REPEAT COMMERCE: 03 - The Repeat Rate Shopify Doesn’t Show You

Jay Myers: ​[00:00:00] So your Shopify store says 40% of your customers come back. That's great. You're above average. You feel great, and you move on with your day. But there's a problem. That 40% is actually answering a different question than the one you think you asked. Shopify is quietly blending three completely different behaviors into one number.

People whose subscription just rebuild, that's one. People wondering, wandering back after two years and, or five years and just happen to be purchasing again. And then three, people who actually choose to buy from you again on purpose on a regular schedule. Those are not the same thing, and only one of those numbers actually tells you if you've built something powerful that lasts. And today, we're gonna pull that apart and we're gonna find the number that Shopify doesn't show you directly on their dashboard. [00:01:00] So welcome back to Shopify 1%.

I'm Jay Myers, and this is episode number three of the Repeat Commerce series. Now, if you missed episode one and two, definitely go back and listen to them. Just look for the titles called Repeat Commerce episode one, episode two. You'll see them, you'll see them on the podcast, because we're building on each episode.

Last episode we proved that a dollar of revenue is not always worth a dollar. So same top line, but opposite sign depending where it came from. Now, if you actually ran that math last episode and you brought it up in a budget conversation, tell me about it.

Leave a comment. Leave a note wherever you're watching this. I read every single one. Shoot me an email, jay, J-A-Y, @ shopify1percent.com. So today, it's actually a similar trap, just one level deeper. Last time, a dollar of revenue was hiding its true cost. This time one point [00:02:00] of repeat rate is hiding what kind of behavior it's actually measuring, because one point of repeat rate isn't always worth a point either.

So by the end of this episode, you'll have a number that I'm going to call your elective second purchase rate. I don't know if this is an official term. Let's call it your ES- your ESPR. It's your elective second purchase rate. Now, elective, as in the customer actually elected to come back. They chose to.

No subscription, auto-recharging them, and doing the work for them. And this is the cleanest read you can get on whether people actually want your product again. And I'll show you two ways to pull this. One is super simple, and it, y- it's super easy to get. You just have to know how to do it. Now, one very important thing before we continue.

There is a downloadable piece of co- what's, a scorecard to go along with this episode. Make sure you download it. [00:03:00] Um, for episode one and two, in episode one, we had two numbers. I talked to you how to run them for your store. In episode two, we have three numbers that you pull out and a little calculation.

Each one of them is gonna build on, and each time that there's a new episode, we're gonna add n- a new metric. So make sure you grab it today. Just go down in the comments. There's gonna be a link, and you can download the episode so that... Sorry, the, um, scorecard for this episode, so that you can actually put in the information for your specific store that we are measuring.

So make sure you grab the actual PDF from the link in the comments.

Jay Myers: Okay, let's start with why the number you probably check every morning can't actually answer the question that matters.

Revenue and ROAS both look backwards. Revenue tells you money that came in. ROAS tells you an ad that worked last week [00:04:00] before costs shifted. So it's useful, but neither one tells you the thing you actually need to know, which is whether this keeps happening without you paying to make it happen. So you can post a record revenue month but actually be in real trouble, and we actually covered that, how that happens in the last episode.

If that revenue came from first-time buyers who paid $35 to acquire, the revenue chart looks great, but the P&L statement is having a very different opinion. Now, I wanna be careful here because repeat rate is also calculated from past behavior. It's historical too, and anyone who tells you it's a crystal ball is overselling it.

But here's the difference that matters. Revenue measures the result. repeat behavior tells you whether that result has a chance to compound, and whether last year's customers are quietly becoming next year's demand, or [00:05:00] whether you start every January back at zero renting all of them again from Meta. So the one-liner I'd tattoo on the wall is: Revenue tells you what happened, repeat rate tells you whether you have to go and buy it all over again.

