Google Analytics and GA4: Improve Your Online Sales by Better Understanding Customer Data

1. 1Pick the report that matches the action. Use Behavior (GA4: Engagement > Pages and screens) for broken pages and exit trends. Use Behavior Flow (GA4: a first Explore report) for step-by-step drop-off. Use Audience for demographics, geography, devices, and time of day. Use Acquisition (GA4: Traffic acquisition) for channel and ad spend.
2. 2Set the date selector first. Choose day, week, last 28 days, last 90 days, previous calendar year, or a custom range. Record the baseline for the metric you plan to change. The source frames analytics as a way to judge whether a change is needed and whether a change delivered the desired impact.
3. 3Hunt for broken steps in Behavior Flow. Start at the landing page and follow the red drop-off bars. A step where drop-off is 100 percent points to a failed page load or a button that does not function. On shipping pages it can point to shipping sticker shock or the estimated delivery date.
4. 4Cross-check exit pages. If one page shows a trend as the exit page, ask why customers leave there. Check that page for broken links before you redesign anything. Fix what is broken first.
5. 5Read bounce rate (GA4: engagement rate, its inverse) against the page's purpose. A high bounce on an e-commerce product page is a problem. A high bounce on a one-page festival page with a map is expected. Act only on pages where the purpose requires a next step.
6. 6Use audience and channel data to retarget spend. Read demographics (segment users by age group), countries, and time of day to decide who and when to target. The source notes time of day matters for staffing phone banks. Check which channel (direct, paid, affiliates, social, email) drove the e-commerce conversions, and move effort toward that channel.
7. 7Verify tracking before you trust the data. Hits count how often the tracking code fires. If new, returning, and active users do not report in Audience, troubleshoot the code. Remember that users who block cookies or clear them appear as new users every visit.
8. 8Make one change at a time, then re-run the same report over the same date window. Compare against the baseline. Keep the change only if the metric you targeted moved.

The source describes a behavior-flow scan on an e-commerce site. You open Behavior Flow (in GA4, you add it as your first Explore report) and follow users from the landing page. Most steps show a red drop-off bar, which is normal because some users find what they wanted. At one step, though, drop-off is 100 percent. The source lists the causes to test: a page that failed to load, buttons that do not function, or on a shipping page, shipping sticker shock or an estimated delivery date that drives customers off. You load the page on a phone and desktop and click every button. If something is broken, you fix it. If everything works, you review the shipping cost and delivery date display. You then compare the same 28-day window before and after the change. If drop-off at that step falls, the change worked. If it does not, move to the next cause on the list.

• If a step in Behavior Flow shows 100 percent drop-off, then treat it as broken and test page load and button function before changing copy or design.
• If a page shows up repeatedly as an exit page, then check for broken links on that page before assuming a content problem.
• If bounce is high on a product page, then act on it. If it is high on a blog post or single-purpose page with no required next step, then leave it alone.
• If drop-off occurs on a shipping page and nothing is technically broken, then examine shipping cost and delivery date as the likely cause.
• If the Audience report shows heavy traffic at certain hours or from certain age groups and countries, then shift staffing, ad targeting, and spend to match them.

• Treating every high bounce rate as a defect. The tell: you redesign a festival-style or single-answer page that was working as intended.
• Optimizing on bad data. The tell: new, returning, or active users do not appear in Audience, or returning visitors show as new, because the tracking code is not firing or cookies are blocked.
• Changing several things at once. The tell: the metric moves and you cannot say which change caused it.
• Skipping the before-and-after comparison. The tell: you have no recorded baseline or date range, so you cannot say whether the change produced the desired impact.
• Looking in old locations. The tell: in GA4 you cannot find bounce rate, time on page, or behavior flow because they are now engagement rate, average engagement time, Engagement, and an Explore report.

• ▸Insight source: the report and view where you saw the issue (for example, Behavior Flow or Exit pages; GA4: Explore or Pages and screens).
• ▸Page or step affected: the exact URL or checkout step.
• ▸Baseline metric and date range: the number and window before the change (for example, drop-off or exit rate over the last 28 days).
• ▸Suspected cause: broken page, nonfunctioning button, shipping cost, delivery date, wrong audience, or wrong channel or hour.
• ▸Tracking check: confirm the code fires and new, returning, and active users report.
• ▸Action taken: the one change made (fix link, remove step, speed up page, shift ad spend, retarget demographic).
• ▸Change date: when the change went live.
• ▸Result metric and same date window: the number after the change, measured the same way.
• ▸Decision: keep, revert, or test the next suspected cause.

