REPEAT COMMERCE: 02 - Run The Math Before You Raise The Ad Budget

You've probably heard this statistic a hundred times: improve customer retention by five percent, and profits jump twenty-five to ninety-five percent.

It might be the most repeated number in retention marketing. It's on the landing page of every retention app ever built.

It also comes from a 1990 study of banks, insurance brokers, and auto-service companies. Not Shopify stores. Not ecommerce. The paper is fourteen years older than Shopify.

I still believe the argument is right. But a thirty-six-year-old banking study can't tell you what retention is worth to your store. So today we're going to calculate it properly, with Shopify numbers. Yours.

CALLBACK + THE 1% WIN + THE SCORECARD (0:35–2:00)

Welcome back to Shopify1Percent, I'm Jay Myers, and this is episode two of the repeat commerce series.

If you did last episode's homework, you've got two numbers on a sticky note: your returning customer rate and your repeat revenue share. If you skipped it, go back after this one, it's a ten minute job.

And a small announcement about that sticky note. It's graduating. Over this series, we're turning it into a proper scorecard, a running retention audit of your own store. Every episode adds one number or one action, and by the end of the series you'll have done a full retention teardown without ever sitting down to do a full retention teardown. Episode one gave it two lines. Today it gets three more.

Because by the end of this episode, you'll run one calculation, about ninety seconds, that tells you what a five point lift in your repeat rate is worth to your store in dollars per year. Then a second version, maybe two more minutes, that tells you something better: whether your next marketing dollar should buy ads or buy retention. Most founders have never run that comparison. Not because it's hard. Because nobody told them it was a decision.

[PRE-ROLL SPONSOR SLOT]

WHAT BAIN ACTUALLY SAID (2:00–3:15)

Quickly, the famous stat, because I don't want to build on a foundation I haven't checked.

The 25-to-95 claim comes from a 1990 Harvard Business Review paper by Frederick Reichheld and Earl Sasser called "Zero Defections." Reichheld later invented the Net Promoter Score, so, credible guy, real research. But the findings were per industry, in services: cutting customer defections five percent generated about 85 percent more profit in one bank's branch system, around 50 percent in insurance brokerage, thirty percent in auto service.

The direction of that finding has held up for three and a half decades. Keeping customers is worth more than almost anyone budgets for. Reichheld even used a leaky bucket to explain it, which is where everyone's favourite retention metaphor comes from.

But if you run a Shopify store, quoting a 1990 banking study at your CFO isn't a strategy. So let's build a store and run the experiment ourselves.

THE FRAMEWORK: LEAKY BUCKET MATH IN 3 STEPS (3:15–12:30)

Step 1: Build the store and price the leak (3:15–6:00)

We're building this store on the whiteboard, together, right now. It's fictional, and the numbers are round on purpose so you can follow along while you're driving.

It's a coffee brand. A million dollars a year in revenue, fifty dollar average order value. Over the last twelve months it acquired fifteen thousand new customers, and 25 percent of them came back for a second order. Two more inputs, and I'm stating them out loud because they matter: contribution margin before marketing is 40 percent, so each fifty dollar order leaves twenty bucks after product, shipping, and fulfillment. And it costs 35 dollars in ad spend to acquire a customer.

[natural moment] Now, your margin might look nothing like 40 percent, and your CAC might make 35 bucks look like a dream. I'm picking these to keep the math round, not because they're gospel. The whole point of today is that you'll rerun this with your numbers, and yours are the only ones that matter.

So, the leak. A 25 percent second-purchase rate means 11,250 of those fifteen thousand customers bought exactly once. Run the economics on one of them: their order left twenty dollars of contribution, and they cost thirty-five to acquire. That customer is a fifteen dollar loss, locked in forever unless they come back.

Eleven thousand two hundred and fifty of them, at fifteen dollars each. That's about 169 thousand dollars this store has quietly lost on customers who never returned.

And what does the founder do when growth slows? What we all do. Raise the ad budget. Refill the bucket faster. Nobody in the Monday meeting says "what if we made the hole smaller," because the hole doesn't have a dashboard. ROAS has a number. Churn just has a vibe.

Step 2: Price the plug (6:00–9:00)

Same store. Now we run the experiment: move the second-purchase rate five points, from 25 to 30 percent.

[natural moment] One clarification before someone attacks me in the comments, and this is genuinely worth thirty seconds because I see merchants trip on it constantly. I'm modelling a five-percentage-point lift, 25 to 30. That is not the same as improving 25 percent by five percent, which would barely move anything. Five points on this store means 750 additional customers come back. It's also a bigger move than what Bain modelled, so I'm not recreating their study. I'm running a clean Shopify scenario and showing you the machinery so you can turn the dials yourself.

So: 750 additional second orders, at our fifty dollar AOV. Seven hundred and fifty times fifty is 37,500 dollars a year in new revenue.

And that's the floor, deliberately. It assumes each of those customers comes back exactly once and nobody orders a third time. Your own repeat customers' actual behaviour, how big their reorder baskets are, how many come back again, is sitting in Shopify's cohort reports, and we'll pull it in a minute. For the whiteboard, we stay conservative.

