Black Friday buys you a cohort acquired at a discount, and most of it never returns at full price. The four weeks after the weekend decide whether that cohort was an acquisition or a giveaway, and almost nothing in the standard BFCM playbook covers them.
A note on evidence: this post carries the weakest owned data of the five. Andrew should supply a cohort retention figure from any account, or the post ships one experience marker short of our standard.
Why the discount cohort behaves differently
Someone who bought at 20% off learned something about your pricing. Specifically, that it exists at a lower number and will probably return there.
That is not fatal, and it does change what the next message has to do. A standard welcome sequence written for a full-price customer assumes the buyer chose you on value. A discount buyer may have chosen you on price, and the job of the following month is to give them a second reason.
The practical consequence is that the December sequence should lead with product experience and outcome rather than with another offer. Another discount confirms the lesson they already learned.
The four weeks that matter
| Week | Job | What to send |
|---|---|---|
| 1 | Delivery certainty | Shipping updates, what to expect, when |
| 2 | Product experience | How to use it, what good looks like |
| 3 | Second reason | Range, brand, what else solves their problem |
| 4 | The considered ask | Replenishment or subscription, not a discount |
Week one is not marketing, it is operations, and it matters more than the rest combined during a period when delivery expectations are strained. Shipping and return communication tests carry the highest win rate of any category in DRIP’s experiment database at 41.8%, and that is in normal trading.
The subscription conversion window
For consumables, weeks three and four are the highest-value moment of the whole quarter.
The customer has now used the product, which means the offer is no longer asking them to commit to something unfamiliar. Subscriber lifetime value runs between $350 and $800 against $168 for one-time buyers, and replenishment models sit at the low end of subscription churn at 5% to 8% monthly.
Match the cadence to actual consumption rather than to the pack size. Shipping a 30-day supply on a 30-day cycle means anyone who misses a few days accumulates stock and cancels, which is the most common cause of replenishment churn. We cover the mechanics in our guide to subscription ecommerce.
Ask them why they bought
The confirmation page during peak weekend is the highest-volume research opportunity you will get all year, and almost nobody uses it.
Two questions, open text: what almost stopped you from buying, and how did you first hear about us. A hundred coded responses from peak weekend is more customer research than most brands gather in a quarter, and the themes point directly at the next year’s roadmap. Our guide to post-purchase survey questions covers the question design and the coding process.
Read the support tickets too. Peak volume surfaces friction that normal trading hides, and a week of tickets is a diagnostic document nobody has to commission.
Measure it as a cohort, not as a month
This is where most brands lose the thread. December revenue gets reported as a number, the BFCM cohort disappears into it, and nobody finds out whether the customers acquired that weekend were worth acquiring.
Track the November cohort separately: what share placed a second order within 60 days, within 90 days, and what their average order value was on that second purchase compared to a full-price cohort from October.
Two years of that data and you can answer the only question that matters about your discount depth, which is whether the customers it buys ever come back.
What to do with the cohort that does not return
Accept some of it. A share of Black Friday buyers were always going to be one-time purchasers, and chasing them with escalating discounts trains the rest of your list to wait.
The better use of the data is upstream. If the cohort’s repeat rate is poor, that is information about the offer you ran rather than about the customers, and it should change next year’s discount depth rather than this year’s win-back campaign.
Frequently asked questions
Do Black Friday customers come back?
Some do, at lower rates than full-price cohorts. The four weeks after purchase, and specifically whether you lead with product experience rather than another discount, make a large difference.
When should I make a subscription offer to a Black Friday buyer?
Around weeks three and four, once they have used the product, rather than in the first email. For consumables it is the highest-value moment of the quarter.
Should I discount again in December to bring them back?
Generally no. Another discount confirms that your prices are negotiable and trains the list to wait for the next one.
How do I know whether my Black Friday cohort was worth acquiring?
Track it separately: second-order rate at 60 and 90 days, and second-order value against a full-price cohort from a normal month.
The pre-peak work is in our BFCM CRO checklist. The retention and post-purchase layers are what our Shopify CRO and ecommerce CRO programmes test year-round.