1. 1Open Behavior reports in Universal Analytics, or Pages and screens under Engagement in GA4, and list the pages users reach. Page views and metric data live here, and this is where you start a performance audit.
2. 2Split sessions by device category (mobile vs. desktop). The source says analytics shows whether users engaged on a mobile device or a desktop computer. A page that works on desktop and fails on mobile is a responsive-design fault.
3. 3Check how quickly the site responds to a customer's request. The source names response speed as something digital analytics reports. Record the figure for each key page before you change anything, so you have a baseline.
4. 4Pull the Final URL report under Google Ads. It shows the exit page for visitors who arrived from AdWords. Look for pages where those visitors leave, such as the shipping page, and ask whether the cause is sticker shock or something broken.
5. 5Build a funnel for the checkout path: GA4 uses Explore with a customer behavior funnel, and UA uses goal pathing. Read the step-to-step drop-off and flag any step with an abnormal loss.
6. 6For each flagged step, test it yourself on a phone and on a desktop. Check whether the page design is confusing, whether the terms and conditions or shipping rules explain the exit, and whether anything fails outright.
7. 7Hand the specific failing step to your web developers with the funnel evidence. Fix it, then compare the same dashboard over time to see whether the change had the effect you wanted.
8. 8Keep the master view unfiltered and apply any filters only to a separate filtered view. Bad filters cause data loss, and that would corrupt the performance data you are relying on.

The source describes a checkout funnel where customers selected products, entered their address and credit card information, and accepted the terms of service. They reached the final review step, and 90% of them left there. The author calls this a significant problem and says it needs to go to the web developers. The analytics cannot say whether the cause is confusing page design, special shipping terms, or a broken step. The step is the finding. These are customers who had already committed their payment details, so the author says there is no clearer example of analytics driving revenue. The sequence is: find the drop-off in the funnel, identify the step, test it for breakage, hand it to developers, then recheck the funnel after the fix.

• If a high share of visitors exit at one page, such as shipping, then test the page for breakage before concluding it is a pricing problem. The source gives both explanations: sticker shock or something broken.
• If a drop-off happens late in the funnel, after the customer has entered payment details, then treat it as the top-priority fix. The 90% final-review loss is the source's clearest revenue example.
• If performance differs between mobile and desktop users, then investigate the responsive design of the affected pages first.
• If you are tempted to apply a filter to clean up data, then apply it to the filtered view only and never to the master view.
• If you change something on the site, then compare the dashboard over time to confirm it had the desired impact before moving on.

• Reading an exit-page report as a pricing problem only. The tell: you adjust prices while the page is actually broken, and the exit rate does not move.
• Reviewing only aggregate numbers and never splitting by device. The tell: sales or engagement look fine overall while mobile users quietly fail at a step.
• Never building a funnel for the checkout. The tell: you know total orders but cannot say at which step customers abandon.
• Applying filters to the master view. The tell: data goes missing or changes unexpectedly, and the original cannot be recovered.
• Making a site fix without a baseline. The tell: you cannot say whether the change helped, because there is no earlier comparison to read against.

• ▸Page or step name: write the page being audited, such as shipping or final review.
• ▸Device split: record mobile vs. desktop behavior for this step.
• ▸Response speed baseline: note how quickly the page responds before any change.
• ▸Drop-off at this step: record the percentage leaving here, from the funnel or Final URL report.
• ▸Suspected cause: choose confusing design, shipping terms or pricing, or a broken step.
• ▸Manual test result: note what happened when you walked through the step on phone and desktop.
• ▸Developer handoff: write the specific problem you gave the developers and the date.
• ▸Post-fix comparison: record the same metrics after the change to confirm the impact.

1. 1Define each desired outcome before touching settings. The passages name purchase, newsletter signup, eBook download, finished game, contact form completion, and blog interaction. Write each one as a single observable action.
2. 2Choose the tracking mechanism per outcome. In Universal Analytics, goals live under Admin > Goals. In GA4, goals are rolled into events under the configure tab. Events are custom actions (downloads, page loads, time on page, contact form completions, scroll depth, video watches, newsletter signups), and you build them in GA4 under configure.
3. 3For a destination goal, point it at the page that only appears after completion. The source example sets the checkout goal's destination to the final page, a thank-you HTML page. When Analytics sees that page with the code installed, it records a goal completion.
4. 4Set a goal value only when the value is knowable. A $0.50 blog referral gets $0.50. A SaaS with a fixed price or coaching courses with a set rate get that amount. Variable order values (the checkout example) get no goal value. Google does not calculate this number. You enter it, and it enables cost per conversion alongside conversion rate.
5. 5Confirm e-commerce is configured both on the website and in Analytics. It then reports under Conversions (Monetization > shopping behavior in GA4). Read revenue, average order value, and products purchased from there.
6. 6Build the funnel view. In GA4, go to Explore and add a customer behavior funnel (checkout behavior). List the steps: landing page, billing, shipping, payment, review, purchase completion. Read the drop-off at each step.
7. 7Use the reverse goal path to trace completers backward from completion to site entry. Check for unnecessary steps. Extra redirects during checkout are friction, so remove them.
8. 8Escalate the largest single-step drop to your web developer with the step named and the percentage stated. Check broken links on high exit pages, and re-read the funnel after the fix.

The source's e-commerce report shows $3,164 in revenue for the week, with an average order value of about $64–65. The checkout funnel (billing, shipping, payment, review, completion) shows that fewer than 10% of sessions at the final review step reach the completed transaction. That is a drop-off of more than 90%. These visitors had already chosen products, entered a shipping address, entered payment information, and accepted the terms of service. The source offers two candidate causes: a confusing review page design, or special shipping terms and conditions. Because every earlier step carried visitors through, the review step is the one to take to the web developer. The source stresses that fixing this one step could drive revenue substantially. Revenue and AOV tell you what happened. The funnel shows where it is being lost.