Now profit. At our 40 percent contribution margin, that 37,500 of revenue leaves 15,000 dollars of contribution profit. And notice the best part: we already paid to acquire these 750 people. There's no CAC on the second order. That fifteen grand arrives clean.

Quick callback to last episode's numbers, the SimplicityDX research: the average new customer arrives at a 29 dollar loss, while the average repeat sale generated about 39 dollars for merchants. Different data set, same shape. The first order pays the toll. The second order keeps the money.

[MID-ROLL SPONSOR SLOT]

Step 3: Price the alternative (9:00–12:30)

Now the comparison this episode exists for, and the reason for the title.

Say the founder skips all of this and buys that same 37,500 dollars of growth the normal way, through acquisition. At a fifty dollar AOV, that's 750 brand new customers.

Their orders generate the same 15,000 dollars of contribution before marketing. But acquiring them costs 750 times 35 dollars, which is 26,250 in ad spend.

Fifteen thousand in, twenty-six and a quarter out. That's an 11,250 dollar loss.

Put the two paths side by side. Same store. Same 37,500 dollars of new revenue. Through retention, it adds fifteen thousand of profit. Through acquisition, it loses eleven thousand. Same top line. Opposite sign. A 26 thousand dollar swing on a single budget decision most founders don't realize is a decision.

A dollar of revenue is not always worth a dollar. Where it comes from changes what's left over.

Now, two honest caveats, because I want you to trust this math enough to argue with it. First: acquisition builds the base retention feeds on. You can't retain people you never acquired, and nothing in this series will ever tell you to turn off your ads. Second: your numbers might show acquisition IS profitable. Plenty of stores have first-order economics that work. Great. The point isn't to rig the math so retention always wins. The point is that almost nobody runs the comparison before deciding where the next dollar goes. Run it once and you'll never unsee it.

[natural moment — Jay: this one's yours to personalize, swap in the version you actually remember] I watched this exact pattern for a decade at Bold. A merchant would have a killer month, revenue up, ROAS up, high fives all around, and their returning-customer line hadn't moved in a year. Then ad costs would rise, or an iOS update would land, and suddenly the whole business looked broken. The ads hadn't stopped working. The ads were the only thing working. That's the difference between a growth engine and a life-support machine, and from the revenue chart alone, you genuinely cannot tell which one you're running.

THE TACTICAL PART: RUN IT ON YOUR STORE (12:30–15:00)

Two versions. The quick one tells you the revenue opportunity. The adult version tells you whether to actually move budget.

The 90-second version. Three inputs from your Shopify admin. One: new customers acquired in the last twelve months. Analytics, then Reports, look for "New customers over time." Two: your average order value, it's a card on the main Analytics dashboard. Three: the number 0.05.

New customers, times 0.05, times AOV. That's your conservative annual revenue floor for a five point lift in second purchases.

One nuance, and it sets up a future episode: use a mature window. Someone who placed their first order last Tuesday hasn't had a fair chance to reorder yet, so if your store is growing fast, recent cohorts will make your repeat rate look worse than it is. Trailing twelve months is a reasonable default; just know the freshest customers in there are still "in progress."

The adult version. Take that revenue floor and multiply by your contribution margin, the percent of each order left after product, shipping, and fulfillment, before marketing. That's the profit the lift would add. Then price the other path: your first-order contribution (AOV times that same margin) minus your blended CAC. If that second number is negative, every new customer is a bet that they'll come back, and your repeat rate is literally the odds. If it's positive, congratulations, and the comparison still tells you which dollar works harder.

Shopify's cohort reports, look for "Customer cohort analysis" under Customers, will also show you what your actual repeat customers spend and how often they return, so you can replace my conservative assumptions with your real behaviour.

DO THIS TODAY + THE SCORECARD (15:00–16:00)

Your one percent win: run the 90-second version. New customers times 0.05 times AOV. Write the dollar figure on the scorecard, under your two numbers from last episode. If you've got three more minutes, run the adult version and add both profit numbers: the contribution from the lift, and your first-order contribution after CAC.

That's five lines on your scorecard now. It's becoming a real picture of your retention economics, and it cost you about twenty minutes across two episodes.

Then say the lift number out loud in your next marketing conversation. Not as a plan, just as a fact. "A five point repeat lift is worth about forty grand a year to us." Budgets have a way of noticing numbers with dollar signs attached, in a way they never notice "we should really work on retention."

RECAP + CTA (16:00–16:45)

The recap fits in one breath: on our whiteboard store, the same 37,500 dollars of growth adds fifteen thousand of profit through retention and loses eleven thousand through acquisition. Same top line, opposite sign. And your version of that math is three inputs in Shopify Analytics.

Next episode: you're tracking the wrong number. Why revenue and ROAS can both look great while the metric that predicts your survival quietly rots, plus the exact queries to find your true repeat rate with subscribers stripped out. I promised those queries in episode one, and I'm making good.

If you got one percent better today, follow Shopify1Percent wherever you're listening. And if you ran the math and it surprised you, leave a comment or a review and tell me the number. I read all of them.

See you in the next one.