And I'm not gonna go and pretend that it's the only number that matters. Margin matters, cash flow matters. There's so many other things. If I said repeat rate was the single thing that predicts everything, some MBA would email me very politely and s- tell me to change or that I was wrong. And you know what?

They'd probably be right. But it's one of the clearest signals of whether your growth is compounding or resetting, and that's the honest version of this claim. So track repeat behavior, good. Except the repeat number sitting on your dashboard in Shopify isn't quite the one you think it is. So let's clear something up, because I've heard people, including me, uh, sometimes say that Shopify's returning customer rate is [00:06:00] misleading. Now, that's not fair, and I wanna correct it.

Shopify's number isn't wrong. It calculates exactly what it says. Your returning customer rate is of all customers who order- ordered during a period you selected, how many of them ordered at some point before, okay? That's Shopify's returning customer rate. Now, that's it, and that's a real correctly computed number.

But the catch is what merchants think it means versus what it actually means. So most people read it as what percentage of my customers come back? It's a loyalty number, but it actually isn't. It's a customer mix number.

It tells you the blend of new versus previously seen buyers in a given window. Two different questions, and you can feel the gap once you say them out loud. Shopify's returning customer rate asks of everyone who ordered this period, how many of them had ordered [00:07:00] before?

The question you actually care about asks, of the customers I acquired, how many came back with a, within a meaningful window and deliberately chose to buy again? That's what you want to know. That second question is, is the one that predicts everything. And there are two things kind of muddying that first number that keeps it from answering the second.

And one I planted back in episode one, we talked a little bit about it. It counts everyone who ordered, even a one-time buyer from three years ago who just re- happened to buy again last month. It's fine to include, but it pads the number and doesn't give you what you actually wanna know. And genuinely, I don't know why the dashboard version reaches all the way back like that instead of using a window. My best guess, it's just simpler to compute at scale, but that's a guess. If someone at Shopify knows the same, the, the reason, I'd honestly love to hear it. In fact, I'd love to have someone at Shopify, [00:08:00] maybe from the analytics team, on the podcast because I'd love to hear it, uh, explained. So the second thing kind of really muddying the water is the big one that I wanna talk about today, uh, which is subscriptions.

So if you run a subscribe and save or a replenishment subscription of any kind, every recurring order counts as a returning purchase. And technically, that's true, but I wanna be precise here because it's easy to be cynical about this, and I don't wanna be. A subscription renewal is real revenue and real retention.

The customer signed up on purpose, and they haven't canceled, and they keep accepting the product every month and the charge, and that's genuinely a good thing. Trust me, that's great. You want subscribers. It's just measuring a different kind of retention. It tells you that the customer hasn't left. It doesn't tell you they came back and made a fresh decision to buy from you again, and those are both worth knowing, and they're not the [00:09:00] same.

A renewal is a continuation. An elected second purchase is a new yes. So you actually, you need both numbers. And here's the insight I want you to take from this episode. A strong subscription program can completely coexist with a weak repeat business everywhere else, and I've seen this many times. The subscriptions can be humming along while the rest of your customers are a leaky bucket, and if you only look at the blended dashboard, you will never see it.

A healthy subscription program can hide a weak retention repeat program on the other side, and that's why we separate them. So let's put some real numbers behind it. I'm gonna go back to the whiteboard here. In episode two, we went to the whiteboard with, uh, with a coffee brand again, so it's familiar. And I'm gonna build this out as...

I'm gonna build out a proper cohort here because if I get sloppy with the, the populations here, the whole number kind of [00:10:00] falls apart, and I'd rather show you the careful version. So first, the dashboard view. So our store's Shopify returning customer rate, let's say it reads 40%. Now remember what that means now.

40% of the customers who ordered in this window had ordered sometime before Renewals, old customers resurfacing, and a real deliberate reorder who a customer chose to reorder from you are all mixed into that 40%. It's not wrong, but it is a blend. And now let's build the honest number separately. And notice I'm not gonna subtract one dashboard figure from another because they measure different things.

