Black Friday E-commerce Prep: What Gets Decided Before the Checklists

Short answer. Black Friday 2026 falls on Friday, November 27, with Cyber Monday on Monday, November 30. Before the first send checklist, four decisions determine what your CRM will actually be able to do: one readable offer instead of a pile of promo codes, a reinforcement mechanic prepared while nobody is under pressure so you can adjust without destroying margin, an October versus November trade-off that moves list building and testing to where they cost least, and a segmentation that separates volume objectives from margin objectives. Those four decisions belong to August or September. The operational checklists come afterwards, and their value depends on them.

This guide draws on three public pieces from the MS4D group, of which Deliver is the CRM brand: two webinars hosted by Maxence Vanderswalmen, one on cross-functional preparation and one on offers and conversion mechanics, plus a round-up of advice from the acquisition, CRM, and SEO leads. These are practitioner trade-offs from live accounts, not a theoretical framework.

The operational detail already exists elsewhere on this site: the 7 Klaviyo flows that must be running before the peak, the 9 BFCM segments to build, and the 60-days-out to Day+7 calendar. This page covers what gets decided upstream.

1. Black Friday does not run like a seasonal sale

Plenty of teams treat the period as one more commercial moment, modelled on end-of-season sales. The distinction drawn in the first webinar is more useful: a seasonal sale gives the whole market the same starting line at the same hour. Black Friday works on a first-come, first-served basis. Someone who has just bought a desk from one merchant will not buy a second one from the shop next door because it is showing ten more points of discount.

Three direct consequences for a CRM programme.

The delay between your reminders becomes a conversion lever. During the window, shortening the interval between messages in an abandonment sequence is not gratuitous aggression: it is the only way to reach a contact before their budget is spent elsewhere. Outside the period, that interval is justified by recipient comfort. Inside it, the justification is competitive.

Early access is a timing trade-off, not a symbolic reward. Opening to your best customers ahead of the general base is widespread, but the question raised in the webinar is cannibalisation: how do you duplicate a private-sale sequence without stripping the Black Friday sequence itself, given that your base has a finite amount of attention to give. The answer depends on your addressable volume and on what you know about your own vertical.

The length of the operation is a choice, not a default. The pattern the team observes most often is two weeks, with the preceding weekend as early access, sometimes three. Some highly competitive verticals run a full month because certain players start very early. The stated decision criterion is not catalogue size but whether the team can actually sustain the cadence for the length announced.

To that you have to add a constraint few plans anticipate. Black Friday creates an abnormal structural overload in the life of a company, on teams that are already stretched, with heavy leadership pressure on Q4. Product, brand, IT, acquisition, CRM, and logistics have to move together when they spend the rest of the year working in silos. CRM is one of the first functions to pay for it: an offer that shifts on the commerce side without reaching the flows, and you send a reminder showing an obsolete promotion at the worst moment of the year.

2. One offer, readable, repeated everywhere

Both webinars start from the same value equation, borrowed from Alex Hormozi. Perceived value rises with the outcome the customer wants and with the odds they give you of delivering it. It falls with delay and with the effort demanded of them.

What makes that frame useful for Black Friday is that it moves the discussion off the discount percentage alone. A dream kitchen delivered in twelve months when the competition delivers in three is worth less to the customer whatever the discount on the sticker. A kitchen they have to assemble is worth less than one installed turnkey. Two of the four levers are therefore reassurance elements, and they are exactly the ones CRM can carry: guaranteed delivery dates, extended returns, installation included, instalment payment. The first webinar explicitly recommends pushing these differentiating bonuses forward as the period progresses.

Then comes the most repeated rule of both sessions: one single offer. A customer landing on the site should have one direction to take. A subscriber opening an email should read one message and immediately understand the benefit. That single offer is then repeated everywhere: in the navigation, on the first product image, next to the add-to-cart button, in the cart itself, in signup popups, and in every email. The phrasing used in the webinar: people forget, and teaching is repetition.

The anti-pattern is lived, not hypothetical. The founder describes running, as marketing director of a furniture brand, a period where two offers coexisted: free shipping site-wide, or a discount on a product selection, both required to appear on every banner and every email. The result was an unreadable set-up, the exact opposite of the friction reduction the period calls for.

