Customer Accounting Creating Value with Customer Analytics

1. 1Define UNCER as perceived turbulence across three named domains: technology, market forces (competition and customer preferences), and regulation. Score each domain separately for your firm, since the source frames UNCER as exposure to shifts in market, technology or regulatory conditions, not a single vague 'volatility' feeling.
2. 2Anchor the scale to a relative benchmark. The source measures performance relative to main competitors and last fiscal year, and takes the UNCER item (Q18) from Ramani and Kumar (2008) and Song and Thieme (2006). Reuse the same wording and the same relative framing rather than inventing a new scale.
3. 3Place the uncertainty question late in any survey or diagnostic. The source put UNCER (Q18) and PERF (Q19) at the end so that earlier items on customer metrics, integration and sharing do not prime the answers.
4. 4Test uncertainty against architecture before touching performance. Check whether perceived uncertainty correlates with Marketing–Accounting & Finance integration (INTEG) and information sharing (SHARE). The source found r = 0.369 for integration and r = 0.428 for sharing (both p < 0.01), which supports the claim that more exposed firms align structure faster.
5. 5Treat competition as a separate dial from uncertainty. Record how competitive your markets are, because customer accounting is more likely in highly competitive environments, and customer-centric structure pays off 'only in specific competitive markets' (Lee et al. 2015).
6. 6Set the forecasting horizon from the environment. Estimate how long technology and competition stay roughly stable. Company.net uses 3 years as the window over which the business environment would not substantially change, and computes expected margin of the next 1,000 customers at month 1, month 6 and month 36.
7. 7Make the customer-analytics investment conditional on uncertainty and competition. Germann et al. (2013, 212 Fortune 1000 executives) found that more intense industry competition and rapid change in customer preferences strengthen the positive effect of analytics on performance. Pair that with the finding that data and IT support are necessary internal conditions.
8. 8Weigh benefits against coordination cost. A customer-centric structure raised customer satisfaction but added coordination costs (Lee et al. 2015). Commit to deeper restructuring only if the environment score justifies it.

Company.net, a subscription online service, launches in a new country and needs a CLV horizon. Its customer base arrives in cohorts: 25,016 customers in January (8,614 billed in month one, 13,437 in month two) and 38,862 in February (15,086 billed in the first period, 27,082 in the second). Because technology and competition are expected to hold roughly steady for about three years, the firm caps the horizon at 3 years rather than a true lifetime. It computes expected margin of the next 1,000 customers at month 1, month 6 and month 36, and accepts that later cohort-averages carry more error because fewer data points remain as cohorts mature. The environment's perceived stability therefore sets how far out the model reaches. In a more turbulent market, the same logic implies a shorter horizon, an inference rather than a figure from the source.

• If perceived uncertainty is high (technology, market or regulatory change), then raise Marketing–Accounting integration and information sharing; the source links UNCER to both (r = 0.369 and 0.428).
• If competition is intense or customer preferences shift fast, then expect customer analytics to pay off more (Germann et al. 2013); prioritise it, but only once data and IT support exist.
• If your market is not highly competitive, then be cautious about a costly customer-centric restructure; the structural benefit appears only in specific competitive markets and adds coordination costs.
• If the environment is expected to change substantially within the horizon, then shorten the CLV window; Company.net's 3 years is a stability judgment, not a constant.
• If customer orientation (STRAT) is high but integration is low, then do not assume the two move together; STRAT correlated with sharing (r = 0.418) but not with integration.

• Treating uncertainty as one score. The tell: you cannot say whether technology, market or regulation drives the number, so you cannot tie it to any structural change.
• Reading correlation as proof of causation. The tell: you claim uncertainty 'causes' integration, when the source reports only cross-sectional correlations from an exploratory survey.
• Restructuring around customers without checking competitive intensity. The tell: satisfaction rises but coordination costs and decision complexity grow with no performance gain.
• Deploying customer analytics without data and IT support. The tell: sophisticated metrics exist on paper, yet the competition-driven performance benefit never appears, because Germann et al. found these supports necessary.
• Using a fixed lifetime horizon in a fast-moving market. The tell: CLV forecasts extend well beyond the period the business environment can be assumed stable, and late-cohort estimates rest on thin data.

• ▸Technology change (1–7): rate how fast your core technology shifts; note the specific change.
• ▸Market/competition intensity (1–7): rate competitor pressure and customer-preference shifts relative to main competitors.
• ▸Regulatory change (1–7): rate expected regulatory shifts affecting customer data or pricing.
• ▸Overall UNCER score: average the three; record whether you used the Ramani–Kumar / Song–Thieme item wording.
• ▸Current INTEG level: rate Marketing–A&F integration of customer accounting information.
• ▸Current SHARE level: rate information sharing between Marketing and A&F.
• ▸Stable-environment horizon (years): state how long conditions hold; Company.net used 3.
• ▸Data and IT support in place (Y/N): confirm before relying on analytics for performance.
• ▸Expected coordination cost vs. benefit: note the customer-centric gain against added complexity and whether your market is competitive enough to justify it.

1. 1Define the churn event in one sentence before you measure anything: the exact state change (cancellation, non-renewal, lapse) and the period in which it is counted. The source treats retention as a behavioral state, so a fuzzy definition corrupts every downstream number.
2. 2Report retention and churn on the same customer base, using the language the source uses for subscription firms: new subscriber acquisitions, customer dropouts, revenue per subscriber (ARPU), and cost of customer acquisition. Retention then sits beside the other value drivers instead of floating alone.
3. 3Slice the base by acquisition cohort, following the CLV Scorecard and cohort analysis logic in Chapter 2. Track each cohort's survivors period by period. A blended churn rate hides whether newer cohorts leave faster than older ones.
4. 4Treat retention rate as one of the three CLV inputs, alongside margin per customer and cost of customer acquisition. Change one input at a time and see what it does to CLV, so you know how much value a retention gain is worth.
5. 5Convert churn into an expected relationship length. The source says every CLV model that values a customer's long-run contribution depends on this. Start with a parsimonious model whose parameters are easy to obtain, even in Excel (Fader and Hardie 2007b), before moving to statistical churn-probability models.
6. 6Discount the resulting future cash flow (or profit) at the firm's average cost of capital, per the classical CLV definition (Kumar 2008a). Undiscounted lifetimes overstate what retention is worth.
7. 7Use CLV as the ceiling on retention and acquisition spend: do not spend more than a customer's CLV unless you accept losing money. Apply it per customer or segment, and direct retention budgets to high-CLV customers first.
8. 8Report the retention and churn figures consistently over time (the source argues for MD&A disclosure of customer dropouts and related metrics). Consistent reporting lets managers and investors see how churn feeds sales, earnings and stock return.

