Customer lifetime value strategy: grow your business, no excuses

This is part 5, the last in a five-part series on no-excuses customer experience, and the part about customer lifetime value strategy. Part 1 introduced the excuses and the seven rules, part 2 covered rules 3 to 5, part 3 covered rule 6 and part 4 covered rule 7. This part is about the number underneath all of them.

Somewhere near the end of every one of these conversations, a sensible person asks the question the whole series has been circling. Fine, we start with the customers we already have. We act on the signals. We fix the return path, we equip the frontline, we sell it inside, we measure. But which customers, exactly? And how much is any of this worth?

That is the lifetime value question. Customer lifetime value is the profit you expect from a customer over the time they stay with you, and a customer lifetime value strategy is the set of decisions you make with that number instead of around it: who gets the effort, how much friction removal you can afford, and what counts as success.

Without it, every rule is a good idea competing with other good ideas for the same budget. With it, the rules become arithmetic.

Key takeaways

• Customer lifetime value is the profit expected from a customer over the time they stay, and a rough segment-level estimate is enough to make the seven rules add up.
• Rules 1, 2, 4 and 7 each depend on lifetime value directly: which customers, what a signal is worth, how much friction removal you can afford, and what keeping a customer was worth.
• Four numbers per segment (order value, frequency, margin, tenure) are enough to start, and any missing one can be estimated as long as the estimate is written down.
• A value-at-risk table, one row per segment, sorted by signals multiplied by value, tells the frontline where its hours go this month.
• The excuse “our data is not good enough” confuses precision, which is a luxury, with direction, which is a necessity.

What a customer lifetime value strategy is

Lifetime value in its simplest form is three numbers multiplied: the margin you make on a customer in a typical period, the number of periods in a year, and the number of years the customer stays. For a retailer that might be margin per order times orders per year times years of tenure. For a subscription business it is monthly margin times expected months. The formula can be made more elaborate, with discount rates and survival curves, and the more elaborate versions are worth building later. They are not what the seven rules need.

What the rules need is a strategy: a small set of decisions that are made with the number rather than by instinct. Which segment gets the callback hours. Whether a waived fee is a gift or a rounding error. Whether the renewal flow for the top tier deserves a dedicated fix. Whether a retention result is worth what it cost. Each of those is a coarse decision, and coarse decisions need a direction, not a decimal.

Two things a lifetime value strategy is not. It is not a finance-grade projection, and it will disappoint anyone who expects it to be audited. And it is not a customer-level score that has to exist before anyone can act. The argument over whether lifetime value is a waste of time is mostly an argument between people who mean those two different things by the same phrase.

What lifetime value does for each of the seven rules

The estimate can be crude and still be useful. But notice first how much of the series quietly depends on it.

1. Start with the customers you already have.
2. Act on what you already know.
3. Talk to customers like a person: their channel, their timing, their words.
4. Make coming back easy: remove friction at the moment of return.
5. Give the frontline a reason and a way.
6. Sell it inside before you sell it outside.
7. Measure it, or it did not happen.

Rule 1 needs to know which ones. All customers are not worth the same effort, and lifetime value is the only honest way to say so. A callback program that treats a one-time buyer and a ten-year account identically will run out of hours long before it runs out of customers.

Rule 2 needs to know what the signals are worth. A drop in usage in a low-value segment is a data point. The same drop in your top segment is a fire. The signal is identical; the value at risk is not, and value at risk is what decides who gets the call today.

Rule 4 needs a budget. Every friction fix costs something: engineering time, a waived fee, a skipped verification step and its small risk. Lifetime value tells you how much friction removal you can afford, because it tells you what a lost return is worth.

Rule 7 needs a unit. Retention rate tells you how many customers you kept. Lifetime value tells you what keeping them was worth, which is the version of the number that survives a budget meeting.

Rules 3, 5 and 6 lean on it too, more quietly. Relevance is easier to justify when you know what the relationship is worth. The frontline responds to “this customer matters” far better when someone can say why. And the internal pitch lands harder with a figure in it.