• If a step in the funnel loses a large share of visitors, such as 90% between review and completion, then fix that step before improving anything upstream.
• If the order value varies, then leave the goal value blank. If the price is fixed (SaaS, set-rate coaching), then enter that amount so cost per conversion appears.
• If the goal can be reached on a distinct confirmation page, then use a destination goal on that thank-you page. If it is an in-page action such as a form submit, video watch, or scroll depth, then use an event.
• If a page has a high exit rate and the funnel loses people there, then check for broken links and page problems first. If a high bounce rate sits on a content page with no required action, such as a one-page festival map, then treat it as expected. If it sits on an e-commerce product page, then treat it as a problem.
• If someone suggests adding steps or redirects to the checkout so customers see more products, then reject the redirects. Showing related products inside the cart view is acceptable, but extra steps are friction.

• Entering a guessed or arbitrary goal value. Google does not compute it, so the tell is cost-per-conversion figures that no one can trace to a real source.
• Pointing the destination goal at a page that visitors can reach without completing, or leaving off the code. The tell is a completion count that does not match the orders page or the newsletter signup notifications.
• Following old Universal Analytics paths in GA4. The tell is searching Admin for Goals, Behavior, or Conversions and finding nothing. Goals are now events under configure, behavior is Engagement, and goal paths and funnels are in Explore.
• Reading conversion rate or revenue without the step-by-step funnel. The tell is a healthy revenue number while the 90% final-step loss goes unseen.
• Treating bounce rate or exit pages as a verdict without context. The tell is trying to fix a page whose job is only to be read, or ignoring a high bounce rate on a product page.

• ▸Goal name: the single action, such as completed checkout, newsletter signup, eBook download, or contact form submit.
• ▸Tracking type: destination goal (Universal Analytics), or event under configure (GA4).
• ▸Trigger: the exact thank-you page URL or event action that fires the completion.
• ▸Goal value: a fixed dollar amount if one exists (for example $0.50 per referral), or 'none, order values vary'.
• ▸Funnel steps in order: list each page from landing to purchase completion.
• ▸Drop-off per step: the percentage lost between each pair of steps, taken from the funnel report.
• ▸Largest drop-off step: the one step to take to the developer.
• ▸Suspected cause: confusing design, shipping terms and conditions, broken link, or extra redirect friction.
• ▸Verification check: compare reported completions with the orders page or signup notifications.

1. 1Set up the account for sales measurement first. In the Google Analytics setup prompts, choose website, measure across devices and platforms, increase conversions, and analyze sales data because the site is e-commerce. Set the property's currency (U.S. dollars in the book's example), time zone (New York time), and business category (home and garden) correctly, because every revenue figure inherits these settings. The book advises against rushing this step.
2. 2Install the tag so purchases can be recorded. On WordPress, paste the measurement ID into the Jetpack settings and save. On Shopify, go to Online Store > Preferences and paste the ID into the Google Analytics account field. If you add the tag to the site code yourself, insert the full HTML tracking code in the head file, not just the measurement ID. Hand this to a developer if you are unsure. In GA4 the tag sits under data streams.
3. 3Verify the tag fires before you trust any revenue number. Check that new, returning, and active users appear in the Audience reports. If they do not, troubleshoot the tracking code before analyzing anything. A cookie-blocking or cookie-clearing visitor counts as a new user every time, so expect some distortion in user counts.
4. 4Open the e-commerce reporting and read the sales performance report. It collects transaction data, including the revenue from each transaction. Set the date selector (day, week, last 28 days, last 90 days, previous calendar year, or a custom range) and compare against a prior period. This tells you whether a site change moved revenue.
5. 5Build the checkout funnel and read the drop-off at each step. In GA4, go to Explore and add a customer behavior funnel. Walk the steps from product view through shipping, payment, review, and final transaction, and note the percentage lost between each pair of steps.
6. 6Find the largest drop-off and send it to your web developer with the exact step and percentage. Frame it as a specific defect, for example 'we see a 90% drop-off between review and completion.' Check whether the cause is a confusing page design or special shipping terms and conditions.
7. 7Trace revenue back to its source. Break traffic into segments such as paid search, organic, referral, social, and affiliates, and see which marketing campaigns turn into sales. Expand those and cut the spending that does not convert.
8. 8Re-measure after each fix over the same date window. Compare the funnel and revenue figures to the pre-change period to confirm the change had the intended effect.

The book's own funnel example shows a store whose customers get through the hardest steps and then quit. Only about 10% of viewers who looked at the item with shipping carried on. At the final review step, customers have already entered a credit card, payment information, and a shipping address, and accepted the terms of service. Fewer than 10% of them go on to the final transaction step, which is a drop-off of over 90% between review and completion. The author calls this the clearest example of Google Analytics driving revenue. These are customers who selected products, supplied all payment details, and then left. The next action is to sit down with the web developer and name the step and the 90% figure. Possible causes to check are a confusing review page design and special shipping terms. The author says correcting this could drive revenue substantially, and the fix is quick once you know it exists.

