the pattern the source says triggers a manipulation assumption (95% keyword anchors versus a natural 20%).

• ▸Page URL and purpose: state what the page offers and the visitor action you want from it.
• ▸Candidate phrases: list the possible search phrases for this page.
• ▸Relevance check: for each phrase, write yes/no on whether the page truly delivers what it describes.
• ▸Traffic and competition notes: record how often the phrase is searched and how contested it is.
• ▸Current ranking (existing sites only): note today's position for the phrase, or 'none'.
• ▸Chosen primary phrase: pick one phrase for this page and justify it in one line.
• ▸Placement checklist: tick URL, title tag, header, body, image alt text, and anchor text, each written naturally.
• ▸Quality audit: record ad footprint, duplicate content, and thin content as pass or fail, with the fix for any fail.
• ▸User-first test: write one sentence confirming no change harms the reader or deceives crawlers.

1. 1Audit your link profile first. The chapter objective is to find the weaknesses in a site's link profile that keep it from ranking on relevant searches. List who links to you, which pages they link to, and what anchor text they use.
2. 2Classify every link you find as natural/organic (relevant, fits the article, works as a citation for the reader) or manufactured. Search engines favor the natural kind, and your goal is as many reputable sites as possible linking to you in that manner.
3. 3Check anchor text distribution. Natural profiles use a wide variety of words, including non-descriptive ones like "click here" or "here." If a high proportion of your backlinks use anchor text that perfectly targets your keywords, you look like the over-optimized sites Google Penguin penalized.
4. 4Remove or stop the low-value tactics: blog comment links, forum signature links, linky parties, free article directories, and secondary sites built only to link back. The source says these earn little credit, can trigger penalties, or get sites blacklisted.
5. 5Never buy links for SEO value, automate link creation, or hide links in other sites' code. These are named black-hat practices, and Google penalizes more easily than it forgives. Treat them as off-limits.
6. 6Earn links through repeated online marketing wins that give reputable sites a reason to cite you. The source states there is no substitute for hard work: success comes after multiple, successful, repeated wins that generate inbound links, not from a single tactic.
7. 7Fit link effort to page relevance. Search engines rank pages, not sites, so decide which page you want each earned link to point to, and match it to the keyword it is mapped to.
8. 8Keep blog comments only as a traffic play. A genuine comment that adds to the discussion can bring visitors, but you should not count it as an SEO asset.
9. 9Monitor Google Webmaster Central for guideline updates. Sincere sites get penalized by accident, so recheck your link practices against current recommendations. If penalized, fix the problem immediately and file a formal reconsideration request documenting the steps taken.

The source's Entrepreneur.com article shows what a natural link looks like. Its outbound links are outlined in green because they are relevant, fit naturally into the article, and work as citations the author gives the reader. Search engines read each one as an endorsement, and your off-site goal is to be the cited page in as many such articles as possible. Compare the black-hat history the source describes: a program that auto-generates thousands of pages pointing at one target, purchased links, and links hidden in hacked sites. These inflated original PageRank, and Google answered with algorithm updates. Penguin targeted the resulting pattern, in which a high proportion of backlinks carry anchor text that exactly matches the target keywords. The natural pattern is varied wording, including "click here." Practically, you earn editorial citations and let anchor text vary on its own, rather than requesting exact-match anchors.

• If a link tactic can generate thousands of links quickly (forum signatures, comment spam, article directories), then treat it as near-worthless and a penalty risk, and skip it.
• If a high share of your backlinks use exact-match keyword anchor text, then you are exposed to a Penguin-style penalty and should stop acquiring that pattern.
• If a link would exist only because you paid for it or built it yourself (secondary site, purchased links), then it is black-hat and should be avoided entirely, since penalties are easier to get than to recover from.
• If you leave a blog comment, then keep it substantive and treat the payoff as traffic, not SEO.
• If you are penalized, then fix the violation immediately and submit a documented reconsideration request.