The source gives no numeric churn case, so the numbers below are illustrative and not from the book. What the source does supply is the evidence that retention matters: a 2016 Forbes Insights survey found firms that increased retention spending over the prior 1–3 years had nearly 200% higher likelihood of increasing market share than firms that raised acquisition spending. A 2017 Salesforce report found 70% of consumers believe technology makes switching brands easier. Suppose a subscription firm tracks one acquisition cohort of 1,000 subscribers and sees 900 still active after the first period. That is a period-one churn of 10%. It then enters that retention rate with margin per subscriber and cost of customer acquisition into a discounted CLV calculation at its cost of capital. Suppose that CLV comes out at 240 per customer, a figure invented for this example. The source's rule then sets the ceiling: spending more than 240 to acquire or retain that customer loses money. A retention offer costing 60 is justified only if it raises expected CLV by more than 60. Comparing this cohort's churn curve with later cohorts shows whether retention is improving or worsening.

• If a retention or acquisition outlay for a customer or segment exceeds its CLV, then cut or redesign it, because CLV is the upper limit of what you should spend without losing money.
• If blended churn looks stable but newer cohorts decay faster than older ones, then treat it as a retention problem in the recent acquisition channel and investigate that cohort, not the average.
• If you must choose where to put retention budget, then rank customers by CLV and nurture the high-CLV ones first.
• If you lack data for a stochastic churn model, then start with a parsimonious model with easily obtained parameters (Fader and Hardie), and move to statistical churn-probability models once data supports them.
• If you are presenting churn to investors or the board, then show it with new subscriber acquisitions, ARPU and cost of customer acquisition, so churn is read as a value driver and not an isolated rate.

• Churn is defined differently across teams or periods (cancel date vs. lapse date). The tell is that the same cohort yields different retention numbers in different reports.
• Using one blended churn rate and skipping cohorts. The tell is a flat headline retention rate alongside falling revenue per acquired cohort.
• Computing lifetime value on undiscounted cash flows or with no expected-lifetime logic. The tell is a CLV that never changes when retention assumptions change, or that ignores the cost of capital.
• Treating retention as a satisfaction score instead of a behavioral state. The tell is that survey sentiment is high while customers keep defecting; the source treats churn as a behavior, a proxy for satisfaction and quality, not a substitute for it.
• Funding acquisition and retention with no CLV ceiling. The tell is spend per customer that is not tied to any forward-looking value estimate, so budgets are justified by volume rather than by lifetime value.

• ▸Churn definition: write the exact event and the period it is counted in.
• ▸Cohort and size: acquisition period and starting number of customers.
• ▸Active customers per period: count of survivors at the end of each period.
• ▸Retention rate per period: survivors divided by prior-period survivors.
• ▸Margin per customer: per-period margin or cash flow attributed to the customer.
• ▸Cost of customer acquisition: average cost to acquire one customer in this cohort.
• ▸Discount rate: the firm's average cost of capital.
• ▸Expected relationship length: estimate from the retention curve or a churn-probability model, and name the model used.
• ▸CLV and spend ceiling: discounted lifetime value per customer; the maximum acquisition or retention spend is this value.

1. 1Fix the vocabulary before any calculation. CP is revenue earned from a customer less associated costs over one period. CLV is the discounted future margin per customer, extended across the relationship. Customer equity (CE) is the value of the customer base as a firm-level asset. The source defines customer accounting as 'all accounting techniques that measure individual customer's and/or customer segments' contributions to firm profitability', so decide first whether your unit is the individual or the segment.
2. 2Build the customer data foundation. Go beyond transaction history to the sources the passages list: marketing interactions, clicks, web or mobile navigation patterns, online and offline behavior, and data from customer-owned connected devices. Tie every record to a customer ID so revenue and cost can be assigned to a person or segment.
3. 3Compute CP per period first. For each unit, take revenue earned in the period and subtract the costs associated with serving it. CP is the retrospective, auditable layer, and CLV inherits its cost logic.
4. 4Extend CP to CLV by projecting future margin per customer and discounting it. The source notes that CLV formulae come in sources and variations (its Chapter 2.2), so choose one variant explicitly. Write down the retention, margin and discount assumptions so the number can be reproduced and challenged.
5. 5Sum CLV across customers to get CE, and track it by cohort. The source's subscription-business application uses a CLV scorecard as a performance measurement system and a cohort analysis. Compare cohorts over time to see whether newer customers are worth more or less than older ones.
6. 6Create the Marketing–A&F interface. The source says CFOs and management accountants must coordinate with CMOs and marketing managers to monitor attraction, conversion and retention, and it flags a dearth of knowledge about how this works. Assign joint ownership of the CP/CLV definitions so Marketing does not own the data while Finance owns the cost logic.
7. 7Report selected customer metrics regularly to top management and the board as key performance indicators. Choose a few; do not report all of them.
8. 8Link rewards to the metrics, split by role. Reward sales/account managers for increasing customer equity. Incentivize relationship managers to extend the profitable lifetime duration of customers. Check the design for dysfunctional triggers before launch.

The passages give no customer-level CLV figures, so this example uses their organizational cases. Wells Fargo realigned around a two-tiered sales structure. A relationship manager supplies the external, interaction-oriented focus, and a product specialist supplies the internal technical input. Read through CLV, the relationship manager's job is the lifetime duration and margin of each customer. The product specialist's job is the offer that earns the margin. This changes the selling question from the product-centric 'How many customers can we sell this product to?' to the customer-centric 'How many products can we sell to this customer?' CP tells you whether a customer is profitable this period. CLV tells you whether it pays to keep investing in that customer. Texas Instruments shows the control side. It reportedly adopted a reward system with three marketing metrics: marketing gains for three consecutive years, efficient and timely services, and better understanding of customers. For your own design, borrow the principle from these cases: put the metric in the evaluation system, and match it to the role that can move it.

