Skip to content
Faisal Mahmud TusherDigital marketing, SEO & analytics
with Faisal Tusher

Marketing Objectives › Retention and Repeat Purchase

Retention Through Email

Email as a retention channel: onboarding after the first order, replenishment timing, win-back before lapse, preference centres and cohort holdouts.
Objective guideintermediate9 min readUpdated

Why email suits retention

Retention needs a channel that can speak to one customer about their own history, at a time set by what they did. Email does that better than any other route. It reaches everyone who agreed to hear from the business, it can carry enough content to help someone use what they bought, and its cost barely rises with the number of messages. A lifecycle programme can therefore send the right message to each customer at their own moment, not a single campaign to everyone at once. The lifecycle marketing guide explains why triggers based on behaviour, rather than calendar schedules, are what make that possible.

Permission is the asset. Under the UK regulator's direct marketing guidance, existing customers may sometimes be emailed about similar products under a soft opt-in, provided they were given a chance to refuse when they bought and in every message; other markets have their own rules. Every retention email spends a little of that permission, and Google's sender guidelines require bulk senders to offer one-click unsubscribe and keep spam complaints low. A retention programme that over-sends loses the channel for the customers it most needs to keep.

The limit is honesty about cause. Customers who were already going to buy again tend to open and click, so email dashboards credit the programme with orders it did not create. Attributed revenue measures who opened before buying, not who bought because of the email. The tactics below include the one that answers this question, because without it a retention email budget cannot be defended or cut on evidence.

Email tactics for retention

TacticWhat it does for retentionPlaybook
Onboard after the first purchaseHelps new customers get value from what they bought, which predicts whether they buy againLifecycle marketing
Time replenishment from cohort dataSends the reminder when the typical customer runs out, read from real purchase intervalsCohort analysis
Win back before customers lapseContacts customers whose gap since the last order passes the usual interval, before they are lostRepeat purchase and loyalty
Offer a preference centreLets customers reduce frequency or choose topics instead of unsubscribingEmail list hygiene
Prove lift with cohort holdoutsHolds back a share of each cohort to measure how much the programme changes repeat behaviourOnboarding email cadence test

Onboard the order, not just the account

Most welcome sequences are written for subscribers. A retention onboarding sequence is written for a customer who has just bought something specific: how to set it up, how to get the most from it, what people who bought it usually need next, and where to get help. The second order is the hardest step for most retailers, and it is decided largely by whether the first experience went well. Emails that arrive after delivery and help the customer succeed with the product do more for the second order than a discount code sent the day after checkout. Branch the sequence by what was bought, because a customer who bought a grinder and one who bought beans need different help. The tutorial on building a post-purchase email flow builds that sequence beside the order emails, triggered from delivery rather than the order date.

Send the reminder when they run out, not when the calendar says

Replenishment timing should come from the store's own data. A cohort analysis of time between orders, split by first product, shows when a typical customer re-buys, and the reminder belongs just before that point. The cohort analysis guide covers how to read the intervals. The tutorial on finding what customers buy after their first order builds the split by first product from the order history. Products with no natural cycle need a different trigger: a complementary item, a seasonal occasion or the end of a typical lifespan. One reminder at the right time beats three on a fixed schedule, and stock-up promotions should be used carefully because they push the next order further away.

Contact the customer before they count as lost

Lapse should be defined relative to each customer's usual interval, not a fixed number of days for everyone. Once a customer passes the point at which the chance of another order falls sharply, which the time-between-orders distribution shows, they move into a win-back sequence. Size the effort to the customer's value: a personal note or a meaningful incentive for high-value customers, a short automated sequence for the rest. Lead with something new or useful rather than an announcement of a discount, which trains customers to wait. When a customer has ignored the whole sequence, reduce them to occasional contact rather than mailing them indefinitely. The repeat purchase and loyalty guide covers the design of win-back offers for retailers.

Give customers a quieter option than leaving

For retention, an unsubscribe is a lost channel to a current customer. A preference centre that lets people choose fewer emails, pause for a period or pick topics keeps many of them reachable. Link it beside the unsubscribe link, honour it in every flow, and read the choices: a rise in "fewer emails" is an early warning that the programme is sending too much. The email list hygiene procedure covers keeping the list clean so that the customers who stay subscribed can still be reached.

Hold out part of every cohort

The tactic that makes the rest defensible is a holdout. Exclude a random share of each new customer cohort from the lifecycle programme, or from one flow at a time, and compare repeat rate and lapse between the two groups over the cohort's buying interval. The difference is what email added. Run it as an A/B Test with a control that receives nothing, decided before the results arrive. The onboarding email cadence test shows a test design built on activation and unsubscribe guardrails that adapts to post-purchase flows.

Budget posture and KPI targets

Invest in lifecycle automation and data before campaign volume. The retention email budget is mostly the platform, the customer data work that feeds it (order history, product, interval and consent status in one record) and the content for each lifecycle stage. Send volume is nearly free and should not be how the budget grows; more sends to the same customers raise opt-outs faster than they raise orders.