• Chasing link volume: you use forum signatures, comment links, or directories to build thousands of links. The tell is a large link count with little ranking movement or a penalty notice.
• Over-optimized anchors: you request exact-match keyword anchors from every linking site. The tell is a backlink profile with almost no "click here" or branded variety.
• Joining link schemes such as linky parties or article directories: the tell is that sites in the scheme get penalized or blacklisted together.
• Treating link building as a one-off tactic: you expect results without repeated marketing wins. The tell is a stalled link count after one campaign.
• Accidental violations: you do not follow Google's webmaster guidance and get penalized despite good intentions. The tell is a sudden loss of rankings or removal from results for weeks or months.

• ▸Target page URL: the page you want to rank and that links should point to.
• ▸Target keyword phrase: the phrase mapped to that page.
• ▸Linking site and URL: where each backlink lives.
• ▸Link type: natural/organic citation, or manufactured (comment, forum signature, directory, secondary site, purchased).
• ▸Anchor text used: copy the exact clickable words.
• ▸Anchor category: exact-match keyword, branded, or non-descriptive ("click here").
• ▸Share of exact-match anchors: count exact-match anchors divided by total backlinks, to check for Penguin-style over-optimization.
• ▸Action: keep, stop, remove, or earn more like this.
• ▸Next earning move: the content or marketing win that would give reputable sites a reason to cite this page.

1. 1Name one target phrase for the page and write down what the searcher wants. The source's example: someone typing 'buy Nike running shoes' wants good webpages that sell Nike running shoes. A page that merely mentions the phrase does not serve that searcher.
2. 2Count what is on the page besides the phrase. The source's failing page contained only the words 'Monarch Butterfly Migration Patterns' and a butterfly image. Fail the page if a reader gains no information from it.
3. 3Put the meaning in words, not images. Search engines don't 'read' pages the way humans do and can't 'see' a picture to know what it shows. They examine characteristics of the words on the page, so anything you want counted must exist as text.
4. 4Match the phrase to the content in both directions. Relevance is how well the search phrase matches the page's content. Check that the phrase appears where the page actually discusses it, and that the page covers the topic the phrase promises.
5. 5Cover specific topics, not only broad ones. The source notes that broad lists fail someone looking for 'monarch butterfly migration patterns'. A narrow page that answers a narrow query fully is a valid asset.
6. 6Add enough substance to earn links. Google's 1998 algorithm combined relevance with popularity (the number and reputation of linking pages), and popularity served as an indicator of quality: the better a page, the more links it is likely to gather. Ask of every page: would another site link to this, as the Gizmodo article linked to Belkin's WeMo pages?
7. 7Build enough content to merit repeat visits. The source says most sites earn little direct traffic because few have enough content to generate habitual visits. Treat depth and breadth across the site as a traffic input, not decoration.
8. 8Revisit pages on a schedule. The source book itself is updated periodically so it reflects current best practice. Apply the same discipline: review each page, correct outdated material, and record the date.

Take the source's own test with two pages targeting 'monarch butterfly migration patterns'. Page 2 holds only that phrase as a heading and an image of a butterfly. Page 1 has actual content about the topic. Under the pre-Google engines of the mid-1990s, which ranked purely on how closely page words matched the search words, Page 2 would appear before Page 1. Google's 1998 algorithm added popularity (links from other pages, weighted by their reputation), so the page worth linking to gained ground. Run the same audit on your own page. Step one: the phrase is on the page, so relevance passes. Step two: remove the phrase and ask what is left. For Page 2, nothing is left, so quality fails. Step three: confirm that any meaning carried by an image is also in text, since the engine can't see the butterfly. The fix is to write the substance Page 1 already has. After that, the page can earn the links that signal quality.