• If the decision is about one period's pricing, service cost or account pruning, use CP. If it is about how much to spend to acquire or retain a customer, use CLV.
• If a manager's job is acquiring or growing accounts, reward increases in customer equity. If the job is retention, reward the profitable lifetime duration, not duration alone.
• If you can only report a few customer metrics to the board, choose those that tie to the CLV drivers (attraction, conversion, retention) and that a named role can influence.
• If Marketing and A&F produce different customer numbers, stop using either for incentives until one shared definition exists.
• If the business is subscription-based, track CLV by cohort rather than as a single blended figure. The cohort view shows whether value is improving or eroding.

• Treating one period's CP as if it were CLV. The tell: customers are ranked by this year's margin, and early-stage customers with strong future margin are cut.
• Staying product-centric while reporting customer metrics. The tell: the question asked in reviews is still how many customers will buy a product, and no one owns the customer's total value.
• Metrics sit in a dashboard but not in the incentive system. The tell: managers hit their rewarded targets (e.g. sales volume) while CLV or CE stagnates.
• Rewarding lifetime length without regard to profitability. The tell: retention rises while margin per retained customer falls, because relationship managers are not held to profitable duration.
• Marketing owns the data and Finance owns the costs, with no interface. The tell: the CLV formula's assumptions are undocumented, and the CFO and CMO cannot reproduce each other's figures.

• ▸Unit of analysis: individual customer or named segment; state which and why.
• ▸Period CP: revenue minus associated costs for this unit over the stated period; list each cost line included.
• ▸CLV variant: name the formula variant used and its source, since CLV formulae vary.
• ▸Projection assumptions: retention, margin trend and discount rate, each written out with its basis.
• ▸Data sources: list transaction, interaction, click/navigation and device data feeding the calculation, each tied to a customer ID.
• ▸Cohort definition: how customers are grouped (e.g. by acquisition period) for scorecard and cohort tracking.
• ▸Joint owners: the named Marketing and A&F owners of the definitions and the numbers.
• ▸Board-level KPIs: the few customer metrics reported to top management and the board, and how often.
• ▸Role-linked incentive: account manager tied to customer equity; relationship manager tied to profitable lifetime duration; note any dysfunctional-trigger risks.

1. 1Test the business model first. The source's parsimonious model targets subscription-based enterprises (SBEs): firms that charge for access to a product or service per period. If the firm is non-contractual, flag that CLV is harder to compute and treat any CE figure as lower-confidence.
2. 2Fix the definition before computing: CE = the sum of customer lifetime values (CLV) of current and future customers. Decide explicitly whether your number covers the installed base only or also customers not yet acquired, and label it accordingly.
3. 3Collect the SBE value drivers named in the source: customer base, revenue per subscriber, service cost per user, customer turnover (dropouts), new subscriber acquisitions, and cost of customer acquisition. For an outside valuation, pull these from public disclosures. For an internal one, pull them from billing and CRM data.
4. 4Estimate CLV by observing behavior, not assuming it. Use cohort analysis: group customers by acquisition period, track their purchase and retention behavior over time, and link that to CLV and firm performance. This closes the gap between theoretical CLV models and practice.
5. 5Aggregate cohort or per-user CLVs into one CE figure. Keep the driver breakdown visible (base, revenue, cost, turnover, acquisition) so the total can be decomposed and challenged.
6. 6Place CE outside the IFRS/US GAAP balance sheet. Customer franchise is a key intangible asset that those statements do not capture. Present CE as a supplementary valuation measure and never as a restatement of reported assets.
7. 7Wire CE into control. Report selected customer metrics among the KPIs sent regularly to top management and the board. Link evaluation and rewards to them: for example, reward sales/account managers for increasing customer equity and relationship managers for extending profitable customer lifetime.
8. 8Disclose externally. The authors argue the MD&A should carry new subscriber acquisitions, revenue per subscriber, customer dropouts, and cost of customer acquisition, so investors can rebuild CE themselves.

The source gives no numeric CE calculation, so the example below shows the structure and a real incentive case without inventing figures. Take a subscription Internet or media firm. Its CE rests on four public inputs: subscribers, revenue per subscriber, service cost per user, and dropouts, with new-subscriber acquisition and acquisition cost feeding future customers. Group subscribers into acquisition cohorts and observe each cohort's retention and spend. Each cohort's margin per user (revenue minus service cost), carried across its observed lifetime, gives a CLV, and the sum across current and expected future cohorts gives CE. The balance sheet shows none of this. Link it to management behavior the way Texas Instruments reportedly did, with a reward system built on three marketing metrics: marketing gains for three consecutive years, efficient and timely services, and better understanding of customers (Kumar 2008b). Under the source's logic, a CE-based variant would reward account managers for raising CE and relationship managers for extending profitable lifetime.

• If the firm sells subscriptions or other contractual per-period access, then compute CE from the disclosed drivers. If it is non-contractual, then state the retention-estimation challenge up front and discount your confidence.
• If you are valuing from outside the company, then limit yourself to publicly disclosed metrics (subscribers, revenue per subscriber, dropouts, acquisition cost) and keep the model parsimonious.
• If cohorts show different retention or spend, then value by cohort instead of using one blended CLV, because observed purchase behavior is the input.
• If CE is used in incentives, then reward both growth in CE (account managers) and profitable lifetime extension (relationship managers) so that neither acquisition nor retention is neglected.
• If the board or investors see only GAAP figures, then add CE and the four customer metrics to the reporting pack (MD&A for external audiences) rather than waiting for accounting standards to change.

• Treating CE as a recognized asset or booking it against GAAP equity. The tell: CE appears in a balance-sheet reconciliation instead of a supplementary disclosure.
• Counting only current customers while calling the result customer equity. The tell: no assumption on new-subscriber acquisition or its cost appears in the model.
• Using one blended CLV that ignores cohort differences. The tell: the CLV does not change when the acquisition mix or the retention curve changes.
• Applying a subscription model to a non-contractual business without addressing retention uncertainty. The tell: churn is entered as a fixed number with no behavioral evidence.
• Reporting CE but leaving rewards on transactional or product metrics. The tell: managers hit sales targets while CE and customer lifetime stagnate, a dysfunctional incentive the source warns about.