There is still a clean way to stack. The value-stacking logic says a top-up offer should complement the base offer rather than compete with it, and should be expressed in the same unit of measure. A percentage discount is reinforced by an additional percentage, never by a euro amount per spending tier. Two reasons given: mixing units makes the offer hard to read for the customer, and it makes your own average discount rate impossible to track during the operation.

3. Reinforcements: the ammunition is prepared cold

The reinforcement mechanic comes from retail and mail order, and it is probably the most transferable idea in these webinars for a CRM team.

The principle: you set a base offer, for example a discount across a large share of the catalogue, and you prepare one or more top-up offers in advance that you will only trigger if a threshold is crossed. Under target on volume while margin holds, you fire the reinforcement. Over target, you keep it in reserve. The trigger can be site-wide or restricted to specific segments by email.

What this demands of CRM, entirely upstream: the target segments exist and have been tested, the promo codes are created, the templates are dressed, the exclusion rules are written. The reinforcement then becomes a fifteen-minute decision rather than a project.

The benefit is as financial as it is operational. The warning in the webinar is blunt: in the Thursday-evening rush, when a team realises it is behind and improvises an extra discount, nobody recalculates the discount rate for the operation, and margin absorbs the decision. Pressure caps and last-minute arbitration are covered in the 7-days-out plan.

4. Sow in October, harvest in November

This is the cross-channel trade-off that changes the CRM result the most, and it is not made inside the CRM.

Media costs are structurally lower in October than during Black Friday itself, when everybody is fighting over the same inventory. The conclusion the acquisition team draws is that an engagement budget spent in November is badly placed: at equal budget, October buys more.

Three uses for that window.

List building. Campaigns whose objective is list signup, run in October, feed an audience that will have time to go through a complete welcome flow before the commercial window. A contact captured on November 25 has time for nothing. The second webinar also recommends rebranding signup popups and welcome sliders to the upcoming operation from that phase onward, so the promise is coherent end to end. The lever is detailed in our guide to popups and signup forms.

The retargeting pool. The same October budgets build an audience of site visitors and video viewers you can retarget when conversion is at its strongest. This phase advance gives you, in the phrase used, more ammunition than your competitors at the moment that counts. It is the same logic as first-party data collection.

Creative testing. Creative iteration inside a window as short as Black Friday is difficult, expensive, and hard on teams. The recommendation is therefore to test the boldest concepts in October, review performance in early November, then launch only pre-validated concepts during the period. Inside the window, you prune what is not performing and iterate on a narrow sample, not across the whole creative matrix.

The same anticipated-learning logic applies on the paid search side. The products you plan to discount are not necessarily your hero products the rest of the year, and the algorithm does not know how to serve them. Hence the recommendation of a dedicated campaign for that selection, launched at least two weeks ahead, ideally a month, the delay considered optimal for clean learning.

On the CRM side, the exact equivalent of that anticipated learning has another name: deliverability.

5. Deliverability is tested in October, on three layers

The argument fits in one sentence repeated in two of the three sources: deliverability does not get repaired in 24 or 48 hours, and depending on how bad the situation is, getting back to normal can take several weeks. A test run on November 20 therefore opens no options, it merely documents the problem.

The recommendation is to test from early October, on three complementary layers.

The insistent point in the practitioner round-up: this is not about picking one of the three, it is about doing all three, because each tests something different. And if a problem is detected late, the consequence is immediately a CRM one: you have to segment differently, pull back on the least engaged segments, and give up part of the planned volume. The full set of settings is covered in our deliverability guide.

6. Segment for margin as much as for volume

The audit observation is clear: the most common segmentation on the accounts observed is still openers and clickers at 30, 60, or 90 days. It is described as a fairly limited baseline for this specific period, because it rests on non-purchase behaviour. The recommendation is to cross it with an RFM model: you have customers with excellent RFM who never react to email, and highly email-responsive customers whose purchase value is low. Crossing the two axes creates enough boxes to build genuinely different sequences.

Then comes the segment BFCM guides almost always ignore: bargain hunters, the customers who only buy on promotion. Two identification methods are given. By purchase timing first, checking whether all of a profile's orders fall during commercial operations. By custom property second, populated during the operation itself, set to 1 if the purchase was made under promotion and 0 otherwise.

