Hey everyone,
Welcome back for another bite to chew on.
Every consumable brand has two kinds of repeat customers:
Subscribers, who already said yes.
And everyone else: the one-time buyers and lapsed customers who come back when they feel like it, on their own schedule.
Most brands pour their retention budget into the first group and point one generic winback flow at the second.
We did the same thing at Obvi for years.
We built a subscription program we are proud of, and it earns its keep. But no brand converts every repeat buyer into a subscriber, and we knew a meaningful share of our reorders came from customers who buy on their own terms.
We just did not have a system built for them.
That’s why we started running rePete by Bold, an AI reorder agent that learns each customer's buying pace and nudges them at the right moment.
The results were strong enough that Bold wrote a case study on it:
Today, we want to walk you through the receipts.
Let's get into it.
On the Menu:
The two repeat audiences we were treating exactly the same (and what that cost us)
Why a fixed calendar can never predict a reorder, and what does
The receipts: what a purely additive reorder channel did to our repeat revenue
Claim $10K in reorder revenue
If your repeat revenue strategy is "subscriptions plus a winback flow," you are leaving money with the customers who buy on their own terms.
Those customers do not want a billing cycle.
They want a reminder when they are actually running low.
rePete is built for exactly that segment.
rePete by Bold is an AI reorder agent for Shopify brands. It studies every customer's order history and predicts when each one is likely to run out of each product.
At that moment, it sends a friendly nudge by email, SMS, on-site, or push.
One click later, the shopper is checking out a restock with the right add-on already in the cart.
Per-customer reorder predictions: A prediction for every customer and every product, based on their real buying rhythm instead of a generic 30-day cycle. Every reorder makes the next prediction smarter.
Smart Cart: When a customer clicks the nudge to confirm a reorder, Smart Cart suggests the add-on that fits, turning a simple restock into a bigger basket.
Subscribers stay untouched: rePete works only with the customers outside your subscription program, so every reorder it drives is additive. And you only pay when it produces reorders.
Chew on This readers get their first $10,000 in reorder revenue courtesy of rePete!
The two repeat audiences we were treating the same
Subscribers need patience. Everyone else needs timing.
Obvi's repeat business had two very different audiences inside it, and for a long time we ran one global outreach strategy across both.
Subscribers had already committed. What they needed from us was patience and flexibility: easy skips, easy swaps, no pressure.
One-time and lapsed buyers needed the opposite. They needed well-timed encouragement to come back.
One strategy for both meant we got it wrong in both directions.
We were too aggressive with loyal subscribers who did not need a nudge, and not aggressive enough with the one-time buyers who were quietly slipping away.
Nobody churns loudly. They just don’t come back.
We assumed we knew when customers were done
Underneath that sat a trickier problem: we had no real way of knowing when a customer was actually finishing their product.
Outreach went out when it was convenient for us to assume they were done, not when they were ready to buy.
We never really considered when somebody was finishing their product. We just assumed we knew, and that is when we would ask them to buy.
We never flipped it around and asked: is it convenient for you yet? Are you ready?
Once you see that, it is hard to unsee. Every replenishment email on a fixed delay is a guess dressed up as a strategy.
Consumption is personal. Calendars are not.
The rePete team frames this with three shoppers buying the same tub of protein.
The athlete finishes it in 21 days.
The household buyer, feeding a family, gets through it in 35.
The weekend buyer takes 46 days, and a monthly subscription would just pile up in the cupboard.
Same product, three completely different reorder windows.
A 30-day replenishment flow is early for one, roughly right for another, and more than two weeks late for the third.
Multiply that across a catalog with as many flavors and formats as ours, and a calendar-based approach was never going to catch the customers who buy by the scoop.
Why a calendar cannot predict a reorder
Learn the rhythm, then predict the window
rePete starts by studying every customer's order cadence and product behavior.
From there, it builds a reorder prediction for each customer and each product: when that specific person is likely to run out of that specific item.
Not a segment average. A per-person, per-product window.
When the timing is right, a nudge goes out.
The channel adapts to where the customer actually is, whether that is email, SMS, on-site, or push.
The message reads like a helpful reminder rather than a promotion. And every reorder that comes back through the system feeds the next prediction, so the model gets sharper the longer it runs.
The nudge ends in a bigger basket
We also turned on Smart Cart.
When a customer clicks a nudge to confirm their reorder, Smart Cart suggests the add-on that pairs with it.
For a brand like ours that launches new flavors and formats constantly, that matters more than it sounds.
Every drop now gets a warm, well-timed audience that is already in a buying moment.
That is the difference between a reorder reminder and a reorder channel.
One brings back the same order. The other grows it.
Additive by design, and priced like it
The part we cared about most: none of this touches our subscribers.
rePete works exclusively with the customers outside the subscription program. Every reorder it drives is purely additive, not pulled forward from a subscription we would have collected anyway.
The pricing matched the promise.
What sold us was simple: if it does not do anything, you do not pay anything.
We are all spending a lot of money on a lot of tools right now, and we need every one of them to work.
Setup took about 20 minutes on a single call, and the first reorder predictions were live within 24 hours.
The receipts
The return, without a dollar of acquisition spend
rePete was built around three goals: turn one-time buyers into repeat customers, get repeat customers reordering more often, and get them spending more when they do.
We are seeing results on all three.
On the return itself, rePete has driven a 36x ROI for Obvi.
Not one dollar of that went toward acquiring a new customer.
All of it came from people who had already bought from us once and simply needed to be asked at the right time.
Better orders from better-timed buyers
The order economics surprised us more than the topline.
Reorders through rePete carry a 28% higher average order value than our other reorders.
Customers who reorder through rePete generate 13% higher revenue per customer than our other repeat buyers.
A lot of that is Smart Cart doing its job.
Nearly a third of rePete reorders (32%) have been upsized with an add-on at the moment of confirmation.
When you reach a customer at the exact point they are running low, they are open to trying the new flavor too.
A channel that did not exist before
rePete is now responsible for 6% of all of Obvi's repeat purchase revenue, and that number is climbing.
That is revenue flowing through a channel we did not have before.
And through all of it, our subscription program is exactly as strong as it was.
rePete did not move a single subscriber. It found the repeat revenue that was going uncaptured.
We scaled Obvi past $100M in five years, and one of the biggest wins we have found lately had nothing to do with ads or subscriptions.
It was capturing the repeat revenue we did not even know we were leaving on the table.
Honestly, rePete is the tool we wish we had from day one.
Sum It Up
Subscriptions cover part of your repeat revenue.
The rest sits with customers who buy when they are ready, and most brands have no system built for them.
On the two audiences: Running one outreach strategy across subscribers and everyone else means over-pressuring the loyal and under-serving the ones slipping away.
On timing: A fixed replenishment calendar guesses when a customer is done. A per-customer, per-product prediction knows, and it gets sharper with every reorder.
On the receipts: For Obvi, a reorder channel aimed only at non-subscribers became a growing slice of repeat revenue, with bigger baskets and zero impact on subscriptions.
To read our FULL case study, click below:
If you sell anything people consume by the scoop, the pour, or the pump, you have a version of this gap in your own numbers.
The first $10,000 in reorder revenue rePete earns you is on them.
Book a call with the Growth Team and see what it finds in your store.
Let us know how we did...
All the best,
Ron & Ash