• ▸Business model check: Confirm for-fee-per-period access (contractual) and note any non-contractual elements that reduce confidence.
• ▸Customer base: Record current subscribers by cohort and period.
• ▸Revenue per subscriber: Enter average revenue per subscriber for each cohort.
• ▸Service cost per user: Enter the cost to serve each subscriber, to derive margin per user.
• ▸Customer turnover/dropouts: Record observed dropouts per cohort per period, the source for lifetime duration.
• ▸New subscriber acquisitions and acquisition cost: Record gross additions and the cost of customer acquisition, used for future customers.
• ▸CLV per cohort: Derive from margin per user over observed lifetime, with the method noted.
• ▸Customer Equity total: Sum CLVs of current and future customers, showing both components separately.
• ▸Control link: Name the KPI reported to the board and the manager reward tied to it (e.g., CE growth for account managers, profitable lifetime for relationship managers).

1. 1Fix the outcome set before you measure anything: profitability against peers, profitability against your own prior years, and, if you are listed, future stock returns. The passages treat these as the results of customer-centric value creation, so each one needs a named metric, a peer set, and a time window.
2. 2Record a baseline from traditional accounting variables (asset- and earnings-based). Lev's foreword argues these have sharply declined in usefulness for investment analysis. The baseline shows what customer-based measures add. It is not meant to be the final judge.
3. 3Value the customer base as an asset. Build customer lifetime value (CLV) for individual customers or segments, then aggregate to customer equity (CE). Document which CLV formula variant you use, because the book treats formulae as having sources and variations. Keep that choice constant across periods so the series is comparable.
4. 4Report CE through a CLV scorecard and a cohort analysis. The scorecard serves as a performance measurement system. Cohorts show whether customers acquired in a given period are worth more or less over time. Together they tie the outcome to specific customer groups rather than to one blended company figure.
5. 5Build the data infrastructure to support that valuation: one integrated, organization-wide view of each customer, regardless of product bought or channel used, at the individual-customer and transaction level. Without it, CE is unreliable and outcome changes can't be traced to customer behavior.
6. 6Audit the three legs of organizational architecture: decision rights, performance measures, and compensation and incentives. Check that people who decide on retention, pricing, and customer segments (a) have the CLV information and (b) are rewarded for acting on it. The source says misalignment among the legs is likely to show up as lower organizational performance.
7. 7Test resource allocation against outcomes. Compare retention and acquisition spending. A 2016 Forbes Insights survey found that firms increasing retention spend had nearly a 200% higher likelihood of raising market share than firms prioritizing acquisition. Use your own cohort data to see whether that pattern holds for you.
8. 8Close the loop on a fixed schedule. Compare CE movement, market share, and profitability (versus peers and prior years) period by period. If you are public, check whether your CE-based valuation anticipated later stock returns better than the accounting baseline did.

Take the Forbes Insights 2016 survey cited in the source: firms that raised retention spending over the previous one to three years had nearly a 200% higher likelihood of increasing market share in the last year than firms that spent more on acquisition. Here is how to turn that into a test on your own firm. First, record your retention-to-acquisition spend split for the last three years. Second, build a CLV cohort table, with one row per acquisition year and CE by cohort. Third, line up market share and profit versus peers and versus last year. If retention-heavy periods coincide with rising cohort CLV and share gains, the survey pattern holds for you. If not, check the architecture first: whether retention owners had decision rights, CLV-based measures, and matching incentives. The book's premise is that value creation depends on that coherence. Treat the survey as a hypothesis to test, not proof for your firm, because the source reports a likelihood and not a causal effect.

• If CE rises but profit versus peers does not, then check whether customer-level data is integrated across products and channels; a fragmented view makes the CE figure unreliable.
• If a CLV scorecard exists but spending and behavior do not change, then examine decision rights and compensation before refining the metric; the architecture needs all three legs aligned.
• If cohort CLV is falling in newer cohorts while headline revenue is growing, then treat acquisition-led growth as a warning and shift spend toward retention.
• If accounting earnings and CE tell opposite stories, then weight CE for forward-looking judgments and the accounting baseline for historical reporting, since the source questions the relevance of traditional measures for investment analysis.
• If you cannot name the CLV formula variant you used, then stop comparing periods until it is fixed; mixed formulae make trends meaningless.

• Judging performance on traditional earnings and asset measures alone. The tell: the customer base is the largest value-creating asset yet appears nowhere in your performance reporting.
• Building a scorecard without incentives. The tell: managers have CLV dashboards but bonuses still follow volume or market-share targets.
• Siloed customer data. The tell: marketing and finance quote different customer counts or values, and no one can give a view independent of product or channel.
• Changing the CLV formula between periods. The tell: CE jumps or drops at the point of a methodology edit rather than a customer event.
• Treating survey statistics as a guarantee. The tell: you cut acquisition spend citing the 200% finding without checking your own cohort data.

• ▸Outcome metrics and peer set: list profitability measures, peer group, prior-year comparator, and (if listed) the stock-return measure.
• ▸Traditional baseline: record the current asset- and earnings-based figures for the same period.
• ▸CLV formula variant: state the formula and its source, and freeze it for all periods.
• ▸Customer equity this period and last period: aggregate CLV, with the change.
• ▸Cohort table: one row per acquisition cohort, with CLV by period.
• ▸Retention vs acquisition spend: amounts for the last 1–3 years and the ratio.
• ▸Architecture check: for decision rights, performance measures, and compensation, mark aligned or misaligned and name the owner.
• ▸Data integration status: record whether one customer view spans all products and channels (yes/no/gaps).
• ▸Review verdict and next action: compare CE movement against outcome movement and state one change to make.

