Updated · 7 min read
Replenishment emails: the lifecycle flow that buys itself
A user buys coffee beans from you. Three weeks later, they're running low. Four weeks later, they've grabbed another brand from the supermarket because yours never asked. That's the silent churn replenishment exists to prevent — a one-message intervention with the highest revenue-per-send of any lifecycle flow, and the one most programs still aren't running.

By Justin Williames
Founder, Orbit · 10+ years in lifecycle marketing
Why this flow earns more per send than anything else you ship
Picture a customer halfway through a bag of your coffee. They'll be out by Sunday. The email that lands on Friday — "running low?" with a one-click re-order — isn't selling them anything. They've already chosen you. It's the reminder that keeps them choosing you instead of grabbing whatever's on the supermarket shelf at 8pm.
That's replenishment. It applies to consumables — coffee, skincare, pet food, vitamins, contact lenses, household supplies — anything with a predictable buy-use-run-out-buy-again cycle. The flow exists to catch the user at the "run out" moment.
User has already chosen your product once. They're satisfied enough to be re-orderable. The replenishment email doesn't persuade. It reminds, and makes the re-order one click.
Revenue per send is typically 5–15× what you get from a marketing broadcast — the regular promotional emails sent to your whole list — because the intent is already there. The job isn't to convince anyone of anything. It's to close the gap between "user is running out" and "user notices they're running out and remembers where to buy." That gap is smaller than any other lifecycle trigger has to close. Smaller gap, same purchase, much higher revenue per send. That's the whole punchline.
Three messages, in the order users actually need them
The flow runs three messages, each one calibrated to a different point on the depletion curve. Most programs over-engineer this. Three is enough.
Message 1 — "heads-up, you're nearly out." Lands 5–7 days before the user's typical re-order window, calculated from their purchase history (more on the maths in the next section). Subject: "Running low on [product]?" Body: one-click re-order of the same product, an estimated delivery date, a link to adjust quantity or switch to subscription if it's relevant.
Message 2 — "you might actually be out by now." Lands 10 days after message 1 if they haven't re-ordered. Subject: "Time to restock [product]." Body: re-order CTA, gentle urgency, alternative sizes if relevant. This one catches the people who meant to re-order, opened the first email, then got distracted by their actual life. A non-trivial share of revenue lives in this message alone.
Message 3 (optional) — "stop having to think about this." Lands 30 days after the original purchase if there's still no re-order. Subject: "Never run out — set up auto-delivery." Body: subscription pitch, usually with a small discount, framed around convenience rather than savings. Converts one-time buyers into subscribers. Lower conversion rate than messages 1 and 2, but every one that lands is a meaningful LTV — lifetime value, the total revenue you'll get from that customer — bump.
,
The hardest part: predicting when they'll actually run out
The whole flow hinges on one number — the day the user is likely to run out. Get it right and message 1 lands while the bag is half-empty. Get it wrong by a fortnight and you're either annoying people who just bought, or arriving after they've already replaced you.
Three approaches, in increasing order of complexity. Pick the simplest one your data supports.
Category default. Every product gets a default depletion window: coffee 30 days, skincare moisturiser 45 days, 5kg dog food 60 days. Use the category default unless you have user-specific data. Simple, fast to ship, accurate enough for most products. This is where 90% of programs should start, and where many should stay.
Per-user average. For a customer who's re-ordered multiple times, use their own re-order rhythm. Someone who reliably re-orders coffee every 21 days gets message 1 on day 15 — not on the category default of day 23. Tighter, more accurate for returning customers. Requires historical purchase data, which means six-plus months of repeat-buying behaviour to be useful.
Product-plus-size-plus-household. Someone who buys the 1kg bag depletes faster than someone who buys 5kg. A household-size signal — pulled from signup data or inferred from past purchases — refines further. Lifts accuracy 20–30% over the category default but needs cleanly tagged attribute data most programs don't actually have.
The path is: start with category default. Move to per-user average once you have six-plus months of repurchase data. Only build the size-plus-household model if the revenue genuinely justifies the engineering. Most programs don't need to. The ones that build it anyway often don't measure whether it actually moved the number, which is its own problem.
Subscription is the real prize — replenishment is the bridge
Replenishment email is, fundamentally, a workaround. The long-term play is subscription — auto-ship at the predicted interval, clear path to pause or cancel, no email needed. Every replenishment send is a customer you haven't converted to subscription yet.
So treat the flow that way. Every replenishment message should carry a subtle "or set up auto-delivery and skip the reminders" option. The path from one-time buyer → replenishment responder → subscriber is the LTV compound that makes consumable businesses actually work — and it's the path most programs leave on the table.