How to estimate lifetime value by segment in a week

You do not need customer-level lifetime value to act. You need segment-level lifetime value, roughly right, and you can usually build that from what you already have.

1. Choose segments people already recognize. Tenure and frequency are enough to start: long-tenure high-frequency, mid-tenure regular, new in the first year, occasional. If your company has a tier structure everyone understands, use that. The point is that people recognize the rows without a glossary.
2. Pull four numbers per segment. Average order value, purchase frequency per year, a rough margin percentage, and how long customers in that segment typically stay. Billing, the point of sale and the finance team hold these between them.
3. Estimate what is missing and write down that you estimated it. If tenure is unknown, use the share of last year’s customers still buying this year and turn it into an expected number of years. A footnote saying which margin and which tenure you assumed is not an admission of weakness. It is what makes the number arguable, and arguable numbers get improved.
4. Multiply. Order value times frequency times margin gives annual margin per customer. Annual margin times expected years gives lifetime value for the segment.
5. Sanity-check against the whole. Multiply each segment’s annual margin by its customer count and add them up. If the total is wildly different from the company’s actual gross margin, one of the four numbers is wrong, and it is usually frequency.
6. Put it on one page, with the assumptions underneath. Four rows, one figure each, dated. Update it monthly, and expect the first two updates to change it a lot.

A worked example (illustrative)

The figures are round and invented, to show the arithmetic rather than to suggest a benchmark.

Take two segments. Long-tenure, high-frequency customers spend 50 an order, six times a year, at a margin of thirty percent: 90 a year in margin. They typically stay another five years, so the segment value is about 450 per customer. New first-year customers spend 40 an order, twice a year, at the same margin: 24 a year, and on average they stay about a year and a half, so roughly 36 per customer.

Now count the rule 2 signals. Suppose 300 long-tenure customers showed an early-warning signal last month, and 1,000 new customers did. The new segment has more than three times as many signals. The long-tenure segment has 135,000 at risk against 36,000. If the frontline has hours for only one list this month, the arithmetic has already chosen it, and it is not the list with the most names on it.

Customer-level vs segment-level lifetime value: which to use when

Both are real, and the choice between them is the most common place a lifetime value strategy stalls, because the customer-level version is what people picture and the segment-level version is what they can build this month.

	Segment-level estimate	Customer-level score
What it needs	Four numbers per segment, some estimated	Joined purchase history per customer, a model, someone to maintain it
Time to first version	A week	A quarter or more
Precision	Right in shape, wrong in detail	Better in detail, still wrong for any individual
Decisions it supports	Which segment gets the hours, what a fix can cost, what a result was worth	Who gets which offer, individual account priorities
Failure mode	Treated as more precise than it is	Waited for, while nothing happens
Use it when	Now, for every decision in the seven rules	After the segment version has been used for a few months

Start with the segment version and keep it in use. The customer-level score is worth building once the decisions it would improve are actually being made, and not before. A precise number nobody uses is worth less than a rough one that sets the frontline’s list every month.

The value-at-risk table

This is the one thing I would build before anything else in the series, because it turns all seven rules into a single page that a leadership team can argue about productively.

Four columns. One row per segment. Update it monthly.

Segment	Customers showing early-warning signals	Rough value per customer	Value at risk
Long tenure, high frequency	Count from the rule 2 signals	Segment lifetime value estimate	Signals multiplied by value
Mid tenure, regular	Count	Estimate	Signals multiplied by value
New, first year	Count	Estimate	Signals multiplied by value
Occasional, low frequency	Count	Estimate	Signals multiplied by value

Early-warning signals are the rule 2 inventory, counted. The customers whose usage halved, whose ticket went silent, whose renewal notice sat unopened. Do not weight them yet. Just count them, per segment.

Value at risk is the product, and it sets the row order. Sort descending. The top row is where the frontline’s hours go this month, where the return audit starts, and what the measurement plan should be watching.