1. 1Rewrite the selling question first. Replace 'How many customers can we sell this product to?' with 'How many products can we sell to this customer?' Apply it in planning documents, sales targets and product reviews. The source treats this as the core test that separates a product-centric firm from a customer-centric one.
2. 2Audit your orientation against the Table 1.1 rows: basic philosophy (sell products vs serve customers), business orientation (transaction vs relationship), positioning (features vs benefits for individual needs), organizational focus (internal and marketing-department-owned vs external, with employees as customer advocates), and selling approach. Mark each row product-centric or customer-centric with one piece of evidence. The rows you cannot evidence are your gaps.
3. 3Build one organization-wide view of the customer base, irrespective of the products purchased or channels used. Collect, track and integrate data at the individual-customer and transaction level. Budget for this as a substantial IT-related investment commitment, because the source treats it as the foundation of the performance measurement system.
4. 4Measure what each customer adds to your bottom line, not only what the customer values. Use lifetime value of customers (CLV) per customer and customer equity (CE) in aggregate. Use these figures to guide tactical and strategic resource allocation.
5. 5Segment explicitly and quantify the segments. Aim operations at delivering the greatest value to the best customers for the least cost. Assign each segment its own service level and spend ceiling based on its measured value.
6. 6Answer the source's strategic questions per customer or segment: volume or margin? How many products can we sell? How do we build profitable relationships over a long time span? Which segments and processes are more productive? Can we influence this customer to interact differently and more profitably? Record the answers as decisions with owners.
7. 7Build interaction response capacity: the degree to which you deliver successful products and services by exploiting the feedback of a specific customer. Route individual customer feedback into product and service decisions rather than into a marketing report.
8. 8Align the three architecture elements together: organizational structure, performance measurement system and reward system. The source says they are jointly determined and complementary. Do not change the metrics without changing who owns customers and what gets rewarded.
9. 9Test budget priorities against the retention evidence. In the 2016 Forbes Insights survey, firms that raised retention spending had nearly a 200% higher likelihood of increasing market share than firms that raised acquisition spending. Review where your marginal dollar goes.

The source describes a real shift in advertising that shows the logic. Television advertising pursued 'brand awareness', which is product-centric broadcasting. Internet advertisers paid only when a user clicked through to their website, so spend became 'direct and measurable' customer acquisition. This works because data reveals actual customer behavior rather than intentions, and behavior is available for the entire customer population, so sampling is unnecessary. A customer-centric firm takes the next step: it links each click-through customer to the product and channel history held in one unified customer record. It then asks what that customer adds to the bottom line, using CLV. Spend is judged by lifetime value, not by clicks. The retention evidence reinforces this. The Salesforce 2017 'State of the Connected Customer' report found 70% of consumers believe technology makes switching brands easier. The Forbes Insights survey found that firms increasing retention spending had nearly a 200% higher likelihood of gaining market share than firms favoring acquisition spending. Together these suggest weighting spend toward retaining the customers whose measured value is highest.

• If a customer or segment's CLV is high, then invest in retention and cross-selling (more products to this customer). If CLV is low, then lower the service cost to match its value rather than serve it at the same level as the best customers.
• If your data is held by product line or channel, then fix the unified individual-customer record before launching customer-centric incentives or campaigns.
• If a decision is framed as 'how many units can we move?', then restate it as 'volume or margin with this customer?' before approving it.
• If you change performance measures to customer metrics, then change the reward system and ownership in the same cycle. Otherwise structure and rewards will keep pulling toward product targets.
• If budget is split between acquisition and retention with no value evidence, then shift the next increment toward retention and track its effect on market share and CLV.

• Treating customer-centricity as a marketing-department initiative. The tell: customer relationships appear only in marketing objectives, and other functions' employees are not described or rewarded as customer advocates.
• Serving every customer equally well. The tell: no quantified segmentation, and the best customers get the same service level and cost as the least valuable.
• Measuring satisfaction or sales without measuring the value the customer adds. The tell: no CLV or customer-equity figure exists, and resource allocation is justified by revenue or volume.
• Fragmented customer data. The tell: different teams cite different customer counts or histories for the same person because product and channel systems are separate.
• Changing one element of the architecture alone. The tell: customer metrics are introduced but structure and rewards still follow product P&Ls, so behavior does not change.

• ▸Table 1.1 row scores: For each of the five rows (philosophy, orientation, positioning, organizational focus, selling approach), write 'product' or 'customer' plus one line of evidence.
• ▸Unified customer view: State whether one record exists per customer across all products and channels, and at what data level (individual/transaction). Name the gap and the owner.
• ▸Value metric in use: Record whether CLV and CE are computed, how often, and which decisions use them.
• ▸Segment definitions: List segments, their measured value and the target service level and cost per segment.
• ▸Strategic question answers: For each priority segment, answer volume vs margin, number of products to sell, long-term relationship plan and behavior to influence.
• ▸Feedback loop: Describe how individual customer feedback reaches product and service decisions, and who acts on it.
• ▸Architecture alignment: Record the current structure, performance measures and reward system, and the change needed in each so all three match.
• ▸Acquisition vs retention split: Record current spend on each and the evidence for it. State the next shift planned.
• ▸Review date and owner: Name who reviews the sheet and when.

1. 1Build the availability inventory first. List the metrics: number of customers, usage/traffic, gross customer additions, net acquisitions, ARPU, churn/retention, cost of service, cost of acquisition (CoA), and Customer Equity. Score each 1 if it is actually produced and 0 if not, the way the source's AVA dummies do. A metric someone could compute does not count.
2. 2Split churn into instant and historic, and add yield and margin. Company.net's internal report carries instant churn, historic churn, yield and margin. Cut them by dimensions such as country, subscription flight and telco. The report fits a subscription business where every transaction is logged online.
3. 3Set a monitoring cadence for the leading indicators. The source says these forward-looking drivers must be watched daily to confirm results fit the strategic objectives. Assign an owner to each metric and compare each reading to the strategic target.
4. 4Cut the metrics by marketing campaign and acquisition channel. Track CLV and CoA for an average customer acquired through Facebook or Google, and compare against Yahoo or Bing. Check retention behavior (instant and historic) per channel, because channels churn differently.
5. 5Compute cohort margin per thousand customers (MpK). Average each month's gross margin across the cohorts that have data for that month. Use a minimum of 2 and a maximum of 12 months of past cohorts, whatever your data supports. Expect fewer data points for later months, because a cohort that started in month one has 5 months and one that started in month two has 4.
6. 6Normalize before forecasting. Check the standard deviation of each averaged month. Where outliers bias the average, remove them and recompute.
7. 7Extrapolate beyond observed data. Take the last observed month-on-month decrease, apply it, then shrink the rate by a fixed factor each month (95% in the source example) to reflect declining churn. Extend the projection to the horizon you need (the source table runs to month 36).
8. 8Read every metric movement as a trade-off. Before reporting an improvement, ask what moved elsewhere: a price rise (ARPU) predictably raises churn, and more marketing spend raises CoA and may raise churn by bringing in lower-quality customers. Use the scorecard to test these cause-and-effect links between actions, customer metrics (churn, CoA, margin, yield) and financial results (CLV, CE).
9. 9Extend availability into use. For each available metric, record whether it is used for planning and control, for targets and rewards at each level (top management, business unit managers, sales managers, middle managers, customer care), for external reporting, and for valuation. Also record who owns the data and how accurate it is.

