Every store owner who sells something people use up hears the same pitch sooner or later. Add subscribe-and-save. Recurring revenue. Predictable months. Customers who come back without being asked. And it is all true. We have said it ourselves: in our piece on repeat purchases, we called subscribe-and-save the strongest retention tool there is for a consumable product, because it turns a repeated decision into a single default.
What the pitch leaves out is that the checkbox on your product page is not a feature. It is a second business, bolted onto the first. A one-time store sells an order and is done with it. A subscription store makes a promise to deliver the right amount, at the right time, at the right price, every month, and then has to keep that promise to hundreds of people at once. It forecasts differently, stocks differently, ships differently, bills differently and talks to its customers differently.
Some stores should absolutely run that business. Plenty should not. Here is how to tell which one you are before you add the box, and how to build it properly if you do.
Check the Product Before You Check the Box
Start with the product, because no offer and no email sequence will rescue a product that does not fit the model.
Matt Holman, who runs a consultancy devoted entirely to subscription programs, put it bluntly on Kurt Elster's podcast: subscriptions are "just a mechanism for repeat purchases." If the repeat purchase is not already there, the mechanism has nothing to work on. His examples of stores that burn time forcing it: bed sheet subscriptions, pillowcase add-ons, and a giant jar of chili powder that gets used a tablespoon a week. His test is the one worth remembering. If you have to educate a customer twice, once on why your product is good and again on why it would ever be a subscription, you are fighting uphill.
Run your product through four questions:
- Does it get used up? Coffee, pet food, supplements, skincare, furnace filters, razor blades. Not sheets, not a cast-iron pan, not a jacket.
- Does it get used up at a steady rate? A daily coffee runs out on a schedule. A seasoning, a party snack or a special-occasion drink runs out whenever it feels like it, and a fixed delivery date will be wrong most months.
- Can they grab it on the way home? Holman points out that snacks and heavy beverages are consumables, yet they make poor subscriptions, because the gas station and the grocery store are more convenient than your box. A subscription has to beat the errand.
- Is running out a real annoyance? Nobody wants to discover the dog food bag is empty at 9pm. If running out costs your customer nothing, never running out is worth nothing to them.
If you pass all four, you have a candidate. If you are not sure, the cheapest test in ecommerce is available to you: turn the option on, leave it quiet, and watch. Holman's rule of thumb is that if five to ten percent of buyers take the subscription without any push, there is real demand to build on. If almost nobody does, the product has told you something no amount of merchandising will change.
The Discount Is the Weakest Reason Anyone Subscribes
Most stores build their subscription offer the same way. One-time purchase at full price, subscribe-and-save at 10 or 15 percent off, a toggle between the two. The discount is the whole pitch.
That has a problem you can see in your own data this afternoon. When the only difference between the two options is price, a slice of your customers will pick the subscription, take the discount, and cancel before the first renewal. The team behind the How Shoppers Think channel describes exactly this pattern and the check that exposes it: open your subscription app and count cancellations on day zero, the same day as the first order. If that number is high, you do not have subscribers. You have a coupon with extra steps, and every one of those orders went out at your subscription margin.
It gets worse for the discount. Space Goods, a subscription-first mushroom-coffee brand that does roughly 80 percent of its business on subscription, ran a clean split test: one version of its subscription priced 25 percent cheaper than the other, sold side by side at the same time. Price was the top reason customers gave when they canceled, so the cheaper version should have kept people longer. It did not. Retention came out about the same in month one, two and three, and customers still cited cost at nearly the same rate. Once people settle into a price, they get used to it, whatever it is.
So what should the subscription offer be built from? Three things work better than a bigger percentage:
- Things that make the product easier to use. Space Goods sells its subscription as a starter kit with a whisk, a mug and a measuring spoon, because the people who make their first cup right are the ones who stay. The gift is chosen to fix the most common reason people drift away.
- Value that grows the longer they stay. Instead of all the savings up front, some stores step it up: a small discount on the first order, a larger one by the third, a free add-on at month four. That rewards the behavior you actually want, which is staying.
- Control. Skip a month, swap a flavor, change the date, pause for a trip. The biggest hesitation before subscribing is fear of getting stuck, so make the flexibility visible on the product page, not buried in a portal.
None of that requires a deep cut. If your offer still feels like it needs one, the problem is usually upstream, and it is the one we covered in You Discount Because You Never Built an Offer.