• If a page's only on-page value is the target phrase and an image, then rewrite it with real content before any other optimization, because a phrase match alone is the failure the source describes.
• If the search phrase signals a transaction (for example 'buy Nike running shoes'), then the page must let the visitor do that thing. Informational text alone does not match the intent.
• If important meaning lives only in an image, then restate it in text, because search engines can't see pictures.
• If a page is a candidate for links, then judge it by whether a reputable site would cite it. A link from CNN.com counts for more than one from a small local news site.
• If a site expects repeat visits or direct traffic, then it needs enough content to merit them. If it lacks that, plan on search and referral traffic and invest in depth.

• Writing to match the phrase instead of the reader. The tell: the page has a headline and keyword repetition but no information, like the source's Page 2.
• Treating relevance as sufficient. The tell: a page is on topic but has no links or citations, which suggests no one considers it worth endorsing.
• Locking content in images. The tell: the page looks complete to a human but contains little crawlable text.
• Aiming only at broad topics. The tell: your pages compete on generic terms while specific queries, such as the monarch butterfly example, go unanswered.
• Publishing once and never revisiting. The tell: pages have no review date and contain information the source would call out of date. The source's own book is updated periodically for this reason.

• ▸Page URL and target phrase: record the single phrase the page is meant to rank for.
• ▸Searcher intent: one sentence on what the person typing that phrase wants to find or do.
• ▸Content without the phrase: describe what useful information remains if you delete the phrase. Write 'none' if nothing remains.
• ▸Intent match (yes/no): does the page deliver what the searcher wants, such as selling the product when the phrase says 'buy'?
• ▸Text vs. image check: list any meaning carried only by images and the text you will add.
• ▸Specificity: is the page a broad overview or a narrow answer? Note the narrower queries it should cover.
• ▸Link worthiness: name one reputable site that might link to this page and why.
• ▸Repeat-visit value: what on the site would bring a visitor back, and is it enough?
• ▸Last reviewed date and next review date: record when the content was last checked and when it will be rechecked.

1. 1Segment every paid dollar by source before setting budgets. The source passages list paid search (AdWords or Bing Ads), display ads, social ads, and affiliate links as separate traffic sources. Judge each on its own KPIs, such as revenue per visitor, rather than on aggregate site numbers.
2. 2Connect your analytics account to your paid search account before spending at scale. Connected, you see site behavior per ad and ad group. Unconnected, you see only click-through rate and must assume all ads perform roughly alike, which hides losers.
3. 3Classify every candidate keyword by searcher intent: buyer (exact product, such as 'best deal on thinkpad x200 tablet'), researcher ('tablet reviews'), or browser (curiosity terms like 'ideas'). Rate buyer keywords as highest value because they convert at a high rate when the ad and landing page match the product.
4. 4Allocate budget in tiers by intent. On a small, tight budget, fund only buyer keywords and make sure ads show only to that group. Add researcher keywords next, once you want to grow budget and reach. Treat browser keywords as brand-building spend that only large budgets can carry.
5. 5Run multiple ads per ad group. Each ad carries keyword relevance (dynamic keyword insertion where the searched term appears in the ad), one unique value proposition, and a call to action with an action verb. Add a time frame such as 'today' for urgency. Your ad is one of about 10 on the page, and every competitor has a UVP of its own.
6. 6Check competitors before you set bids. The passages name SpyFu, SEMRush, and KeywordSpy for online advertising competitive research. Use them to see which keywords and UVPs you are bidding against, especially well-known brands.
7. 7Compare revenue per visitor across sources on a regular cycle. If organic search beats display, the passages give two readings: SEO investment may pay off more than extra display spend, or your display targeting needs work. Diagnose which before moving money.
8. 8Judge social and display by return against the level of investment, not by reach. A modest-reach campaign aimed at a few key targets can earn a large ROI. Keep paid social consistent with your other media rather than running it in isolation.