The source projects margin per thousand customers acquired today. For month 1, the five cohorts with data show gross margins of 2238, 2523, 3828, 3457 and 2482, which average to about 2905.65 (shown as $2906). For month 2, four cohorts (3491, 4530, 4245, 3367) average about 3908. Later months have fewer cohorts, and the standard deviation shrinks from $694 to $567 to $484 to $434. The table's month-on-month rates run +35%, −8%, −17%, −19%, then −17.9% from month 5 to month 6. The forecast applies this last observed decrease to month 5's $2421, which gives $1988 for month 6. The next rate is 17.9% × 95% = about 17.0%, so month 7 is $1650. The decrease keeps shrinking by 5% per step. By month 35 the rate is −3.8% and MpK is $123, and by month 36 it is $118. The source describes a pattern where MpK levels off after an initial period, which is what justifies this extrapolation. Multiplying these MpK values by acquisition volume and comparing them with CoA gives the CLV link the scorecard tracks.

• If a metric is not produced by your systems today, score it 0 and do not report on it. Build it before using it in targets.
• If you plan to raise ARPU (price), then forecast higher churn in the same review. If you plan to spend more on marketing, then forecast higher CoA and possibly higher churn.
• If the standard deviation shows an outlier cohort biasing a monthly average, then remove the outlier before extrapolating.
• If you have fewer than 2 cohorts of history, do not forecast MpK. If you have more than 12 months of past data, stay within the 12-month maximum range the source uses.
• If channels show different retention behavior, then judge channels on CLV and CoA together, not on acquisition volume.

• Reporting all customer metrics improving together. The tell is a deck where ARPU rises while churn and CoA also fall, which the source calls unlikely in reality.
• Counting a metric as available because it could be computed. The tell is a registry full of ones but no owner, no regular report and no one who can give you last week's number.
• Averaging cohorts without checking dispersion. The tell is a monthly MpK driven by one unusual cohort, visible as a large standard deviation relative to the mean.
• Using one blended acquisition cost across channels. The tell is a Facebook or Google customer assumed to have the same churn as a Yahoo or Bing customer.
• Monitoring churn only on a historic basis, or only occasionally. The tell is that problems show up in monthly results rather than in daily instant churn, which the source treats as a leading indicator.

• ▸Metric name: ARPU, churn/retention, cost of service, CoA, CLV, Customer Equity, etc. (one row per metric).
• ▸Available (1/0): 1 only if the metric is routinely produced today.
• ▸Cut dimensions: country, subscription flight, channel, campaign for which the metric is reported.
• ▸Refresh cadence and owner: daily or other frequency, and the named person responsible.
• ▸Data accuracy check: how log-file or transaction data is verified.
• ▸Used for planning/control, targets and rewards: which levels (top management, BU managers, sales managers, middle managers, customer care).
• ▸Used for external reporting or valuation: yes/no, and where.
• ▸Paired trade-off metric: the metric expected to move the other way (e.g., ARPU↔churn, marketing spend↔CoA).
• ▸Cohort history available: number of monthly cohorts (2–12) and any outliers removed.

1. 1Diagnose the silo first. List every product or functional manager who contacts your top customers. The source's test: each one may push a different offering to the same customer without first determining the customer's true needs. If two or more people do this, you have the problem the realignment must fix.
2. 2Define the segments before you draw any boxes. Customer-centricity rests on a carefully defined, quantified segmentation in which operations deliver the greatest value to the best customers at the least cost. Use CLV (retention rate, margin per customer, cost of customer acquisition) to rank segments, so that structure follows value and not org-chart habit.
3. 3Start with lateral coordination, not a full reorganization. The source names lateral coordinating activities as the first stage of realignment. Set up a horizontal structure in which information flows are readily shared among team members, and leave the product lines in place.
4. 4Name an owner for each customer relationship. Candidate roles in the source are customer relationship managers, a Chief Customer Officer, and a Business Intelligence unit. Give each role written authority over specific decisions: volume versus margin with a given customer, how many products to sell to a customer, and which relationships to build over the long term.
5. 5Install a Marketing–Accounting and Finance interface. The source calls this interface crucial. Have finance supply the customer profitability and CLV figures so that the people holding decision rights can see what a customer adds to the bottom line.
6. 6Give decision-makers integrated data. An integrated customer database must be accessible to those managing the relationship so they can understand the 'why' behind behavior. Reduce the number of warehouses, marts, caches and operational stores that must be integrated, since too many integration points are a reason firms fail to get value from customer data.
7. 7Tie the rights to measures and rewards. Organizational architecture holds only if decision rights, performance measures and incentives cohere. Reward account managers for increasing customer equity and relationship managers for extending profitable customer lifetime, so the people with authority are pushed toward the same objective.
8. 8Price the structure. Corporate-level customer-centric structure raises satisfaction but adds coordination costs (Lee et al. 2015), and it pays off only in specific competitive markets. Before you scale up, check whether your market is competitive enough and whether benefits are expected to exceed the added costs.