• If the funnel shows a steep drop at a single step, fix that step before adding traffic, because more visitors only feed the leak.
• If a drop-off occurs after the customer has already entered payment and shipping details, suspect the page design or the shipping terms and conditions, and escalate to the developer.
• If a marketing campaign's traffic converts to sales, expand it. If it does not, cut the spending.
• If new, returning, or active users do not show in Audience reports, fix the tracking code before interpreting any sales data.
• If you changed the site, compare revenue over the same date range before and after, and keep the change only if the numbers support it.

• Analyzing revenue on an unverified tag: user counts are missing or flat while sales exist. The tell is that Audience reports show no new, returning, or active users.
• Wrong property settings (currency, time zone, e-commerce option not selected), so revenue is misreported. The tell is figures in the wrong currency or sales reports that are empty.
• Judging performance on traffic alone and never building the funnel. The tell is high visits but flat sales with no step-level view of where customers leave.
• Reporting a drop-off to the developer without naming the step and percentage, so nothing gets fixed. The tell is a vague ticket like 'checkout isn't converting.'
• Mistaking cookie-driven new-user inflation for real growth. The tell is a high new-user count driven by visitors who block or dump cookies.

• ▸Date range analyzed: set the window and the comparison period.
• ▸Revenue and transactions (sales performance report): record the revenue figure and the transaction count for the window.
• ▸Tag verification: note whether new, returning, and active users appear in Audience reports (yes/no).
• ▸Funnel step counts: record the number of users at each step from product view through shipping, payment, review, and completion.
• ▸Largest drop-off: name the step pair and the percentage lost.
• ▸Suspected cause: list the candidates, such as confusing design or shipping terms.
• ▸Developer request: write the one-sentence issue with the step and percentage.
• ▸Top converting source: name the channel or campaign that produces sales, and the one to cut or expand.
• ▸Post-fix result: record the same funnel and revenue figures after the change.

1. 1Create the account with deliberate settings. Use a Google account (a dedicated company Gmail is fine, not required). Name the account, tick the technical-support option for access to an account specialist, and accept the terms of service. Optional data-sharing boxes can be unchecked without harm.
2. 2Configure the property to match the business before any data arrives. In the book's demo: name it 'demo account', set the country to the U.S., set the time zone to New York, set the currency to U.S. dollars, and pick the industry category (home and garden). Wrong time zone or currency corrupts every later report, so set them once and correctly.
3. 3Get the tag identifiers from the right place. In Universal Analytics: Admin > Setup Assistant > Install Tag/Tag Installation, with the tracking ID (the book's example format is UA-92800053-1). In GA4: Admin > Data Streams, where the measurement ID (example format G-XT7ERDBTD0) replaces the tracking ID.
4. 4Install the tag using the lowest-risk path for your platform. In WordPress, use the measurement ID (the easy route). Otherwise paste the HTML tag into the theme head section. The book also recommends Google Tag Manager over raw HTML. Shopify follows a similar pattern, and hand-coded sites need the head-section install.
5. 5Link Google Ads from Admin > Property, then follow the configuration prompts so Ads data flows into Analytics. If you run several websites or domains that each have their own AdWords account, the Google Ads linked-accounts view shows every AdWords account tied to the one Analytics account. GA4 has no standard keyword report, so do not expect keyword data in the old location.
6. 6Define what counts as a conversion. Build events (custom actions such as enhanced e-commerce or a 'contact us' interaction) and goals (purchase, ebook download, newsletter signup, game finished). In GA4, events live under Engagement, and goals and goal paths moved to the Explore tab, which adds premade and custom paths and funnels.
7. 7Verify capture with the Realtime report. It shows which pages visitors are on and how long ago they opened them. Open your own site, confirm you appear on the right page, then confirm that e-commerce and contact-us actions register.
8. 8Check the standard overview for gaps. Confirm that countries, devices, acquisition channels, new versus returning users, Google Ads, and goals all populate. An empty section points to a missing link, event, or goal.

The book builds a demo property for a home-and-garden website. You create the account 'Our Garden site' and the property 'demo account'. You set it to the U.S., New York time (the author is in East Tennessee), and U.S. dollars, with the category Home and Garden. You then open Admin > Setup Assistant > Tag Installation and copy the identifier, which looks like UA-92800053-1 in the old interface and G-XT7ERDBTD0 in GA4 under Data Streams. For the WordPress site you paste only the measurement ID, rather than editing the theme head section. Next you link Google AdWords under Admin > Property and follow the prompts. Finally you define events for enhanced e-commerce and 'contact us' and a goal such as a newsletter signup. The Realtime report then confirms that pages and actions are being recorded.

