Is Your Growth Real? Cohort Quality and Customer Equity Dilution

Short answer: Growth is real when each new group of customers is at least as valuable as the groups before it. You check this with a cohort chart: group customers by the month of their first order and track cumulative revenue or margin per customer over time. If new cohorts start flatter, pay back slower or stop buying sooner, you are diluting your customer base, and spending more will make it worse, not better.

These ideas come from RetentionX's newsletter; we've summarized them and added how we use them in our Meta work. The brands in the examples are made up.

Free download: Cohort Quality and Customer Equity Dilution, the one page summary (PDF). Print it or share it with your team.

Why can growth look fine when it is not?

Because the top line metrics average everything together.

The problem: Revenue is up. Blended CAC (cost per new customer) is steady. ROAS looks fine. But each new month of customers is a little worse than the last. They buy less after their first order, repeat less, and take longer to pay back what you spent to get them. You are filling your customer file with weaker buyers, and none of the usual dashboards show it.

The solution: Look at customers by cohort. RetentionX calls this the one chart that tells you if your growth is real.

What is a cohort chart?

A cohort is a group of customers whose first order happened in the same month. A cohort chart tracks the cumulative revenue per customer for each group over time.

How to build one, following the newsletter:

  1. Group customers by the month of their first order.
  2. For each group, track cumulative revenue per customer at month 0, 1, 3, 6 and 12. Ideally use CM1 (contribution margin after product and fulfillment costs) instead of revenue, because margin is what pays back ad spend.
  3. Show the curves stacked on one chart, or as a heat map with cohorts as rows and months as columns.
  4. Every month, ask one question: is this new group as good as, better or worse than the groups before it?

How do you read a cohort chart?

Compare the newest cohorts against your best past cohorts at three points. This is a paraphrase of the newsletter's healthy vs warning table:

What to check Healthy Warning
The start (month 0 to 1) New cohorts match or beat past ones New cohorts start flatter, even with more spend and better creative
Payback Pays back in the same window as your best cohorts Payback slips later, or never arrives
The tail (month 3 to 12) Keeps adding value Flattens after month 2 or 3, lower order values, more promo dependence
What it means A healthy growth engine Each new cohort lowers the quality of your customer file

RetentionX notes that the warning pattern tends to line up with three things: harder pushes into broad audiences, heavier discounts, or new products that sell easily but do not build loyal buyers.

What is customer equity dilution?

Customer equity dilution is when your customer count grows while the average value of a customer falls.

RetentionX defines Average Customer Equity as the expected future contribution from active customers divided by the number of active customers. When you add a lot of lower value customers, that average drops.

The newsletter's illustrative example: a brand has 200,000 customers, each worth $180 in expected future contribution. A promo heavy quarter adds 80,000 new customers worth $95 each. Customer count is up 40%. But the average customer is now worth less. Working through those illustrative numbers, the blended average falls from $180 to about $156.

Is dilution always bad?

No. RetentionX is clear on this. A customer worth $120 who cost $25 to acquire can be a great deal, even if they lower your average. Dilution is dangerous when nobody sees it, because then budgets keep growing on the assumption that new customers look like old ones.

Signs it is happening unseen:

  • You spend more to get the same revenue.
  • Customers stay for a shorter time.
  • Repeat purchases get weaker.
  • More orders depend on a promo.
  • CAC keeps rising.

What should you measure for each cohort?

The newsletter lists a set of measures to track per cohort. Paraphrased:

  • Contribution margin per customer at 90 and 180 days.
  • The same measure for your existing base over the same window.
  • Cohort quality ratio: new cohort margin per customer divided by base margin per customer. Under 100% means dilution.
  • Repeat rate at 60 and 90 days.
  • First order discount depth, and what share of orders used a promo.
  • Return and refund rate.
  • Categories per customer, meaning how many product types each customer buys.
  • Payback: CAC divided by margin per customer.

The flag to watch: the cohort quality ratio falls and payback gets longer at the same time. One alone can be noise. Both together is a pattern.

How do you find which channel or offer is causing it?

Slice cohorts by where they came from. RetentionX suggests three cuts: acquisition channel, first product bought, and the offer used (a welcome discount, a Black Friday code, or full price).

When you slice this way, you can find combinations that never work. The newsletter's example is a combination of channel, product and welcome discount that never reaches the top quarter of customers by LTV. Once you see that, you can choose channels and offers based on cohort quality, not ROAS.