Measure subscription conversion from replenishment as a separate KPI from one-time re-order conversion. A 10% subscription conversion rate is worth roughly 3× the LTV of a 30% one-time re-order rate. Don't blend them.
So why not just push everyone to subscription from day one? Because subscription has a higher commitment cost for the user — you're asking for a recurring relationship before they've decided your product belongs in their routine. Many first-time buyers aren't there yet. Replenishment bridges the commitment gap: get them through a second and third one-time purchase, prove the product works, then convert when it's clearly part of their week. Push subscription too early and you actually suppress first-purchase conversion. Worst of both worlds.
Measuring whether any of this is working
Four numbers worth instrumenting. The first one is the only one that proves the flow is doing its job; the others tell you why.
Re-order rate in the next 14 days, with a holdout. A holdout — the random subset of users you deliberately don't message, your control group, the way a clinical trial works — is the only honest way to measure incremental lift. Replenishment sent vs holdout suppressed. Expected incremental lift on 14-day re-order rate: 15–30% for consumable categories. Below 10% and the flow needs a copy or timing pass.
Time-to-reorder. Median days from previous purchase to next. Replenishment should shorten this by 5–10 days vs the holdout. Reminded users re-order sooner — that's the whole mechanism, and it's worth confirming it's actually happening rather than assuming.
Subscription conversion from replenishment. Percent of message recipients who set up auto-delivery. 2–8% is healthy. Below 1% means the subscription CTA isn't being noticed — usually because it's buried at the bottom of message three with no real prominence.
Revenue per send. Typically $2–$10 per replenishment send for consumer consumables, vs $0.10–$0.50 for a marketing broadcast. The order-of-magnitude difference is why programs that ship replenishment usually find it's their largest single lifecycle revenue line within a quarter.
A few practical edge cases worth naming. If a user buys multiple consumables, run replenishment per product, not per user — a coffee buyer who also buys filters should get two parallel flows on different schedules. Cap combined frequency at no more than two replenishment messages per week per user, or you flood your highest-LTV customers and train them to ignore you.
For non-consumables, replenishment works less directly. The closest analogue for durable goods is an "accessory or refill" prompt: camera owners need SD cards, printer buyers need toner. Weaker signal than consumable depletion (timing's less predictable, the correlation's noisier), but the structural play is the same.
Cancellation saves get confused with replenishment all the time, and they shouldn't. The pattern looks similar — "before you cancel" and "we'll miss you, here's how to come back" emails — but the trigger is a cancellation event, not a depletion prediction. They belong in their own sequence, not blended with replenishment.
places replenishment almost always first or second in the build order for any program with repeat-purchase consumable products. It's the rare flow with a clear line from "user behaviour" to "revenue per send" — and the rarer flow where the ROI maths is obvious inside the first month. If you sell consumables and you're not running this, that's the one thing to ship on Monday.
Read to the end
Scroll to the bottom of the guide — we'll tick it on your reading path automatically.
This guide is backed by an Orbit skill
Related guides
Browse allBrowse abandonment: the program that sits between ads and cart
Browse abandonment catches the users who viewed a product and left without adding to cart. Smaller per-user lift than cart abandonment. Ten to twenty times the trigger volume. For most programs it's the biggest revenue lever you haven't shipped yet.
Referral program emails — the three flows that make it work
A referral program lives or dies on the lifecycle messaging — the automated emails that prompt, deliver, and confirm — wrapped around it. Three flows do the work: inviter prompt, invitee welcome, reward confirmation. Get the timing and copy right on each and conversion roughly doubles without anyone touching the offer.
Trial-to-paid: the seven-email sequence that converts 20%+ of free users
Trial conversion is the most financially leveraged flow in SaaS — every percentage point compounds directly against CAC. Here's the seven-email sequence that reliably moves trial conversion from 5% to 20%+.
Price increase emails: how to raise prices without a churn spike
A price increase is one of the highest-risk lifecycle moments your program will ever run. Done wrong, it triggers churn, public complaints, and a reputation dent that outlasts the extra revenue. Done right, most users accept the change without friction. Here's the sequence that works.
Review request emails: the timing that actually produces reviews
A review request at the wrong time gets ignored. At the right time, it converts at 15–25% into a submitted review. Here's the timing, the message pattern, and why incentivising reviews almost always backfires.
Product launch email sequence: the five emails that actually sell a new product
A product launch with one big announcement email captures a fraction of the addressable audience. A proper five-email sequence catches multiple attention windows, builds anticipation, and converts the users who needed a second or third touch. Here's the structure that reliably outperforms the single-send version.
Found this useful? Share it with your team.
Use this in Claude
Run this methodology inside your Claude sessions.
Orbit turns every guide on this site into an executable Claude skill — 63 lifecycle methodologies, 91 MCP tools, native Braze integration. Free for everyone.