The Second-Purchase Window: When Should a First-Time Customer Buy Again?
A customer who bought ten days ago is not necessarily a retention problem.
A customer who bought 180 days ago and never came back might be.
The important part is everything between those two points.
At Altimax, we call that the Second-Purchase Window:
The period in which a first-time customer would reasonably be expected to place their second order.
Knowing that window changes how you treat the customer.
Before it:
help them succeed.
Around it:
make the next purchase easy.
After it:
begin asking why the relationship did not progress.
Why the window matters
BS&Co’s 2026 analysis found that among customers who did make a second purchase:
- 50.3% returned within 30 days,
- 76.4% within 90 days. (bsandco.us)
The exact shape will vary by category.
But the important implication is that the second-order opportunity is often concentrated much earlier than brands assume.
A retention program that says:
“Thank you on day 2. Review request on day 10. Win-back on day 120.”
may leave a large behavioural gap in the middle.
The Second-Purchase Window is not one number
Consider:
Coffee
A 30-day replenishment cycle may be reasonable.
90-day supplement bundle
A much longer period.
Fashion
The next purchase may be occasion- or newness-driven rather than depletion-driven.
Furniture
Order number two may be another room or category rather than replenishment.
So the window is part:
- customer behaviour,
- product behaviour,
- category behaviour.
Use the repeat-purchase curve
Imagine 1,000 customers who eventually return.
By day 15:
180 have made order two.
By day 30:
By day 60:
By day 90:
By day 180:
Plot that over time.
Now you can see:
- when the curve rises quickly,
- where it bends,
- when the probability begins to flatten.
That curve is much more useful than:
average time to second purchase = 61 days.
Median is often more useful than average
BS&Co notes that average second-purchase timing can be distorted by a long tail of late buyers, while median timing can better represent the point where half of eventual repeat buyers have already returned. (bsandco.us)
RetentionX similarly exposes both average and median time-between-order measures. (help.retentionx.com)
If a few customers return after 300 days, they can make the average look much later than the experience of most repeat buyers.
Product-level timing can change the window
Klaviyo’s product analysis is built around measuring the time from a product purchase to the next purchase and using that information to time post-purchase communication. (academy.klaviyo.com)
This matters because Product A can have a 20-day second-purchase pattern while Product B has a 75-day pattern.
One brand.
Two windows.
The window has three stages
I think of it like this.
Stage 1: Experience
The customer is still using, receiving or evaluating the first purchase.
Focus on:
- onboarding,
- education,
- confidence,
- service.
Stage 2: Progression
The customer is approaching the point where a second purchase becomes plausible.
Now introduce:
- reorder,
- replenishment,
- next product,
- subscription,
- relevant recommendation.
Stage 3: Overdue
The customer has moved meaningfully beyond expected order-two timing.
Now the question changes:
Why did they not continue?
This is where Revenue Recapture begins.
A worked example
Suppose:
10,000 customers bought Product A.
2,400 eventually make order number two.
Their second purchases occurred:
- 800 by day 30
- 1,600 by day 60
- 2,050 by day 90
- 2,300 by day 150
- 2,400 thereafter
Median second purchase:
roughly day 48.
Now consider:
Customer A at day 18.
Still early.
Customer B at day 50.
Inside the natural purchase period.
Customer C at day 110.
Now meaningfully outside the strongest part of the curve.
The same one-time buyer status means something different at each point.
Why this improves discounting
If the majority of customers naturally buy by day 60, giving everyone a discount on day 30 may subsidize orders that would happen anyway.
A more disciplined strategy might:
- educate early,
- recommend naturally,
- reserve stronger incentives for customers who have moved beyond normal timing.
NICKIS reportedly took this approach by delaying more aggressive incentives until customers had moved materially beyond expected second-purchase timing. (retentionx.com)
The Second-Purchase Window should change by acquisition cohort
A customer acquired through:
- full-price search,
- 50% Black Friday promotion,
- influencer launch,
- referral
may have very different repeat behaviour.
So calculate timing by cohort where sample size permits.
If Black Friday customers take longer and repeat less, that is acquisition intelligence.
Use the window to evaluate flows
Ask:
Does our post-purchase flow cover the period where customers actually make order number two?
If most repeat happens between day 20 and day 70 but the flow ends at day 10, the answer is no.
BS&Co specifically argues that many post-purchase flows end before the majority of the repeat-purchase window has played out. (bsandco.us)
The Second-Purchase Window is not the same as the Lapse Window
The Second-Purchase Window applies specifically to:
order one → order two.
The Lapse Window can apply later:
repeat customer → unexpected absence.
A first-time buyer has no individual cadence yet.
A repeat customer does.
That is why the two deserve separate logic.
The management question
Do not ask only:
What percentage of customers repeat?
Ask:
When should the second purchase happen, and what percentage of customers have failed to make it by that point?
That tells you much more about what needs attention now.
Related reading
Why Customers Don’t Make a Second Purchase
How Long Should It Take Customers to Buy Again?