The ceiling is set by margin on incremental repeat orders measured against holdouts, not by attributed email revenue. Work out the gross margin on the extra orders and renewals the holdout shows, set it against the platform, data and content costs, and the programme is justified while the first exceeds the second. The break-even calculator helps with the arithmetic, and the retention and payback guide explains why that margin also raises lifetime value (LTV) and so the acquisition cost (CAC) the business can afford. Keep a test allowance of time and content for one new flow or one flow rebuild per quarter. Money and effort move out of any flow that shows no lift against its holdout after a full buying interval, and out of any flow whose unsubscribe rate climbs while repeat behaviour does not.

The tactics in a worked plan

Illustrative example — not a real result

Brindlewood, an invented online pet-food shop

Brindlewood sells dry dog food in bags that last most customers about five weeks. Of each month's new customers, 31% order again within ten weeks. The programme is a welcome email and a monthly newsletter; email is credited with £38,000 of revenue a month in the platform's reports.

TacticBeforeAfter two quarters
OnboardingGeneric welcomeThree emails after delivery: feeding guide by dog size, switching tips, how to set a delivery reminder
ReplenishmentNoneReminder at day 30, read from the median reorder interval of 34 days
Win-backNoneSequence at day 50 for customers with no second order; a personal note for those with three or more past orders
Preference centreUnsubscribe only"Reminders only" and "pause for a month" options
HoldoutNone10% of each monthly cohort receives no lifecycle email for ten weeks

After two quarters the contacted groups reorder within ten weeks at 41%, against 33% in the holdouts, so the programme adds about eight points of repeat rate. On cohorts of about 1,200 new customers a month, that is roughly 96 extra second orders a month at a £14 gross margin each, £1,344 against £600 a month of platform and content cost. The newsletter is cut to fortnightly after the "reminders only" option is chosen by 9% of customers in its first month.

The plan shows why the posture leans on holdouts: the platform credits email with far more revenue than the extra orders the holdout reveals, and only the second figure justifies the spend.

KPI set for email retention

KPIWhat it tells youRead how often
Repeat purchase or renewal rate by cohortWhether customers acquired in each period come back within the expected intervalMonthly, once each cohort's interval has passed
Time to second purchaseWhether onboarding and replenishment are pulling the second order forwardMonthly
Lapse rate of emailed customers against a holdoutHow many customers the programme keeps who would otherwise have gonePer cohort, after a full buying interval
Revenue per customer at day ninetyWhat each cohort is worth at a fixed age, comparable across cohortsMonthly
Unsubscribe rate on lifecycle emailsWhether the programme is spending permission faster than it earns ordersWeekly, by flow

Lead the report with the holdout comparison; it is the KPI that shows Incrementality rather than activity. Opens are unreliable because of mail privacy features, as the email metrics and testing guide explains, and clicks show interest rather than retention, so treat both as diagnostics for individual emails. The KPI tree for retention shows how these measures roll up into revenue from existing customers, and the lifecycle email programme playbook sets the monitoring and audit routine that keeps flows firing and segments accurate.

When email is the wrong route for retention

Email cannot retain customers it cannot reach. Businesses that sell through marketplaces or resellers often do not hold the customer's address or consent, and businesses whose customers bought as guests without agreeing to marketing start with a thin list; the first job there is capturing permission, not writing flows. Email is also the wrong lever when the reason customers leave is the product, the service or the price. A well-timed reminder cannot fix a delivery that arrived late or a product that disappointed, and a holdout that shows no lift for every flow is often telling the team that the problem lies outside marketing. Finally, for moments that need an answer within minutes, such as a delivery slot or an expiring booking, the SMS and messaging channel page covers the faster route.

Frequently asked questions

How large should a retention email holdout be?

Large enough that the difference in repeat behaviour between the two groups can be read with confidence over the cohort's normal buying interval, and small enough that the business does not give up much revenue while it learns. For most stores that means a modest share of each new cohort, kept out for a fixed period and then released. Smaller businesses can pool several months of cohorts into one holdout before reading it.

Should retention emails carry discounts?

Sparingly and late. A discount in every lifecycle email teaches customers to wait for one and gives away margin on orders that would have happened at full price. Lead with usefulness: how to get value from what they bought, a reminder timed to when they run out, something new. Keep incentives for win-back, sized to the customer's value, and test whether they add orders against a group that receives the same email without one.

What is the difference between retention email and sales email?

Sales email aims at the next order from anyone on the list and is judged on revenue per send. Retention email aims at keeping customers who have already bought, follows each one's place in the lifecycle, and is judged by cohort on repeat rate and lapse against a holdout. The same platform sends both, but the triggers, timing and measures differ, and a retention programme that drifts into weekly promotions usually raises opt-outs among exactly the customers it was meant to keep.

Terms used on this page

Tools to try

Frameworks

Where this page is used

Sources

  1. Email sender guidelines — Google Workspace Admin Help
  2. Direct marketing guidance — Information Commissioner's Office

Related