Your Delivery Schedule Is a Promise About How Fast They Use It
Ask anyone who runs a subscription program why people cancel, and the same answer comes back near the top: too much product. Max Sturtevant, whose agency runs email for large subscription brands, calls it the main reason people leave. The closet fills up with bags nobody opened, and canceling is the only way to make it stop.
That is not a customer problem. It is a scheduling problem, and it is yours. Every delivery interval you offer is a claim about how fast your customer gets through the product. If you sell a pack of drinks on a 30-day cycle to a single person who drinks one every other day, you have promised them a stockpile.
Three fixes, in order of effort:
- Set the default interval from real usage, not from the calendar. Thirty days is a convenient number, not a fact about your product. Look at how long one-time buyers actually take to reorder, and start there.
- Ask when usage varies. One pet-product brand that consultant Daniel Budai worked with sells drops dosed by the animal's size, so a small dog and a large dog empty the same bottle at very different speeds. One question at signup gets each customer the right interval from the first order. If your product finder or quiz already collects this kind of answer, use it here; declared answers like these are worth more than anything you can infer.
- Make adjusting easy, and treat it as good news. Holman describes a study across roughly 200 brands in which subscribers who adjusted their delivery frequency three times were worth about twice as much as those who did not, and subscribers who changed the product or flavor three or more times were worth about six times as much. A customer who is tuning the subscription is a customer who intends to keep it.
Churn Sets a Ceiling, and You Can Calculate It Today
Here is the piece of arithmetic that every subscription store runs into eventually, usually by surprise.
Every month, some share of your subscribers cancel. That share is your churn rate. Every month, some number of new people subscribe. As your base grows, the number canceling grows with it, because 10 percent of 300 is more than 10 percent of 100. Eventually the people leaving each month equal the people arriving, and the base stops growing. Holman describes this as the most common problem he sees in programs that "stopped growing": nothing broke. The math simply arrived.
The ceiling is easy to find. Divide your new subscribers per month by your monthly churn rate.
A small store adding 20 new subscribers a month with 10 percent monthly churn tops out at 200 subscribers (20 divided by 0.10). It will not feel like a ceiling on the way up. After a year it will have about 140, and after two years about 185, and every month the growth gets smaller while the effort stays the same.
There are only two ways to raise it. Bring in more new subscribers, or lose fewer. Double the new subscribers to 40 a month and the ceiling doubles to 400. Cut churn to 5 percent and it also doubles to 400. Holman's advice to smaller programs is that the first lever is usually the faster one, because shaving a point off churn can take hours of work on emails and cancel flows while a better offer can lift sign-ups right away. Both matter. The point is to know which one you are pulling, and to know your ceiling before you plan a year around a number your program cannot reach.
Know, too, where the churn actually happens. It is rarely spread evenly. Budai shared the numbers from one pet-product account: about 15 percent of subscribers had canceled after the first month, close to half by the second, and 60 percent by the third. After that, the losses slowed sharply. That is one store, not a benchmark, but the shape is common. The subscription is won or lost in the first 90 days, which is why the last section of this piece is about the first month.
You Just Became a Billing Company
A one-time store takes a payment once, at the moment the customer is most excited. A subscription store takes a payment every month, at a moment the customer did not choose and may have forgotten about. That changes your job in ways most owners do not anticipate.
The receipt is your riskiest email. On a churn-by-day chart Sturtevant shared from one large client, the single biggest spike in cancellations landed on day 30, the day of the first renewal charge, when the customer received a plain receipt announcing that they had just been billed again. Nothing about the product changed that day. The reminder of the charge did the damage. Your renewal notice and your receipt should do what any good sales page does: remind them what they are getting and why they wanted it, and put "skip" and "change date" ahead of "cancel."
Some payments will fail. Cards expire, get replaced after a fraud alert, or hit a limit. In the pet-product account above, 11 percent of first renewals failed at the payment step. Those are customers who did not decide to leave, and you will lose them anyway unless something retries the card and asks them to update it.
The rules apply to you now. Selling on a recurring charge brings legal obligations that a one-time store never meets. At the federal level, the Restore Online Shoppers' Confidence Act requires online sellers to disclose the recurring terms clearly before taking payment, get the customer's express consent to the recurring charge, and provide a simple way to stop it. The FTC's broader "click-to-cancel" rule was struck down in court in 2025, and the agency reopened the rulemaking in 2026, so expect that area to keep moving. Around 25 to 30 states also have their own automatic-renewal laws, and some go further; California, for example, requires that a subscription bought online can be canceled online. Reminder requirements vary by state and often depend on the length of the term. Check the rules in the states you sell into, or ask your subscription app what it handles for you. None of this is hard, but none of it is optional.