A tablet retailer has a small online ad budget and sorts its keywords by intent. 'Best deal on thinkpad x200 tablet' is a buyer keyword: the searcher has finished researching and wants the exact model, so conversion is typically high when the ad and landing page name that product. This keyword gets first claim on budget. 'Tablet reviews' is a researcher keyword. Some searchers will buy later, but others are doing a class project, so a lower share are serious and the advertiser pays for more unqualified clicks. It waits until the retailer wants to expand budget and reach. Browser-style terms are skipped, since typically only eBay and Amazon buy those cheap clicks to build brand recognition. The same logic applies to a local movie theater. It does better on 'theaters in Birmingham' than on 'date night ideas', where the searcher is looking for blog posts, not tickets. For each funded buyer keyword, the retailer writes two ads with different UVPs and an action-verb CTA with a time frame, such as 'Save today'. It connects analytics to the ad account so it can see on-site behavior per ad, not just CTR.

• If budget is small and tight, then bid only on buyer-intent keywords and keep ads from showing to researchers or browsers.
• If buyer keywords are performing and you want more volume, then add researcher keywords next and expect a lower share of serious buyers per click.
• If revenue per visitor from organic search exceeds that from display, then either shift investment toward SEO or fix display targeting. Check targeting first, since both readings fit the data.
• If analytics and the paid search account are not connected, then connect them before reallocating. CTR alone assumes every ad performs about the same.
• If you lack a large budget and the keyword signals browsing ('ideas'), then do not buy it. Reserve cheap-click, brand-awareness buying for large online brands.

• Buying researcher or browser keywords on a thin budget. The tell: plenty of clicks and traffic, but a low share turn into sales because the searchers were not serious buyers.
• Judging channels on aggregate site KPIs. The tell: you cannot say whether paid search, display, or social produces the better revenue per visitor, so budgets move on guesses.
• Reading only CTR because analytics is not linked to the ad account. The tell: ads in a group look alike on CTR, but nobody knows which ads send visitors who actually behave well on the site.
• Copying competitors' ads or writing generic ads. The tell: your ad has no UVP distinct from the roughly 10 others on the page, or lacks an action verb, and CTR stays flat against brand-name rivals.
• Measuring social or display success by reach alone. The tell: you call a campaign a failure for reaching few people, even though a small, well-targeted audience could return a strong ROI.

• ▸Channel: paid search, display, remarketing, social, or affiliate. Use one row per channel.
• ▸Monthly spend: the budget committed to this channel.
• ▸Primary keywords or audience: the targeting, with each keyword tagged buyer, researcher, or browser.
• ▸Intent tier funded: buyer only, buyer + researcher, or all. Match it to your budget size.
• ▸Revenue per visitor: from analytics, for this source alone, not aggregate.
• ▸Analytics linked to ad account (Y/N): if N, note that you can see only CTR and cannot compare ads.
• ▸Ad variants running: the number of ads per ad group, with the UVP and CTA verb in each.
• ▸Competitor check: tool used (SpyFu, SEMRush, KeywordSpy) and the main competitor UVP found.
• ▸Decision for next cycle: keep, shift budget, fix targeting, or cut, with a one-line reason tied to revenue per visitor.