We build the cohort from scratch. We pull a group of customers whose very first order was a one-time purchase, not a subscription, and we only include customers who've had a full 12 months to come back. So nobody's [00:11:00] in here who just bought last week and hasn't had a fair chance. The maturity rule matters here.

It's the same principle we used in episode two. So that gives us a clean cohort of, let's say, 8,000 one-time fir- sir, one-time first customers. Now, we ask one question of that group. Within 12 months of their first order, within 12 months, not all time, how many placed another one-time order? A real deliberate non-subscription second purchase.

So in this example, that's 2,000 of them. 2,000 out of 8,000. So that's a 25% elective second purchase rate. Okay? And now the critical thing, say that part slowly, the 40 did not become 25. Those two numbers were never the same thing. 40% describes the [00:12:00] mix of everyone who was reordering during that period. 25% describes whether a specific group of first-time buyers made a deliberate second purchase within a year.

One is a snapshot of a crowd, the other is the behavior of a specific cohort over time, and both are true. They just answer very different questions. And so here's the raw number that kind of stings a little bit and but it's clean because it comes straight out of the same cohort. Of those 8,000 one-time first customers, 2,000 came back, which means, you guessed it, 6,000 of them had a full year to return but didn't.

Now that's your mature one and done count Not they vanished last week. They had every chance, and they just chose not to. Now, is 25% good? Honestly, I can't give you a clean benchmark, and I'm not gonna make one up. That repeat [00:13:00] numbers you see quoted around the internet, the 20 to 40% ranges, those are all blended, um, a lot of them blended dashboards, first of all.

So the numbers that stores report might not be right because the reporting, like I talked about earlier, on the three different blended types of repeat customers, and it's completely different for every vertical too. So almost nobody actually publishes a repeat rate with subscriptions stripped out and for a fixed cohort window, which is kind of the point of this whole episode.

So for now, just don't compare yourself to some industry average report that's measuring something completely different. Compare this cohort to your last one, and compare it to how often someone should naturally rebuy what you're selling. Coffee should be high, mattresses should be low, and that's fine.

Side note, this is actually the kind of exact number that we should publish at Bold to finally give this category of customer a true benchmark. Maybe we will

[00:14:00]

Jay Myers: Okay, now let's kind of get into the tactical part here, how to actually get your number. And there's two ways to pull this, and they're not really beginner versus advanced, they're just two different comfort levels. So same destination. So path one, have Sidekick do it, and this is a great option.

Sidekick is, hopefully you know, but if not, Shopify's built-in AI assistant, little purple glasses icon up on your top in min. You should definitely be using Sidekick and playing with it. It's pretty powerful. Anyways, it's free on every plan, so anyone listening should be able to do this. And it can actually build this cohort report from just a plain English request.

So you describe the report, and it will build the query and, and any filters needed and actually run it. So I'm not gonna read you a, a paragraph-long prompt while you're driving. That's what the show notes are for, but I will have it pasted in there, and it is as well [00:15:00] going to be in the scorecard PDF, so make sure you download that because the exact query to run is in there word for word, so you can just copy and paste it.

But conceptually, you're asking Sidekick for one thing, a cohort of customers whose first order was a one-time purchase, who've had at least 12 months to return. Okay, and I just wanna pause here. What does that mean? Well, if your average reorder time is 30 days, then that should be back 13 months. If the average reorder time is 60 days, it should be ba- back 14 days, so that it means these customers have had enough time to reorder.

So again, I'm gonna just start over. You're looking for a cohort of customers whose first order was a one-time purchase. Second thing, who've had at least 12 months to return to reorder, and you're asking how many of them placed another one-time order in that window, and what is the percentage? And [00:16:00] here's the part I kind of really want you to hear because it's the difference between using AI well and kind of getting burned by it.

Don't just accept the number that Sidekick gives you. Inspect the filters, read the prompt, make sure it's grouped by customer and not by order or by subscription contract. Make sure first order type is actually one time. Just sometimes it doesn't quite get it right, so read the, the query and read the filters when it shows you.