• If the site is on WordPress, use the measurement ID install; if it is hand-coded or on an unsupported platform, paste the tag into the theme head section.
• If you can use Google Tag Manager, prefer it over raw HTML, which the book calls the better option.
• If you run Google Ads, link it under Admin > Property before judging campaign performance, since without the link the Ads and keyword data does not arrive.
• If you run multiple websites or domains with separate AdWords accounts feeding one Analytics account, use the Google Ads linked-accounts view to confirm every account is attached.
• If you are on GA4, look for events under Engagement and goals or funnels under Explore, and do not expect a standard keyword report.

• Wrong property settings (time zone, currency, country): the tell is traffic-by-hour or revenue figures that do not match your actual business hours or sales.
• Tag installed incorrectly or not at all: the tell is an empty Realtime report while you browse the site yourself.
• Google Ads never linked: the tell is that the Google Ads and campaign reports are empty even though ads are running.
• No events or goals defined: the tell is that conversions, e-commerce, and contact actions cannot be reported, so campaigns cannot be tied to sales.
• Reading GA4 with Universal Analytics expectations: the tell is hunting for the old goals area or a keyword report that has moved or no longer exists.

• ▸Platform (WordPress, Shopify, hand-coded, other): note how the tag is installed.
• ▸Property type (Universal Analytics or GA4): determines where each setting lives.
• ▸Tracking ID / Measurement ID: record the identifier and where it was found (Setup Assistant or Data Streams).
• ▸Property settings: country, time zone, currency, industry category: confirm each matches the business.
• ▸Install method (measurement ID, head HTML, or Tag Manager): record the method used.
• ▸Google Ads linked (yes/no, which accounts): confirm under Admin > Property.
• ▸Events defined: list each custom event (e.g., enhanced e-commerce, contact us).
• ▸Goals/conversions defined: list each (purchase, download, signup).
• ▸Realtime verification result: note the test action, page seen, and pass/fail.

1. 1Create the Google account first. Every Google product requires one, so use a dedicated company Gmail to manage Analytics rather than a personal address. Name the account and keep the technical support option checked so you have access to an account specialist.
2. 2Fill out the property setup honestly and slowly. Set the property name (the book uses 'demo account'), country (U.S.), time zone (New York for East Tennessee), currency (U.S. dollars), industry category (home and garden), and business size (10 in the example). Google uses these answers to propose settings, and you can change them later.
3. 3State your goals in the setup prompts, because they shape which reports Google populates. For the book's e-commerce site, select: maximize user engagement, website (not app), measure across devices and platforms, optimize marketing cost, increase conversions, analyze sales data, and measure lead generation. Choose content monetization only if you place ads on the site. Then read and accept both terms of service boxes.
4. 4Install the HTML tracking tag on the website. It creates a cookie for each user and records where the user came from (the traffic source, such as Google AdWords or a Facebook referral), the device type, and how the user moves through the site. In GA4, the tag lives under Data Streams. Confirm data is arriving before you trust any report.
5. 5If you sell online, enable e-commerce tracking on both the website and in Google Analytics so the two exchange transaction data. Then set up goals (registrations, purchases, newsletter sign-ups). Without both sides configured, the e-commerce report stays empty.
6. 6Read the home page as a baseline. Use the date selector (day, week, last 28 days, last 90 days, previous calendar year, or custom range) to compare periods. Check the channel graph (direct, search, paid, affiliates, referral, email in GA4), then cycle through sources and mediums to see which channels grow and which do not.
7. 7Pull the four visibility layers from their specific reports. Audience: age, gender, market segment, location, browser language, and technology (browser, OS). Device: desktop, mobile, tablet (in GA4 under User > Tech). Acquisition: traffic by source (in GA4 under Traffic acquisition). Behavior: pageviews, sessions, and session duration.
8. 8Learn the GA4 relocations so you do not conclude data is missing. Referral traffic is rolled into all acquisition sources, Sessions is relabeled engaged sessions, goals and goal pathing moved to the Explore tab, and e-commerce channel reporting sits under Monetization > Ecommerce purchases with channels added under sources and mediums.

The book's demo home page shows how much one screen reveals once the tag is installed. On February 13th the channel graph shows six users from affiliates, a few from search, and 1,200 from direct traffic. Cycling to sources and mediums confirms 1,200 users from direct, plus some Google CPC (paid, the AdWords campaigns). Creator Academy on YouTube is listed as a top referral. It sent nobody on the 13th but sent at least one user on the 14th, which shows why you compare dates and do not judge a channel from a single day. The device graphs then split visitors into desktop, mobile, and tablet. With e-commerce enabled, the bottom of the page adds registrations, purchases, and goal data. The owner now knows where traffic comes from, on what device, and whether it converts, none of which was visible before the tag was installed. In GA4 the same channel data appears as a horizontal graph that includes email, and the date selector moves to the top right and changes every graph on the screen.