1. 1Write the conversion first. Name the one action you want (purchase, lead, add-on). Targeting precision means reaching people likely to take that action, not the most people.
2. 2Build a candidate list and apply the relevance filter before anything else. Relevance is the most important factor: a watch seller does not target 'running shoes'. Delete any phrase that fails this test, however large its volume.
3. 3Add the words your audience uses, not just your own. If a pergola company's customer does not know the word 'pergola', they may search 'gazebo'. Mark such synonyms 'relevant, to verify' until further research confirms them. Prefer specific phrases like 'Jacuzzi gazebo', which are highly relevant to the business.
4. 4Pull search volume for each surviving phrase and its related phrases. Google's Keyword Planner gives monthly estimates. A phrase nobody searches has no benefit, so cut it.
5. 5Estimate competition by searching the phrase yourself. Record (1) which sites appear (Amazon.com and Overstock.com-type behemoths are a warning) and (2) whether their title tags are perfectly optimized, meaning the phrase in exact word order.
6. 6Match phrase to site size. A small site targets long tail phrases ('gold Rolex watch', 'sterling silver earrings') rather than fat head terms ('watches', 'earrings'). Volume is lower, but relevance is higher and ranking is easier. Group similar long tail phrases on one page, since one page can rank for several.
7. 7Read the intent in the phrase. In the source story, Darrell searches 'wakeboards', then 'wakeboards under $400'. The second query is a shopper with a price in mind, which makes him the 'perfect customer' for Steve's shop. Rank phrases with purchase-ready modifiers above bare category terms.
8. 8For an existing site, add current ranking as a fourth factor and prioritize phrases where you already rank and can improve. Then check results: a page that earns a higher click share than expected for its position can be moved up, so test title tag, meta description and URL for keyword match and clarity.
9. 9Judge results with attribution in mind. Google Analytics defaults to last-click, which over-credits search. Use first-click, linear or time-decay views as a cross-check, knowing none is a gold standard.

A pergola company lists candidate phrases. 'Pergola' is relevant. A customer who does not know the word may search 'gazebo', so 'gazebo' is marked relevant pending research. 'Jacuzzi gazebo' is more specific and highly relevant to the business, so the company wants to rank for it. Running shoes fails the relevance filter and is dropped, as it would be for a watch seller. Next the team checks Keyword Planner for the monthly volume of each surviving phrase and related phrases, then searches each phrase to see who ranks and whether their title tags use the exact word order. If Amazon.com-sized sites dominate 'gazebo', the small company shifts to long tail variants of 'Jacuzzi gazebo', accepting lower volume for higher relevance and an easier path to page one. Similar long tail phrases are grouped on a single page so one page can rank for several. Later, results are read with last-click attribution's search bias in mind.

• If a phrase is irrelevant to what you sell, drop it regardless of volume, because relevance outranks traffic and competition.
• If Keyword Planner shows negligible monthly searches, drop it, since ranking for a phrase no one searches has no benefit.
• If the top results include behemoths with exactly optimized title tags and your site is small or new, move to a longer, more specific variant.
• If two long tail phrases share most of their words, target both with one page instead of building separate pages.
• If your reports use last-click attribution, discount credit given to search and look at first-click, linear or time-decay before cutting spend on other targeting.

• Claiming a page 'ranks #1' without naming the keyword. The tell: a ranking report with no phrase or volume attached, so the #1 may be for a search nobody runs.
• Choosing high-volume phrases that are not relevant. The tell: traffic rises but conversions do not.
• Chasing fat head terms with a small site. The tell: no movement from page-one-or-nothing phrases such as 'watches' while giants hold the results.
• Using only your own vocabulary. The tell: your keyword list lacks synonyms customers use, like 'gazebo' for 'pergola'.
• Judging targeting by last-click alone. The tell: search and search ads look like heroes while earlier touchpoints appear worthless, though they may have built the confidence that led to the final click.

• ▸Target conversion action: state the single action the phrase should lead to.
• ▸Candidate phrase: the exact search phrase, including any synonym a customer might use.
• ▸Relevance (pass/fail): would this phrase's searcher want what you sell? If no, stop here.
• ▸Monthly search volume (Keyword Planner): record the estimate for the phrase and any related phrases.
• ▸Competition check: list the top-ranking sites and note any behemoths.
• ▸Title-tag optimization of competitors: record whether they use the phrase in exact word order.
• ▸Fat head or long tail: label the phrase, and note which long tail phrases can share one page.
• ▸Intent signal: note any purchase-ready modifier (price, product type) in the phrase.
• ▸Current ranking (existing sites only): record the current position for the phrase.
• ▸Attribution lens used for results: last-click, first-click, linear or time-decay, with the time window you chose.