Just make sure that that is all proper and also make sure that the window is fixed. So I've had Sideca- Sidekick like com- nail this, and I've had it, had it sometimes make kind of easy mistakes. So it, uh, generally gets it right, but check it. So just treat its first answer as a smart draft that you verify, not the final number that you screenshot, okay?

That habit will actually serve you well with any AI tool you ever use, not just Sidekick, but in general good, good practice. Okay, so then [00:17:00] path number two, you can build the query yourself. So if you're comfortable in Shopify's QL editor or maybe you just wanna save this as a report that you rerun every quarter and you don't have to go to Side-Sidekick, you can build the exact same cohort by hand.

And the fields you're, you're working with are the ones that flag a customer's first order type, whether an order is subscription or one time. Exact query text is gonna be in the show notes. I've written it out so you can copy and paste it. But one kind of heads-up here those field names, they do get renamed by Shopify from time to time.

Uh, so I don't know when you're listening to this, but if you're listening to this a year from now, that query might, you might not be able to copy and paste, but it'll be a very similar version of that query. But either way if they do get renamed by Shopify you can access the raw editor and, and it should predict which fields you need as you type them in.

So if it throws an error, 90% of the time [00:18:00] that's all it is, it's a renamed field. And the easiest fix, it's actually kind of funny paste the broken query into Sidekick and just say, "This errored. Can you fix it for me?" So if that happens, let Sidekick fix it. The plain English path and the power user path are collapsing into each other very quickly. This is happening with a lot of things in life right now with AI. So either way, you land on the same two numbers, your elective second purchase rate and your mature one and done count.

So now you're gonna take out that scorecard that you printed, the PDF, and you're gonna write them in there. There's a spot to write them. Write them down. Keep this. This is going to be building out a very powerful audit for your store. Now, what do your numbers actually mean? So A number by itself doesn't tell you what to do.

So before you kind of close that down, put your store on a simple grid. So two things, how strong your subscription base is and how [00:19:00] strong your elective second purchase rate is. And that gives you-- So you we're gonna picture like a four-by-four grid here. So, it gives you four profiles, high subscription, high elective, and that's the dream.

People happily auto-replenish, and your one-time buyers happily come by on their own. And if you're, if this is where you are, uh, your job's done. Just don't break it. Or do you have high subscription but low elective? And this is the sneaky one, and this is actually the whole reason for this episode is your subscription engine looks great and revenue's fine, but if you strip that away, the rest of your customer profile barely repeats, and you're more dependent on people not canceling than you'd like to admit.

A lot of people get in this situation. They have what they call gym membership subscriptions. They don't even wanna message the customer. They don't even wanna email the customer because they're afraid they might cancel. People aren't choosing [00:20:00] to come to the gym and pay every month. They're just on a subscription that they forget to cancel, or it's too much effort to cancel 'cause it's not that expensive.

That's not someone choosing every month to come and pay. So the work here is gonna be building elective repeat outside of the subscription, so the business isn't the one thing kind of holding up the business, okay? And then you're gonna have another quadrant, and I-- This is on that scorecard, by the way.

On the PDF, there's a visual of this. But another quadrant is gonna be low subscription, high elective, and this is your customers come back on their own without being locked into anything, and that's actually very healthy. And there might be a subscription opportunity here, but be careful. Forcing a subscription on people who are happily rebuying can actually annoy your best customers.

Don't fix what's not broken. We work with a brand who I... is a perfect example of this. Uh, they are using our RePete app, which is an [00:21:00] AI reorder agent that nudges people to reorder with a click, so not a subscription. People who buy things one time And just choose to come back. And they also have a subscription.

The reorders through RePete, R-E-P-E-T-E, the app, are four times what their new subscriptions are every month. So, and their LTV on those customers is higher. So the customers who choose to come back are actually spending more, they're buying more, and they're buying more often. They just don't want a subscription, and that's totally fine.