• If one channel (such as direct at 1,200 users) dominates the channel graph, then open sources and mediums before drawing conclusions. Compare against the rough-thirds trend, but judge 'normal' by your own business goals and data, not a fixed ratio.
• If you sell online and the e-commerce graph is empty, then check that e-commerce tracking is enabled on both the website and in Google Analytics, because they must communicate transactions.
• If you cannot find a report you used before, then check the GA4 note for its new location (Traffic acquisition, User > Tech, Monetization > Ecommerce purchases, Explore for goals) before assuming the data is missing.
• If a channel shows zero on one day, then widen the date range (28 or 90 days, or custom) before dropping it, since referrals like Creator Academy send traffic irregularly.
• If you want to judge marketing and SEO impact on the bottom line, then use the acquisition report with e-commerce selected and toggle users, sessions, and behavior metrics next to conversion rates, because it ties each channel to revenue.

• Installing the account but not the tracking tag (or placing it incorrectly). The tell: the dashboard stays flat or empty, and no source, device, or demographic data appears.
• Enabling e-commerce in only one place. The tell: the e-commerce and goal sections show nothing even though sales occur.
• Rushing property setup or entering false business details. The tell: Google proposes settings that do not match your goals, such as a wrong time zone, currency, or missing monetization and lead-generation options.
• Judging channels from a single day or an unchanged default date range. The tell: you dismiss a referral source that sent zero users on one date but one or more the next.
• Mixing up old and GA4 layouts. The tell: you report that referral data or Sessions is missing when it has been rolled into all acquisition sources or relabeled engaged sessions, or you spend time on the Treemap, a visual repeat of data you already saw, which the book calls not immediately useful.

• ▸Account and property: Gmail used, property name, country, time zone, currency, industry category, business size. Confirm each was entered accurately.
• ▸Goals selected at setup: list the use-case boxes checked (engagement, cross-device, marketing cost, conversions, monetization, sales data, lead generation).
• ▸Tracking tag status: installed on the site? Where the tag lives (GA4: Data Streams). Date you confirmed data was arriving.
• ▸E-commerce status: enabled on the website (yes/no) and in Google Analytics (yes/no). Goals defined (purchase, registration, newsletter).
• ▸Channel snapshot: for the chosen date range, record users per channel (direct, search, paid, affiliates, referral, email) and the top source/medium.
• ▸Device split: record desktop, mobile, and tablet shares from the device report (GA4: User > Tech).
• ▸Audience facts available: confirm you can see age, gender, market segment, location, browser language, browser, and OS. Mark any that are blank.
• ▸Behavior metrics: record pageviews, sessions, users, and session duration. Note any session count that exceeds users.
• ▸GA4 relocation notes: list any report you could not find and its new location, so the next review does not repeat the search.

1. 1Open the acquisition reports and set the date range first. In Universal Analytics, use the selectors at the bottom of the home-page graph (day, week, last 28 days, last 90 days, previous calendar year, or custom). In GA4, use the date selector in the top right corner; it changes every graph on the screen at once.
2. 2Read the channel graph to answer one question: how did visitors find the site? UA shows channels vertically on the home page. GA4 shows them as a horizontal graph and adds channels such as email. Record users per channel for the day or period.
3. 3Drill from channel into sources and mediums. A channel total says 'direct' or 'paid'; the source/medium view shows where the traffic comes from, such as Google CPC (the paid-per-click AdWords traffic, also written PPC).
4. 4Check referrals separately. In UA, go to Acquisition > All Traffic and open the referral view. In GA4, the Traffic Acquisition report already contains all sources, including referrals, so no extra tab is needed. List the top referring sites and note the dates they actually sent users.
5. 5Segment the traffic. Split paid, organic, direct, affiliates, display, referral and social. Compare them by device (desktop, mobile, tablet) or by conversions. In GA4, add segments through the 'add comparison' tab.
6. 6Separate users from sessions and flag returning visitors. Direct traffic is mostly returning users, so compare new versus returning users before you read a direct spike as new demand.
7. 7Verify the data is trustworthy before acting. Confirm the tracking tag fires (GA4: the tag sits under data streams), and that new, returning and active users appear in the audience reports. Missing numbers mean the tag needs troubleshooting.
8. 8Compare periods and write the finding in one line: which channels are growing, which are not, and which one you will act on. Check conversions per channel, not just volume.

The source book's own home-page view shows February 13th: six users from affiliates, just a few from search, and 1,200 users from direct traffic. Drilling into sources and mediums confirms the 1,200 are direct, with a small amount of Google CPC (the AdWords campaign) running. Referrals are listed next: Creator Academy on YouTube is one of the top referrals, yet it sent no traffic on the 13th. On the 14th it sent at least one user. Reading this, direct is the dominant channel by a wide margin, which fits the book's note that most direct traffic comes from returning users, so it is not proof of new-audience growth. Paid is small, affiliates are tiny, and a top referral can go a full day at zero. The next move is to look across 28 or 90 days, not a single day, to see which channels are growing and which are not before changing spend.

• If one channel (like direct at 1,200 users) dwarfs the rest, then check new versus returning users before calling it growth, because direct is mostly returning users.
• If a known top referral shows zero on a single day, then widen the date range to 28 or 90 days before judging it; one day is noise.
• If you run Google CPC/PPC and it shows only a few users, then look at it in source/medium detail and tie it to conversions before changing the budget.
• If you need paid versus organic or mobile versus tablet, then segment (GA4: add comparison) instead of reading the blended total.
• If new, returning or active users do not appear, or demographics are blank, then fix the tag or enable demographics in the admin property settings before trusting channel numbers.

