A scary amount of brands have a retention problem hiding right under their noses.

It usually looks like this: the Shopify dashboard says repeat customer rate is 40% and NPS score is in the 80s, but something doesn’t add up. Growth has plateaued, CAC is climbing, and the team has to discount harder every season to hit their numbers.

The root cause is much deeper than stale creative or a few weak SKUs.

It’s that the standard set of retention metrics are just wrong. Measuring the wrong thing on the wrong time horizon, and the picture they paint is overly optimistic by design.

We sat down with Cathy Quain (former VP of Ecommerce at Burton, now on the advisory side at Compass + Nail) to unpack all of this, and most importantly what to do about it.

Go deeper than just “repeat customers”

We often talk about customers in binary ways: bought vs didn’t buy, or first-time vs repeat buyers.

But reality is much more nuanced than that. Cathy calls it the customer activation cycle.

Customers are moving through distinct mindset states: prospect, first-time customer, then to second, third, and fourth purchase. Eventually, they become what Compass + Nail calls a "cheerleader:" the ones who not only keeps buying, but actively advocate for the brand.

This distinction matters because the economic value at each stage is wildly different.

For example, Compass + Nail’s analysis found that Cheerleaders are worth roughly 12 times what a first-time customer is worth!

So yes, you want to convert first time customers to repeat buyers. But it’s ultimately about ascending them from first-time customers all the way to Cheerleaders - and having clear visibility each step of that ladder.

And it doesn’t help that many of the standard metrics are unclear at best.

For example, she was explicit about why the standard Shopify number is misleading her own clients. One brand she worked with was looking at a 40% "returning customer rate" and an 80s NPS score and was convinced retention was a strength.

When Compass + Nail ran their algorithm against the actual customer data (minimum three years, measured across a two-year activation cycle so customers have real time to experience the product and form a relationship)  the picture was completely different.

She called out $8 million in missed opportunity at that one brand, most of it concentrated at specific fall-off points between purchases 1 and 2, and between purchases 2 and 3.

Put simply, as she told us, "The Shopify model is so flawed in how it reports that it's giving everybody this false sense of success in actually keeping customers."

Why NPS stopped working

The other shoe Cathy dropped was about NPS, which is still the most-cited "proof of loyalty" metric in apparel decks.

Her argument is that NPS was designed to predict business success by measuring willingness to refer. And while that may have worked in the past, NPS just isn’t a great measure of that now.

"It doesn't work anymore because of grade inflation, review exhaustion and timing of reviews when they go out. We do the opposite. If you’ve proven that these customers are coming back and migrating through this cycle at the right rate, then you're going to get brand advocacy. You're going to get people who are doing those natural referrals."

In other words: behavior is more predictive than stated intent.

An 80 NPS score tells you what customers said to a survey after a purchase… but that’s about it.

A healthy migration rate from second to third purchase tells you what they actually did six months later with their wallet. One of these is a lot more reliable than the other.

So, What Should Brands Actually Measure?

This is the part most useful for apparel brands even if they never work with Compass + Nail directly — it's a useful diagnostic framework to apply against your own data.

A few things they look at:

Segment sizes

How many customers sit in prospect, casual, first-time, repeat, and cheerleader segments? Not just the top-line count, but the ratio between them.

Migration rates between segments

Of customers who made purchase one, what percentage made purchase two within the activation window? Of those, what percentage made purchase three? This is where the real health diagnostic lives. Cathy said the benchmarks aren't as aggressive as people assume — you need to retain less than 20% of first-time buyers to hit a healthy number, but most brands score so low they think they're doing great.

Segment value and CAC ratios

What does each segment cost to acquire and what is each segment actually worth over the activation cycle? When cheerleaders are worth 12x a first-time buyer, small shifts in migration rate are enormous in dollar terms.

Fall-off point analysis

Where is the leak? Is it that you can get people to second purchase but lose them at third? Is it that you're getting cheerleaders but they aren't worth what they should be (usually a sign you're over-discounting to bring them back)? The diagnosis drives the remediation.

They've packaged all of this into what they call the Brand Equity Index, a 0-to-500 score that works like a health diagnostic for the brand, with the added output of showing how much profit is on the table if the brand were hitting benchmark migration rates.

The reason the framework clicks for us is that it doesn't try to replace attribution; it’s more like a higher-level health score.

Cathy was clear about that: "It doesn't get into any attribution at all. There's more than enough Triple Whales and whomevers out there doing that."