Here is a made up example. Groundswell sells a $289 raised garden bed kit. Suppose its spring cohorts from a deep first order discount have strong first month revenue but a much flatter tail than full price cohorts, because the discount brought in people who were never going to buy soil, covers or a second bed. The ROAS on that offer looked great. The cohort says it was buying weaker customers. All of this is illustrative.

What is the rule for scaling or fixing?

RetentionX puts it as scale or fix:

  • If new cohorts are as good as or better than your best past cohorts, scale, staying inside your payback and LTV limits.
  • If new cohorts are weaker, you have a quality problem, not a budget problem. Check the channel mix, any new product or bundle, the offer or discount, and your onboarding and lifecycle emails.

Never spend through falling cohort quality. More budget on top of a dilution problem buys more of the wrong customers.

How do we use cohort quality in our Meta work?

This part is our own application.

  • Tag where each cohort comes from. We tie customers back to the Ad Set, angle, offer or discount code that brought them in, wherever the tracking allows.
  • Judge winners on new customer margin, not just CPA or ROAS. An ad that wins on CPA but brings in promo only buyers is not a winner in our book.
  • Look at cohorts by Creative Testing Ad Set. Different concepts can attract very different customers. The cohort view shows which concepts bring buyers who come back.
  • Check cohort quality before a Scaling move. Before we price the next customer with marginal CAC and push more budget behind a winner, we want to know the customers it brings are holding up.
  • Report customer count and cohort quality ratio side by side for promos. A promo that adds a lot of customers and drops the ratio needs to be shown honestly.
  • Seed lookalikes with the best customers. We build value based lookalike audiences from top margin customers, not all purchasers, so Meta looks for more of the right people.
  • When blended CAC rises, check dilution first. Then check incrementality, meaning whether the ads caused the sales at all.

Cohort quality connects directly to customer base half life. Weaker cohorts leave sooner, which shortens half life and means more of your future spend goes to replacing them. Jake's shorthand for clients: cheap customers are only cheap if they stay.

Key takeaways

  • Revenue, blended CAC and ROAS can look fine while each new cohort is worse.
  • A cohort chart tracks cumulative revenue or margin per customer by first order month.
  • Customer equity dilution is a growing customer count with a falling average customer value.
  • Dilution is fine when you see it and price it; it is dangerous when it is invisible.
  • Flag when the cohort quality ratio falls and payback lengthens together.
  • If new cohorts are weaker, fix quality before adding budget.

FAQ

What is a cohort in ecommerce? A cohort is a group of customers whose first order happened in the same period, usually the same month. Tracking cohorts lets you compare how each group behaves over time instead of blending everyone together.

Should I use revenue or margin in the cohort chart? Margin is better, ideally CM1, because it shows what actually pays back your ad spend. Revenue is fine to start with if margin data is hard to get.

Do discounts always dilute the customer base? Not always, but they are one of the common causes RetentionX points to. Slice cohorts by offer and compare discounted cohorts against full price cohorts to see what is true for your brand.

Can Meta data show me cohort quality? No. Meta reports the first purchase it can attribute, not what customers do months later. You need order data with customer IDs from your store, then tie it back to the ads where tracking allows.

Sources: RetentionX newsletter, "The one chart that tells you if your growth is real" and "Customer Base Dilution" (cohort method, healthy vs warning signs, customer equity dilution, cohort measures and illustrative examples).

Check whether your Meta spend is buying the right customers

If you want to know whether your growth is building a better customer base or diluting it, we offer a free audit. You get a full account audit plus net new reach and cumulative reach reports. Request it on our contact page.

Check what you learned

Tap an answer to see if you got it right.

1. How do you check if growth is real?

Use a cohort chart.

2. What is customer equity dilution?

Growing count, shrinking value.

3. If new cohorts are weaker, what first?

More budget buys more weak customers.

Ready for the full test? Take the Retention exam and get certified. Want this run on your account? Get my free audit.

About the author

Jake Mehani, founder of Creative Analytics

Jake has spent 15 years running Meta and paid social ads for consumer brands. Meta published his RBX Active Advantage+ case study, he taught the DMAnc Meta and TikTok ads masterclass, and he spoke at Motion's 2024 Creative Strategy Summit. His senior team runs media buying, creative and measurement like an in house team.

Jake on LinkedIn  |  Case studies  |  Free Meta audit

Previous
Previous

Marginal CAC: Price the Next Customer, Not the Average One

Next
Next

Flat vs Compounding Brands: RFM Segments and the Four Numbers for Your Quarterly Plan