• Judging a channel from a single day: you conclude a top referral is dead when it sent zero on the 13th and a user on the 14th. The tell: conclusions change when you move the date range.
• Treating direct traffic as new-customer success: the tell is a large direct number with a high share of returning users, or users who cleared cookies or used a VPN counted as new each visit.
• Looking for referrals in the wrong place after moving to GA4: the tell is a missing referral tab. Referrals sit inside Traffic Acquisition with all other sources.
• Reading channel totals without source/medium detail: the tell is that you cannot say whether 'paid' means Google CPC or something else.
• Trusting data from a misfiring tag or disabled settings: the tell is empty new/returning/active user reports or missing demographics and interest reports.

• ▸Report version (UA or GA4) and date range: record both so periods compare cleanly.
• ▸Channel: direct, organic search, paid search (CPC/PPC), referral, social, email, affiliates, display.
• ▸Users: users from that channel in the period.
• ▸Sessions: sessions from that channel in the period.
• ▸Top source/medium inside the channel: e.g. Google CPC, Creator Academy (YouTube).
• ▸New vs returning users: split, to interpret direct traffic.
• ▸Trend vs prior period: growing, flat or shrinking.
• ▸Conversions from the channel: purchases, sign-ups or goals, to judge quality as well as volume.
• ▸Action: one line naming the channel to invest in, fix or investigate.

1. 1Pull the right reports before interpreting anything. In Universal Analytics, use Behavior: Content Drill-down, Landing Pages, Exit Pages, and Behavior Flow. In GA4, most of this is condensed into Pages and screens under Engagement; landing and exit page reports are not standard, so build them in Explore explorations.
2. 2Start with landing pages. Every page is a landing page, because a visitor may enter on a product page, a blog page, or an ad destination rather than the homepage. Note which pages receive entries and compare each against the week or year before using the comparison selector.
3. 3Read bounce rate in the context of page purpose. A bounce is a visit that leaves without interacting. High does not automatically mean bad, but in most cases it signals a problem. A festival page with a map can bounce high because the visitor has no required next action. In GA4, read engagement rate instead; it is the inverse of bounce rate.
4. 4Check time on page, which is the average seconds visitors spend on a page, measured between the start of a session and the next page view or event. In GA4, this is renamed average engagement time. Add it under Users, then add page/screen host name, to see it per page.
5. 5Read exit pages against the intended path. The exits you want are the checkout page and the thank-you page after checkout. Exits in appreciable numbers on the homepage point to a problem with behavior flow or the site itself.
6. 6Use Content Drill-down to troubleshoot by site section. It shows the directory structure, which sections are most popular, and whether visitors can click through. Isolate the section that looks wrong before changing anything.
7. 7Slice the data with segments (paid, organic, direct, referral, social, mobile vs. tablet, new vs. returning) before concluding. Add the e-commerce segment to see which referral channels and devices produced purchases. In GA4, apply segments through the Explore tools.
8. 8Add events to see what the default reports cannot: downloads, contact form completions, newsletter signups, YouTube video plays, and scroll depth. In GA4, build them under the Configure tab. Re-read the engagement pages with these actions in view.

The source describes a store with e-commerce tracking and a contact-us event set on specific pages. Start with Landing Pages: the homepage drives most of the traffic, so it is the main entry. Next open Exit Pages. Visitors leaving on the checkout page or the post-checkout thank-you page is the healthy result. If many leave on the homepage, that is the warning the source names: a behavior flow problem. Then open Content Drill-down to find which site section is failing, and check bounce (GA4: engagement rate) and time on page (GA4: average engagement time) for the pages in that section. Finally, apply the e-commerce segment to referrals to see which channels and devices produced purchases. Whether a high bounce is a problem depends on the page: the source's festival-map page is expected to bounce, but a homepage that people exit is not. In GA4, run the landing and exit views through explorations.

• If a page's only job is to inform (such as a map page), then treat a high bounce rate as expected and look at whether it fulfills that purpose instead of trying to lower it.
• If exits concentrate on the homepage in appreciable numbers, then treat it as a behavior flow problem and use Content Drill-down to find the failing section.
• If exits occur mainly on checkout or the thank-you page, then the flow is working as intended and you do not need to fix it.
• If a high bounce rate appears on a page that should lead to another action, then treat it as a likely problem and investigate.
• If you need landing or exit data in GA4, then build it in Explore explorations rather than looking for a standard report.

• Treating every high bounce rate as failure. The tell: you are redesigning pages (like a map or information page) that have no required next step.
• Reading only the homepage. The tell: you ignore that any page can be the entry point, so blog and product pages go uncared for.
• Looking for a Universal Analytics report in GA4. The tell: you conclude the data is missing, when landing and exit pages, time on page, and bounce rate are renamed, moved to Pages and screens, or require explorations.
• Interpreting totals without segments. The tell: you cannot say which channel or device is purchasing or struggling because paid, organic, mobile, and tablet are blended.
• Counting visitors as exact people. The tell: new-user numbers look inflated, because someone who blocks cookies with a VPN or clears cookies appears as a new user on every visit.