Your inbox changes shape. Subscribers write in to skip, swap, pause, change an address, ask where the order is, and ask why they were charged. Those requests arrive whether or not you have time for them, and every one answered badly is a cancellation in waiting.
Your shipping rhythm changes too. Renewals cluster around the dates people first subscribed, which often means around your promotions. If you ran a big push on the 1st, a wave of orders comes back every month on the 1st, and your packing table needs to be ready for it.
Subscribers Are the Easiest Demand You Will Ever Forecast
Here is the part of the new business that is genuinely better than the old one.
A one-time store guesses at next month. How many new customers will the ads bring? How many past buyers will come back? Every reorder of inventory is a bet on that guess, and we have written about how much cash those bets tie up in You Did Not Lose the Money. You Turned It Into Inventory.
Subscribers take a large piece of the guessing away. You know who is scheduled, what they are getting and when. Subtract your normal churn and you have a forecast for that slice of your demand that is better than anything you will ever get for new customers. Use it. Plan subscriber stock first, keep a buffer on anything people subscribe to, and treat new-customer demand as the uncertain layer on top.
The flip side is that running out hurts more. When a one-time buyer finds a product out of stock, you lose a sale. When a subscriber's renewal cannot ship, you break the promise the whole arrangement rests on, and a customer who was on autopilot is suddenly forced to reconsider. A skipped shipment you caused is often the first step toward a cancellation.
One more piece of the money deserves a clear look. A subscription discount is not a one-time cost. If you sell a $40 bag of coffee at 20 percent off to subscribers, you give up $8 on every renewal, not just the first. A subscriber who stays eight months has received $64 in discounts. On a product with a 50 percent margin, that discount is eating 40 percent of your profit on every single order. It may well be worth it, because the customer would not have made eight separate trips back to your store on their own. But run that number before you pick the percentage. And if you are weighing how long you can wait to earn back what it cost to win each subscriber, that is its own calculation, which we worked through in Lifetime Value Does Not Pay This Month's Invoice.
The First Month Decides Whether There Is a Second
If most of your churn happens in the first 90 days, most of your effort belongs there too.
The goal of the first few weeks is simple: get the customer to use the product correctly and often enough to see why they bought it. Space Goods checks in with new subscribers at the two-week mark and asks how they are using the product, because it found that the people who drink it first thing in the morning, hot, made with milk, and daily in those first two weeks stay far longer than everyone else. So its first week of emails teaches exactly that routine. Your product has its own version of that pattern. Your best long-term subscribers are doing something specific in their first month, and the job of your onboarding is to find out what it is and teach it to everyone else.
Space Goods found another lever that cost nothing. In a test, new subscribers were asked to set a personal goal for the next 90 days at the start. Same product, same price. Retention in months two, three and four improved by more than 20 percent. People who have said out loud what they want from the product give it long enough to deliver.
Two cautions go with this.
First, send less than you think once the onboarding is done. Every email to an active subscriber is also a small prompt to reconsider the subscription. Sturtevant describes a client whose regular promotional campaigns were triggering more cancellations than new subscriptions among the subscribers who received them. Keep subscribers out of your general sale blasts, especially any sale that undercuts what they are paying, and give them a small number of genuinely useful messages instead. If you do not yet have the basic automated emails in place, start with the ones in Your Store Sends One Automated Email. It Should Be Sending Five.
Second, measure against a group that did not get the messages. The effect of subscriber emails moves around from month to month, and the only way to know whether a check-in is helping or hurting is to hold some subscribers back and compare.
So, Should Your Store Offer One?
Here is the short version.
Offer a subscription if your product gets used up at a steady rate, running out is a real annoyance, and you can deliver it more conveniently than the nearest store. Build the offer from things that make the product easier to use and value that grows over time, not from a bigger discount. Set the delivery schedule from real usage, let people adjust it, and count day-zero cancellations to see whether your offer is attracting subscribers or coupon hunters. Work out your ceiling, stock for your subscribers first, and put your effort into the first month.
Skip it, or leave it as a quiet option, if you would have to talk customers into the idea of a subscription before you could talk them into the product.
Either way, make it a decision about which business you want to run, not a checkbox on the product page. If you want help working through the numbers for your own store, from whether a subscription makes sense to how it fits with your ads and email, that is exactly the kind of thing our ecommerce team does every day. And if you would rather build it yourself, we hope this gives you a clear place to start.