So, um, that's actually a very healthy store. They're choosing to come back every month. And then the fourth quadrant is low subscription, low elective. Okay. This is the one that actually needs real attention, and no shame if it's you because, well, at least now you actually know. And this is a gen-- this is a true repeat commerce problem, and it's usually one of four things: the product experience, the replenishment [00:22:00] timing, reorder friction, or honestly, just product market fit.

And conveniently, those are basically what the next several episodes of this series are going to cover. See, as you can tell, we are setting this up. So find your quadrant, do the math, run the report, find your quadrant, and that's gonna really impact how you listen to the next few episodes.

And it's gonna tell you whether your next move is to protect, build, leave alone, or investigate. So make sure you do this today. Your 1% win today, you're gonna go, you're gonna grab that scorecard if you haven't already. You're gonna get your elective second repurchase rate. Now, if you didn't pay attention during this episode, there is a cheat sheet on the, the scorecard and an example how to get it.

And you're gonna get that, and you're gonna get your mature one-and-done count. Open Sidekick, paste the prompt from the show notes, inspect the filters just to make sure [00:23:00] it's done properly, and then read your two numbers. Add both to your Repeat Commerce scorecard. It's gonna be on line six and line seven, and underneath the numbers from the last two episodes.

And I'll just be straight with you, this is the first scorecard number designed to make you maybe a little bit uncomfortable. The earlier ones were mostly neutral and just kind of like eye-opening. This one, it might sting some listeners, and that's good.

You can't fix a leak that you don't know about. And I have seen so many stores, they think they're crushing it, healthy, healthy. Every time I talk to them, "How are things?" "Oh, busy, doing great, busy." And then just one month they're out of business because there were problems they didn't know about. They thought they were healthy.

They thought they were growing. It wasn't healthy revenue. The right type of repeat revenue wasn't coming in. So even if it stings, at least you know it. So, okay, quick recap. [00:24:00] Revenue and ROAS look backwards, so they can't tell you if your growth is gonna compound. Your dashboard's returning customer rate, it isn't wrong, it just answers a different question.

And I talk to so many stores just because we're in the repeat commerce space and we have a rePete app. I often ask brands, "Hey, what's your, what's your repeat rate? What's your..." And they always quote that number, and they think that's their reorder rate, and it's not. So I bet you 90% of people use that number as their reorder rate.

So it's just a blended number that includes three very different things. It's a mix of who's reordering. It's not whether your customers are actually deliberately choosing to come back. So separate out subscriptions because a renewal is, it is real retention, but it's a continuation. It's a gym membership often.

It's not a fresh decision every month. Now build a proper cohort one-time first customers with a 12-month window to return, and you get your [00:25:00] elective second repurchase rate, plus the count of the one-and-done customers who didn't come back. Then drop your store into the four-profile grid, and you can just mark it on the scorecard to know what to actually do about it.

And next episode, we're gonna start-- stop measuring. I know these first three episodes have been a lot of measuring, and you're probably thinking, "Jay, just tell me what I gotta do to fix it." So, that's coming. We're, um, we're gonna start fixing right where the leverage is the highest, and that is the second order.

The jump from the first purchase to the second, it's the riskiest, it's the most value gap in your whole entire customer likes-- life cycle. There is no order more important than the second order. N- like, we all celebrate orders, but we should all-- Like, last month, how many second orders did you get? I bet you brands don't know.

But it's the most important number, and it's the most [00:26:00] important one to get. And I'm gonna break down in the next coming episodes how to actually engineer it to happen. So that's it for this one. I hope you got 1% better today. That's the whole purpose of this show. If you did, and if you're not subscribed, hit that subscribe button, and if there's anything you're taking away from this episode, please leave a five-star review.

That's how we get found. I read every single one. Leave a little co-comment on it. That would be great. Thank you so much for listening. I will see you on episode four of the Repeat Commerce Series.