The table will be wrong in its details and right in its shape. The segment at the top will almost always be the one everyone already suspected and nobody had put a figure on. Once the figure is on the page, the excuses get very quiet, because the cost of doing nothing finally has a number next to it, and it is usually larger than the cost of any of the seven rules.

What quietly breaks a customer lifetime value strategy

“Our data is not good enough for lifetime value.” This is the excuse I hear most often once the others have been dealt with, and it deserves a direct answer. A rough estimate beats none because the decisions it supports are coarse decisions. You are not pricing a derivative. You are deciding whether the callback list should be sorted by segment. None of those decisions changes if your estimate is off by a fifth. All of them change if you have no estimate at all. The same logic applies to not letting imperfect data stop you acting on what customers tell you. Precision is a luxury. Direction is a necessity.

Revenue used where margin belongs. A segment with high revenue and thin margin floats to the top of the table and takes the frontline’s hours from a segment that actually pays. Use margin, even a rough one.

The average hides the segments. A single company-wide lifetime value figure is a fact about nobody. It cannot sort a list or set a budget. The table needs rows.

The table is built once. Signals change every month and the value estimates should change every quarter. A value-at-risk table dated last year is a slide, not a strategy.

When lifetime value is the wrong lens

When the purchase does not repeat. Some products are bought once a decade by nature. There is no tenure to estimate, and whether the product suits relationship marketing at all comes before any lifetime value work. The effort belongs in the first purchase and in referrals.

When the business cannot wait for the lifetime. Where the planning horizon is a quarter, a five-year figure will be nodded at and ignored. The answer is not to abandon the number but to translate it into indicators that move inside the quarter, so that the lifetime figure sets direction and the quarterly figure sets the pace.

When it is used to justify neglect. Lifetime value says where the effort goes first. It does not say that the bottom row deserves a worse experience, and a company that reads it that way will find its bottom row telling everyone else why they left.

Where to start

1. Pick four segments people already recognize and write them as the rows of one table.
2. Collect the four numbers for each, from billing, the point of sale and finance, and estimate whatever is missing with the assumption written underneath.
3. Multiply to a rough lifetime value per segment and check the total against the company’s gross margin.
4. Count last month’s early-warning signals per segment from the rule 2 inventory.
5. Sort by value at risk and hand the top row to the frontline, with the date of the next update on the page.

FAQ

What is customer lifetime value?

Customer lifetime value is the profit a company expects to earn from a customer over the time the customer keeps buying. In its simplest form it is margin per period, times periods per year, times expected years of tenure. It is an estimate, and a rough estimate by segment is usually enough to decide where retention effort should go.

How do you calculate customer lifetime value simply?

Take the average order value, multiply by the number of orders a year and by the margin percentage to get annual margin per customer, then multiply by the number of years a customer in that segment typically stays. If tenure is unknown, use the share of last year’s customers who are still buying this year to estimate it. Write down every assumption so the figure can be argued with and improved.

What is a customer lifetime value strategy?

A customer lifetime value strategy is the set of decisions a company makes using lifetime value rather than instinct: which customer segments get the retention effort, how much a friction fix or a waived fee can cost, and what a retention result was worth in money. It works with a segment-level estimate and does not require a customer-level model. The practical output is a value-at-risk table that ranks segments by early-warning signals multiplied by value.

Do you need customer-level data for lifetime value?

No. Segment-level lifetime value, built from four numbers per segment, supports every coarse decision about where retention effort goes and what it may cost. A customer-level score improves offer targeting and account prioritization, and it is worth building only after the segment version is in regular use.

What is value at risk in customer retention?

Value at risk is the number of customers in a segment showing early-warning signals of leaving, multiplied by the rough lifetime value of a customer in that segment. Sorted from highest to lowest, it tells the frontline which list to work first and gives leadership the cost of doing nothing. It is updated monthly as signals change.