• ▸Page URL and section: record the page and its Content Drill-down directory.
• ▸Page purpose: state whether it should inform, lead to another page, or convert.
• ▸Landing share: note how much of the entry traffic starts here, compared with the prior period.
• ▸Bounce rate / engagement rate (GA4): record the value and whether it fits the page's purpose.
• ▸Time on page / average engagement time (GA4): record seconds per page and compare with the prior period.
• ▸Exit share: record whether exits are at checkout/thank-you (expected) or elsewhere (flag).
• ▸Segment check: record which segment (channel, device, new/returning, e-commerce) differs from the total.
• ▸Interpretation: state in one sentence where customers engage, drop off, or struggle.
• ▸Next action: name the single change or event to add, with an owner.

1. 1Confirm e-commerce tracking is set up on the site and in Google Analytics so the two coordinate. Without that, the checkout report is empty or wrong. In Universal Analytics, e-commerce sits under Conversions. In GA4, it sits under Monetization > Shopping behavior.
2. 2In GA4, Shopping behavior covers only part of the picture, so build the funnel in Explore. Add a customer behavior funnel with the checkout steps as its stages. Use Funnel exploration for most of what the older checkout report showed.
3. 3Define the steps in order: landing page, checkout page, billing, shipping, payment, review, purchase completion. Keep the list to the steps customers actually pass through, so each drop-off maps to one screen.
4. 4Read the drop-off rate between each adjacent pair of steps (billing to shipping, payment to review, review to completion). Ignore the overall conversion rate for now. A single bad transition is the finding.
5. 5Check how many sessions had no cart activity. These are visitors who added nothing. They are a product-page or traffic problem, not a checkout problem, so keep them out of checkout diagnosis.
6. 6Flag any step where drop-off is extreme. Treat a loss at the last step as most urgent, because those customers already chose products and entered shipping and payment details.
7. 7Use the reverse goal path (customer steps from completion back to site entry) to see what completers did and where they came from. Look for added pages, redirects, or detours that slow them down, since friction is anything that slows a customer from their goal.
8. 8Take the specific step and its drop-off figure to your web developer for diagnosis and a fix. Re-check the same funnel after the change and compare it with the earlier period using the date selector.

In the book's example store, e-commerce reports $3,164 in revenue for the week, with an average order value of about $64 to $65. The funnel from landing page to checkout is described as very tight. The checkout behavior report shows the drop-off between billing, shipping, payment and review, and then how many sessions had transactions. The problem sits at the end. Customers have entered their credit card, payment information and shipping address, and reached the review page. Fewer than 10% of them go on to approve and complete the transaction. That is a drop-off of over 90% between review and completion. The author calls it the clearest example of Analytics driving revenue: these people picked products, entered payment details and then left. The action is to sit down with the web developer, name the step and the 90% figure, and have them find and fix the cause. The book gives no root cause, so this is where the diagnosis starts.

• If one step loses far more users than its neighbors, then treat that step as the defect and send it to the developer with the step name and the drop-off percentage.
• If the biggest loss is after payment and shipping details are already entered, then prioritize it first, because those customers have shown full purchase intent.
• If a proposal adds pages or redirects in checkout to show more products, then reject it. Show related products in the cart instead, and keep the checkout path on as few pages as possible.
• If many sessions show no cart activity, then investigate product pages and traffic quality rather than checkout.
• If you sell fixed-price items (a SaaS plan or coaching at a set rate), then set a goal value so Analytics reports cost per conversion. If order values vary widely, leave the value unset.

• Looking only at overall conversion rate. The tell: the number is low, but nobody can say which step is losing people.
• Treating a high bounce rate as proof of friction. The tell: you flag a blog or one-page festival page that is working as intended. High bounce is bad on an e-commerce product page, not everywhere. GA4 replaces bounce rate with engagement rate, its inverse.
• Using the old Universal Analytics path in GA4. The tell: the Goals and e-commerce reports are missing. In GA4, goals are folded into events, and funnels live under Explore.
• Adding checkout steps to boost product exposure. The tell: extra pages or redirects appear between cart and payment, and drop-off rises at those transitions.
• Reporting the drop-off without acting on it. The tell: the 90% review-to-completion loss sits in a report with no developer ticket, owner, or follow-up measurement.

• ▸Date range: set the period you are analyzing and the comparison period.
• ▸Funnel steps in order: list each step from landing page through purchase completion.
• ▸Sessions entering each step: record the count from the funnel report.
• ▸Drop-off % per transition: calculate the loss between each adjacent pair of steps.
• ▸Sessions with no cart activity: note them separately from checkout losses.
• ▸Worst transition: name the step with the largest drop-off and its percentage.
• ▸Intent signal at that step: record what customers had already entered, such as payment and shipping details.
• ▸Suspected friction: list extra pages, redirects or form issues you can see at that step.
• ▸Developer action and owner: write the specific request, the person responsible and the due date.
• ▸Post-fix re-check: record the same funnel's drop-off after the change.