Attribution tells you which channels deserve credit for a transaction (Meta, Google, Amazon, etc), the activation cycle tells you whether your business is building an engaged customer base that compounds.

Retention is not marketing's job

Did that line make you put your coffee down? It’s spicy, but it also makes sense:

"Retention is the job of every single employee in the company. It's not marketing's job. It's not the retention marketing manager's job. It starts with designing great product, producing great product, pricing your product where the value proposition is there and feels right, having a good supply chain that's going to deliver on time, and that your products aren't going to fail when the customer actually uses them."

Put simply, brands treat retention as a set of email flows and a loyalty program. But in reality, it’s a much more holistic concept.

Cathy's point is that by the time the customer gets to the email flow, the retention outcome has already mostly been determined by product quality, fit consistency, supply chain reliability, pricing integrity, and the post-purchase experience.

The marketing team is the thinnest layer on top of a much larger stack.

For example, most brands send two to three marketing emails a week. Over a year, that's 100 to 150 emails. The typical apparel customer might shop twice a year from a given brand. So you're asking a customer to tolerate 150 emails to get them to two purchases.

If the emails aren't genuinely serving them - eg, if they're mostly "you abandoned your cart" and "here's 15% off" - you're actively training the customer to either tune you out or to only transact at a discount… neither of which are great for your business!

"Is that really serving a customer? Maybe, because they're getting a discount. But is that creating a lifelong loyal customer? Probably not."

The trap: chasing the number instead of the customer

Cathy also described a pattern we’ve seen many time: the late-stage growth brand that's been on a good run, and is now plateauing.

The brand has 5-10 years of double-digit growth, gets anxious when the growth slows, and starts chasing those old growth numbers.

So they start putting out product aimed at a new customer in an attempt to widen the market and start another growth cycle. But more often than not, they just end up moving them away from their core customer and the reason why they started in the first place.

As Cathy says, “That's a downward slide, because now your products aren't resonating as much, now you get promotional, or you've lost your consumer. You're now starting to chase an entirely different set of consumers because you're trying to widen your market."

This matches what we see across the board. The widening-the-market move almost never works. The consumer you're chasing doesn't convert at the rate you modeled, and the core consumer who was carrying your economics walks away from the brand.

Cathy credited this lesson to Jake Burton, which is a good endorsement.

The discipline of staying crystal clear on who your best customer is, and resisting the pressure to dilute the brand in service of top-line growth, is a discipline most brands lose around year 8 or 9 — which is almost exactly when it matters most.

So what should brands do with this?

A few things worth doing regardless of whether you're working with an analytics partner:

Stop trusting your Shopify repeat-customer number as a retention signal

It's measuring something, but it's not measuring what you think. Pull your customer data across at least 24 months and look at migration rates between purchase 1, 2, 3, and 4. The picture will be less flattering and much more useful.

Find your real fall-off point

Most brands have a specific leak — a specific purchase-to-purchase transition where customers drop disproportionately. That's the point where your post-purchase experience, product consistency, or second-order communication is failing. Fix that one point and the economics shift meaningfully.

Audit your post-purchase experience from the customer's seat

Literally walk through what a customer gets after they place an order. The confirmation email, the shipping updates, the review request ("write a review" sent seven days after delivery when the product hasn't been used yet), the abandoned-cart nagging, the re-engagement flows. Is any of this actually serving them, or are you hammering them with widgets and apps?

Treat your cheerleaders as financial assets

If a cheerleader is worth 12x a first-time buyer, your entire marketing calendar should have an explicit strategy for cultivating them — and that strategy should not primarily be discounts.

If you over-discount to reactivate your best customers, you're dragging the value of your most valuable segment down toward the value of your worst one.

Resist the widening-the-market move

If growth has plateaued, the first instinct of most teams is to go hunt for new audiences. The better first move is usually to figure out why your core audience is buying less, and fix that.

It won’t happen overnight. Do it anyway.

The work Cathy is pointing at isn't complicated, but it takes patience.

Clarity on who your customer is. Product that delivers on what the marketing promised. A post-purchase experience that feels like it was built for the customer instead of for a conversion dashboard. Migration rates measured seriously. Fall-off points diagnosed and fixed. The same core customer served deeply instead of a new customer chased shallowly.

There is no version of the market that makes this approach the wrong one, and there is increasingly less room in the market for the approach that skips it.

‍