The source's clearest real case is the Chief Customer Officer. More than a dozen Fortune 1000 firms, including Coca Cola, Hershey, Intel, HP and JD Edwards, created this specialized function to acknowledge the importance of customers. Picture the problem it addresses: a company organized in product silos, where each product or sales manager pushes a different offering to the same customer. The realignment follows the source's sequence. First, lateral coordination links the silos through shared information flows. Second, the customer role gains authority over cross-product decisions. Third, a Marketing–A&F interface supplies CLV so the authority rests on value data. The evidence shows a trade-off. Lee et al. (2015), using Fortune 500 data from 1998 to 2010, found that a corporate-level customer-centric structure improved customer satisfaction but added coordinating costs. The source does not report the outcome of these specific firms' moves. Judge your own move by whether the satisfaction gain outweighs the coordination cost in your market.

• If several product or sales managers pitch the same customer without a shared view of that customer's needs, then assign a single relationship owner and shared data before changing any reporting lines.
• If your market is highly competitive, then lean toward corporate-level customer-centric structure; the evidence says it pays off only in specific competitive markets, so in a weakly competitive market keep it lighter.
• If a new customer role has authority but no CLV or profitability data, then fix the Marketing–A&F interface and data access first; authority without value information repeats the silo problem.
• If decision rights move but compensation stays product-based, then change the rewards (customer equity, profitable lifetime duration) in the same cycle, because the three architecture levers must cohere.
• If coordination cost is visibly rising and satisfaction is not, then stop at lateral coordination rather than adding a full structural layer.

• Treating structure as a free win. The tell: no one has estimated the coordination costs, even though the source reports that customer-centric structure adds them and that the net effect on performance is largely empirically untested.
• Reorganizing without segmenting. The tell: customer groups are defined by habit rather than quantified value, so you cannot say which customers get the greatest value at the least cost.
• Appointing a CCO or relationship managers with no data. The tell: they cannot see what a customer adds to the bottom line, or must pull from many unintegrated warehouses and marts.
• Moving decision rights but not incentives. The tell: sales and relationship managers are still paid on product volume, so they keep pushing products rather than extending profitable customer lifetime.
• Copying a structure into the wrong market. The tell: you adopt a full customer-centric structure in a market that is not highly competitive, where the payoff in the source appears only in specific competitive markets.

• ▸Silo symptom: list the managers who contact the same top customers and the conflicting offers they push.
• ▸Customer segments: name each segment and its CLV basis (retention rate, margin per customer, acquisition cost).
• ▸Decision owner: the role (relationship manager, CCO, BI unit) that holds authority for each segment.
• ▸Decisions delegated: volume vs. margin, number of products sold, long-term relationship investment; mark who decides each.
• ▸Lateral coordination mechanism: the horizontal team or information flow linking product silos (stage one).
• ▸Marketing–A&F interface: the named finance contact and the customer profitability or CLV report they supply.
• ▸Data access: which integrated customer database the owner uses, and the number of systems feeding it.
• ▸Aligned measures and rewards: the performance metric and incentive tied to each owner (e.g., customer equity, profitable lifetime duration).
• ▸Cost-benefit check: expected satisfaction gain vs. expected coordination cost, and whether your market is highly competitive.

1. 1Diagnose the current structure with the passages' organizational-architecture lens. Check three levers separately: the information system, decision rights, and incentives. Integration fails when any one is missing, so write one line per lever describing how Marketing and A&F handle it today.
2. 2Name the shared decisions first. Use the passages' customer-centric questions as the agenda: volume or margin for a specific customer, how many products to sell to one customer, how to build profitable long-term relationships, which segments and processes are more profitable, and how to change customer behavior. Each question needs a Marketing owner for the behavioral input and an A&F owner for the profitability input.
3. 3Build one integrated customer database at the individual-customer and transaction level. It should give a unified, organization-wide view of the customer base, irrespective of product or channel. Marketing and A&F draw from the same records, not parallel extracts.
4. 4Cap the integration points. The passages say firms struggle with customer data because of the excessive number of integration points among warehouses, data marts, caches, and operational data stores. List every store feeding customer metrics, name one system of record per metric, and retire or subordinate the rest.
5. 5Create a lateral coordination mechanism so information moves between functions rather than up and down. Options from the passages: a horizontal structure in which information is readily shared among team members, a Chief Customer Officer role (as at Coca Cola, Hershey, Intel, HP, and JD Edwards), or a split of external-facing and internal-facing roles as at Wells Fargo.
6. 6Define each customer metric jointly, covering CLV, customer equity, and customer profitability. Marketing supplies the behavioral data (preferences, satisfaction, retention). A&F supplies the validation and the valuation formula. Document the formula and its source. The passages note CLV formulae vary, so choose one deliberately.
7. 7Put selected customer metrics among the KPIs reported regularly to top management and the board, and tie rewards to them. Information alone does not change behavior, because managers also need incentives to use it. Check that both functions' incentives point at the same metrics.
8. 8Remove coordination roadblocks explicitly. The passages' case notes that horizontal coordination makes decision-making more cumbersome and that top management had to remove roadblocks. Assign an executive sponsor and review the friction points on a fixed cadence.

Wells Fargo, as described in the source, realigned its organization with a two-tiered sales structure. A relationship manager provides the interaction orientation, which is external focus. A product specialist provides the technical input for product development, which is internal focus. This is the kind of lateral arrangement that replaces product silos, where each product or sales manager pushes offers to the same customer without first establishing the customer's true needs. Read this as an integration design, then add the A&F interface the passages call crucial. The relationship manager's customer knowledge and the product specialist's technical input need a customer-profitability view from A&F, so decision makers see profitability at the customer level, not just the product level. A parallel structural choice is the Chief Customer Officer role at more than a dozen Fortune 1000 firms, which brings customer-centricity issues into the boardroom. The passages do not give Wells Fargo's metrics or results, so any numbers you attach are your own. The sequence to copy is shared customer data, then a joint profitability view, then a role split that routes it to decision makers.

• If each product or sales manager is pushing different offers to the same customer, then the problem is silo structure, not data volume. Fix the lateral coordination before buying more analytics.
• If Marketing and A&F report different customer numbers, then check the integration points first. Reduce the number of data stores and name one system of record before debating definitions.
• If a customer metric exists but no one's rewards depend on it, then add it to the KPIs reported to top management and the board and align the incentives. The passages hold that incentives shape whether information gets used.
• If you need external customer focus and internal technical input at once, then split the roles (relationship manager and product specialist, as at Wells Fargo) and connect both to the same profitability data.
• If coordination is slowing decisions, then treat it as an expected cost of horizontal structure. Have senior management remove the specific roadblocks rather than reverting to silos.

