September is when the decision is really made, not on November 27, because by then the discount table is already printed and the systems are already loaded. The decision is closed now, and it depends on a piece of data that almost no retailer has at hand when they need it: which customer segment will receive what, on which channel and with which message.
That is the real problem behind any customer segmentation for Black Friday in retail: marketing judgment is not what's missing, the way to build the segment on time is. The most common way to solve it is a data cross-check requested from IT weeks in advance, which almost never arrives before the flyer goes to print.
This practical guide is part of Wapping's Retail Year-End Plan 2026-2027, which you can download for free: it walks through what needs to be sorted out before October 16 so that your Black Friday segmentation doesn't depend on a last-minute ticket, and closes with a way to check where your own operation stands today.
Why customer segmentation for Black Friday in retail always ends up as a ticket to IT
The pattern repeats at most retailers. Marketing wants to launch a private sale or a campaign aimed at a specific group of customers. To build that group, someone has to cross data that lives in different systems: purchase history in the POS, the ecommerce registry, the CRM.
That cross-check isn't done by marketing, it's done by IT, by hand, and it takes as long as it takes. By the time the segment arrives, one of two things has usually happened: the date has passed, or the segment is a snapshot of the day it was built, which doesn't update even if the customer keeps buying.
The result is predictable: the campaign goes out with the same message for everyone, because building something more refined involves several departments, data extractions and tests across different systems, and there is almost never time.
Marketing without depending on IT is not an aspiration, it is the condition for Black Friday segmentation to arrive on time.
Before segmenting, solve identity
Here is the step that almost every campaign calendar skips: a segment is only as good as the history it is built on, and that history doesn't exist if the customer isn't identified.
Between 60 and 70% of transactions in physical stores are anonymous. Without identity there is no history, and without history you can't know whether a product family is being taken by repeat customers or by people who were just passing the shop window.
The fourth piece of data missing from most discount tables
Any retailer decides the discount for each family with three pieces of data: how it rotates, what margin it leaves and what the competition is doing. Those are three product data points, and one is missing: who buys each family, and whether they are a returning customer or an occasional one.
Without unified identity across store and ecommerce, that fourth piece of data cannot be calculated. A unique customer identity is the foundation everything else rests on: without an ID that survives the jump between channels, any segment built on top inherits the same problem.
The Black Friday segments that really move margin
With the fourth piece of data solved, segments appear that the turnover report cannot see on its own. An illustrative example helps to see it, with sample numbers, not from a real customer:
Two product families can look like twins in the report: similar units per month, similar margin. But if you cross those sales with customer history, it turns out that seven out of ten units of one are taken by customers with three or more purchases a year, and that eight out of ten of the other are taken by one-visit buyers. Same number in the turnover report, opposite discount decision.
This is exactly what a data-driven campaign in retail makes possible: not treating two families alike when they only look similar on paper, and adjusting the discount for each one according to who really buys it, not just how it moves in the warehouse.
The same pattern changes name depending on the sector, but the criterion is the same. In cosmetics, the serum that regular customers replenish every two months doesn't need the same discount as the gift set that only moves during campaigns. In both cases, the turnover report presents them the same way, and only customer history tells which one sells on its own and which one lives off the poster.
RFM segments for email marketing work with the same logic applied to people instead of product families: who buys, how often, how long ago and for how much, to decide who gets which email and with which offer.
How to build those segments without opening a ticket to IT
Dynamic customer segmentation in retail changes the whole equation: instead of asking for a one-off cross-check that expires the day it is delivered, the marketing team configures the segment once and the system keeps it up to date on its own, with customers entering and leaving according to their behavior in real time.
This relies on more than 50 segmentation parameters that marketing can operate directly, with no code and without opening a ticket to IT. The practical difference is this:
- Before: marketing asks for the cross-check, IT builds it by hand from several systems and the segment arrives as a snapshot of that day.
- With dynamic segmentation: marketing configures the criterion (frequency, value, product family purchased, channel), the system applies it on its own and keeps it alive throughout the whole campaign.
Real ecommerce campaign personalization without IT means this: the team that designs the campaign is the same one that runs it, with no technical intermediaries in the middle of the process.
The two levers that depend on having the segment ready before November
With the segment solved, two levers open up that don't work without it.
The private sale. A window of days before Black Friday, by invitation, for identified customers, designed above all to place the seasonal stock that would otherwise reach the sales with a bigger discount. If your program has tiers, the invitation can be staggered by member tier: that way it stops being an arbitrary privilege and becomes a status benefit. In an illustrative example with reasonable assumptions, moving 1,000 units of seasonal leftovers in a private sale at a moderate discount leaves more total margin than leaving them for Black Friday or for the January sales, with the same merchandise. This can only be executed if there is a list of identified customers with their purchase category, not a generic list.
The benefit returned afterwards. On the big day, the checkout works the same for everyone, with a single price. What changes is what the identified customer receives afterwards: a coupon calculated on their purchase, visible on their phone, with an expiry date, designed to bring them back in December. That calculation depends on knowing who has bought what, at the moment of payment.
Neither lever requires charging different prices to different people on the same day. Both depend on having the same requirement solved before November: identity and dynamic segmentation.
What to do before October 16
October 16 is the date that marks the point of no return: it is when you need to start accumulating identified customer history in order to build, in time, the segments that will be used in the Black Friday window.
Businesses that already work with unified identity and dynamic segmentation record, according to customer results gathered by Wapping, 40% more purchase frequency, 20% more average ticket and 30% more customer retention. All three figures depend on the same thing: knowing who buys, at the moment they buy.
We have turned everything that needs to be in place into fourteen concrete questions, grouped into identity, data for deciding, campaign and benefit, and relationship and measurement. They are not features: they are things that happen or don't happen every time a customer pays in your store or on your website.
You can go through them with your CRM manager, or complete the online self-assessment in five minutes and receive a report with your starting point and the order in which it makes sense to solve each thing before October 16.
