Replenishment reminders and subscriptions solve the same underlying problem, a customer needs more of something they’ve bought before, using two structurally different mechanisms. One asks the customer to decide again each time. The other pre-commits the decision. Neither is universally better, and picking the wrong one for a given product creates friction that shows up as poor conversion in one model or high churn in the other.
Two mechanisms, one underlying job
Both replenishment and subscription exist to solve depletion: a product runs out, and the customer needs a way to get more of it without starting the buying decision from scratch. A replenishment reminder does this by prompting a fresh decision each cycle, timed to when the customer is likely running low. A subscription does this by removing the decision entirely after the first signup, defaulting the customer into a recurring shipment unless they actively intervene.
The job is the same. The customer experience, and the failure modes when the model doesn’t fit the product, are very different.
When replenishment fits better than subscription
Replenishment reminders tend to outperform subscription for products with variable or unpredictable consumption. A product used at different rates depending on season, mood, or circumstance doesn’t fit a fixed-cadence commitment well, the subscription will consistently arrive too early for some cycles and too late for others, generating skips and eventual cancellation regardless of how accurate the initial cadence estimate was.
Replenishment also fits better for products where the customer values choosing the moment of purchase, a considered or occasional-use item where being asked each time, rather than defaulted into a standing shipment, feels appropriate to how the customer thinks about the purchase. Forcing a subscription onto a product like this can generate resentment even among customers who would happily reorder manually when prompted.
When subscription fits better than replenishment
Subscription tends to outperform replenishment for products with steady, predictable consumption, a daily-use item consumed at a consistent rate is the clearest case, where the friction of deciding again every cycle adds no real value and only creates opportunity for the customer to forget, delay, or drift away entirely.
Subscription also fits better where the retention value of removing the decision outweighs the flexibility cost. A product with a strong, stable fit for a customer’s routine benefits more from being locked in via subscription than from being re-earned through a fresh decision every cycle, since each fresh decision is also a fresh opportunity to lapse.
The hybrid signal: letting purchase behavior decide
Rather than assigning a model to a product permanently at the catalog level, the more accurate approach lets individual customer behavior inform which model fits, since two customers buying the identical product can have genuinely different consumption regularity. A customer with tight, consistent reorder gaps is a strong subscription candidate regardless of the general product category. A customer with irregular gaps on the same product is a better fit for ongoing replenishment reminders, even if the product overall skews toward subscription for most customers.
This connects directly to the subscription-conversion timing logic covered elsewhere in this cluster: the second unprompted reorder is the strongest signal a customer has demonstrated the kind of regularity that makes subscription a genuine fit, rather than a commitment being asked for before that regularity is established.
What happens when the wrong model gets applied
Two failure patterns show up consistently when a product gets the wrong model:
- Subscription forced onto an irregular-consumption product generates a pattern of skips almost every cycle, since the fixed cadence rarely matches actual need. Each skip is a moment where cancellation becomes newly possible, and the cumulative effect is a subscription program with high signup but weak long-term retention on that specific product.
- Replenishment reminders relied on for a steady, high-frequency product leaves retention dependent on timing accuracy and the customer’s continued attention every single cycle, missing the retention lock-in a subscription would have provided for a product genuinely suited to one. Even accurate replenishment timing can’t match the retention strength of a well-fit subscription, since it requires the customer to actively decide again every time.
Running both models on the same catalog
Most consumable brands end up running both models simultaneously across different parts of the catalog, and this is generally the right outcome rather than a sign of an unfinished strategy. A catalog with both steady, daily-use products and variable, occasional-use products should offer subscription prominently for the former and lean on well-timed replenishment reminders for the latter, rather than forcing a single model across the whole catalog for consistency’s sake.
Existing customers convert at meaningfully higher rates than new prospects, 60% to 70% likely versus 5% to 20% for a cold prospect, which is the underlying reason both models are worth building well rather than defaulting to whichever is easier to implement. Subscription share of revenue in the 15% to 25% range is a reasonable target band for a consumable brand with a genuine mix of product types, reflecting that even a well-run program typically has a meaningful share of revenue that makes more sense staying in the replenishment model rather than being pushed toward subscription regardless of fit.
Deciding per product, not per catalog
In practice, the decision should be made at the product level, informed by actual consumption data rather than assumption:
- Measure consumption regularity per product, using the same reorder-interval data that drives replenishment timing generally. Tight, consistent intervals suggest subscription fit. Wide variance suggests replenishment fit.
- Layer in customer-level regularity for products that sit in the middle, letting individual behavior refine the product-level default rather than applying one model uniformly.
- Avoid pushing subscription onto a product primarily to hit a program-wide subscription share target. A subscription share target achieved by forcing poor-fit products into the model produces weak retention numbers that undermine the metric it was meant to improve.
- Revisit the assignment periodically. A product’s typical consumption pattern can shift with reformulation, pricing changes, or seasonal demand shifts, and a model assigned once and never revisited can drift out of fit the same way a fixed reorder-timing assumption does.
FAQ
Should every consumable product offer a subscription option?
No. Products with steady, predictable consumption tend to fit subscription well. Products with variable or occasional consumption tend to generate poor-fit subscriptions that skip frequently and eventually churn, and are usually better served by accurate replenishment reminders instead.
How do you know if a specific customer is a good subscription fit, even on a variable-consumption product?
Look at that customer’s own reorder regularity rather than the product’s general pattern. A customer with tight, consistent purchase intervals can be a strong subscription candidate even on a product that’s generally better suited to replenishment reminders for most customers.
Is it a problem to run both replenishment and subscription across the same catalog?
No, this is generally the right approach for a catalog with a genuine mix of product types. Forcing a single model across the whole catalog for consistency usually produces worse results than matching the model to each product’s actual consumption pattern.
What happens when subscription gets forced onto the wrong product?
It tends to generate frequent skips, since the fixed cadence doesn’t match actual need, and each skip becomes a fresh opportunity for cancellation, resulting in high signup but weak long-term retention for that specific product.
Should subscription share of revenue be treated as a target to hit regardless of fit?
No. Pushing poor-fit products into subscription to hit a share target tends to produce weak retention that undermines the metric’s usefulness. The target should reflect genuine fit across the catalog, not be engineered by force.
Close
Replenishment and subscription aren’t competing strategies, they’re two tools suited to different consumption patterns, and the products in a catalog rarely all fit the same one. Brands that assign the model based on actual consumption data, rather than defaulting every product toward whichever model looks better on a dashboard, get stronger retention out of both.