• Treating customer analytics as a Marketing-only task. The tell: the product-centric pattern of the marketing department owning customer relations, with A&F seeing customer metrics only after they are published.
• Building the interface without data integration. The tell: several warehouses, marts, and stores each produce a customer figure, and no one can say which is authoritative.
• Integrating data but leaving incentives unchanged. The tell: dashboards exist, but managers' rewards and the board's KPI pack contain no customer metrics.
• Setting up a coordinating role or title without decision rights. The tell: a Chief Customer Officer or equivalent exists, but the product or sales managers still decide what is pushed to each customer.
• Ignoring the friction of lateral coordination. The tell: decisions slow down, nobody has been assigned to remove the roadblocks, and teams drift back to functional silos.

• ▸Shared decision: write the customer decision both functions must make (e.g., volume vs. margin for a specific customer).
• ▸Marketing owner and input: name the person and the behavioral data they supply (preferences, satisfaction, retention).
• ▸A&F owner and input: name the person and the profitability or valuation input they supply (CLV, customer equity, customer profitability).
• ▸Metric definition and formula: record the agreed definition, the formula chosen, and who validated it.
• ▸System of record: name the single data source at individual-customer and transaction level, and list the other stores that feed it.
• ▸Coordination mechanism: state the lateral structure (horizontal team, Chief Customer Officer, relationship manager/product specialist split) and who sits in it.
• ▸Incentive link: list the customer metrics in the KPIs reported to top management and the board, and the rewards tied to them.
• ▸Known roadblocks: list the coordination frictions slowing decisions, with the executive responsible for removing each.
• ▸Review cadence and date: set when Marketing and A&F next review the metric and decision together.

1. 1Start from the three-legged stool: decision rights, performance measures, and compensation must be designed together. Before touching pay, write down who decides what and which customer metric informs that decision. Incentives only work if the people paid on a metric also control the levers that move it.
2. 2Select a short list of customer metrics (CLV, churn/retention, net additions, Customer Equity) and add them to the key performance indicators reported regularly to top management and the board. A metric that never reaches the board will not carry weight in a pay plan.
3. 3Map each role to one metric it can influence. The source's examples: sales/account managers are rewarded for increasing customer equity; relationship managers are rewarded for extending the profitable lifetime duration of customers. Do not give every role the same composite metric.
4. 4Link the formal evaluation of employees to these metrics and targets. Set targets on the customer metric itself, and write the weighting into the evaluation. Information alone does not change behavior.
5. 5Audit the existing incentive scheme for short-term accounting profit links. In the two banks studied (Ryals 2005; Casas-Arce et al. 2011, 2017), CLV was supplied to employees but incentives stayed tied to short-term accounting profits. Remove or rebalance any component that pays against the customer-centric objective.
6. 6Make the metric trustworthy before you pay on it. Build the Marketing–Accounting & Finance interface so that decision makers get relevant customer-profitability information. Reward only what finance can report consistently.
7. 7Consider a small multi-dimension scorecard, as Texas Instruments did: three marketing metrics covering marketing gains for three consecutive years, efficient and timely services, and better understanding of customers.
8. 8Pilot, then compare by segment. In the bank evidence, shorter-tenure branch managers responded more strongly once CLV became a KPI. Check whether newer managers respond differently from veterans and adjust training or targets accordingly.

Two banks studied by Ryals (2005) and Casas-Arce et al. (2011, 2017) started giving employees CLV as a forward-looking non-financial metric. The information set widened, but both kept the managerial incentive scheme tied to short-term accounting profits. That is the misalignment: staff see lifetime value yet are paid on this period's profit. In the later banking evidence (Casas-Arce et al. 2017), the performance evaluation system began emphasizing CLV as a key performance indicator. Branch managers with shorter tenure showed a stronger response, which suggests CLV information can substitute for work experience. The fix the passages point toward is to follow the aligned examples: reward account managers for increasing customer equity, reward relationship managers for extending profitable customer lifetime, and, as Texas Instruments did, use a small set of marketing metrics inside the reward system. Informing without rewarding leaves the old behavior intact.

• If employees receive CLV or churn data but bonuses still key off short-term accounting profit, then change the bonus. Reporting the metric alone is the bank-case misalignment.
• If a role cannot influence a metric, then do not pay on it. Pair the metric with the decision rights the role holds (e.g. account managers on customer equity, relationship managers on profitable lifetime).
• If finance cannot produce the customer metric reliably, then fix the Marketing–A&F interface before attaching pay to it.
• If the workforce mixes short- and long-tenure managers, then expect a stronger response from newer managers to a CLV KPI and give experienced managers more support, since experience may have been substituting for the metric.
• If the firm is shifting from product- or transaction-centric models, then redesign decision rights and measures together with incentives; changing one leg of the stool produces incoherence.

• Enlarging information without changing pay. Tell: employees can quote CLV but their bonus statements still show short-term profit.
• Paying on a metric the role cannot move. Tell: targets are missed regardless of effort, and managers contest the measure.
• Customer metrics stay off board and top-management KPI reports. Tell: leadership reviews revenue and profit only, so incentives look like an HR experiment.
• Designing incentives in isolation from decision rights and performance measures. Tell: the same customer-centric target is set for everyone, and nobody owns the decision that drives it.
• Treating the incentive design as universal. Tell: the plan is copied from another firm without regard to industry, environment, or technology, which the source says these choices depend on.

• ▸Role: name the role (e.g. account manager, relationship manager, branch manager).
• ▸Decision rights held: list the customer decisions this role actually controls.
• ▸Customer metric assigned: choose one (CLV, churn/retention, net additions, customer equity, profitable customer lifetime).
• ▸Reported to top management/board? Yes/No, and how often.
• ▸Finance/A&F source of the metric: name who computes it and confirm it is reproducible.
• ▸Current pay driver: state what the bonus currently rewards (e.g. short-term accounting profit).
• ▸Conflict check: note where the current pay driver pulls against the customer metric.
• ▸New weighting and target: state the metric's share of evaluation and the target level.
• ▸Tenure split review: record how short- and long-tenure managers respond after rollout.