Usage matters more than method. The phrasing used: with bargain hunters you are in an exclusion logic more than an inclusion logic. The decision depends on the objective set for the period. Revenue objective, you target them to go get volume. Margin objective, you exclude them, because a customer who only returns at degraded margin has no economic interest. That decision is taken at the level of the operation's objective, not at the level of the segment.

The corollary assumed in both webinars is profitability on the first order. One case reported during the mechanics session: a brand finished its operation close to 200,000 euros in the red, having overspent on acquisition while betting on the future lifetime value of customers recruited at a slashed price. The argument against that bet: uncertainty on the lifetime value of these customers is too high, they have a particular status, and they should not be treated like customers acquired in a normal period. A profile who buys at Black Friday, then at winter sales, then nothing until June private sales is a weight, not an asset.

The CRM treatment that follows is simple: record the information during the operation rather than reconstructing it in January, then let the post-purchase flow sort out who buys again at full price in the following weeks.

Conversely, the proof that preparation pays is in these sources too: three accounts managed directly by one consultant on the team passed one million euros in revenue in under 24 hours during a Black Friday. Segmentation is not incidental to that.

7. Stacking, cadence, and channels: the trade-offs made in September

Stacking with the welcome offer. Your popup promises a discount on signup. Is it combinable with the Black Friday offer? The question has to be settled before, not during. If stacking is possible, the instruction is to make a lot of noise about it. If it is not, write it clearly, because the disappointment otherwise gets paid in unsubscribes at the worst possible moment.

Overlap between flows and campaigns. Check your tool's anti-collision rules so an abandonment reminder and a newsletter do not go out minutes apart to the same profile. Promotions embedded in evergreen flows, sunset or winback sequences, need reviewing if they conflict with the public offer. The structural distinction is covered in flows versus campaigns.

The send map. Lay out, segment by segment, which message goes out when, with reinforcements marked as optional and tied to a threshold. That document is what lets you adjust live without improvising, and what stops you waking up the day before behind target.

Channels beyond email. SMS, WhatsApp, and push notifications cost more per message sent but deliver higher read rates. In a saturated inbox, the question is no longer only what to say but where to say it. Our omnichannel approach and our SMS marketing guide cover the conditions to respect.

8. Governing the week: decision latency costs more than mistakes

The question put to the first webinar's audience is concrete. What happens if, at 5pm on Friday, all your campaigns stop because the budget is spent?

Two possible answers, both to be validated before the period. A fast decision channel, called a hotline or a skip-level principle, that gets a budget increase or an extra send approved in minutes without climbing three levels of hierarchy. Or autonomy thresholds granted to teams in advance: if the nominal ceiling is 10,000, the person knows they can go to 12,000 without asking.

The line that justifies the arrangement: there is nothing worse during this period than the cost of the missed opportunity. There are 365 days in a year, but only one Black Friday. Realising on Saturday morning that you could have doubled volume the day before is not recoverable.

On the CRM side, the same rule applies to reinforcements: who is allowed to trigger, on what quantified criterion, within what delay. And this is exactly what preparation buys. Responsiveness and composure during the period depend on the work done before, because that is what leaves mental bandwidth available for the unexpected, and there will be some.

9. Do not debut anything inside the window

The second webinar is categorical: Black Friday is not a testing period. The case reported is a fashion brand that tried to launch a complex golden-ticket gift system never tried before. The operation did not go badly, but it underperformed compared with other periods of the year, for a heavy implementation cost.

The opposite approach works better: take a mechanic that has already been validated, adjust the percentages, change the dressing and the name. An evergreen offer rebranded in the colours of the period works because it gives a reason to buy now to purchase behaviour that already exists. Nobody is trying to create that behaviour, people are already actively looking for deals. You adapt to them.

That leaves the question of where to test instead. Earlier commercial windows are exactly for that: French Days, private sales, brand anniversary, a themed day in your sector. The case cited is a client who validated a progressive bundle mechanic during French Days, then reused it at Black Friday with different percentages and a different presentation. The mechanic itself was already proven.

Is your CRM set-up ready to absorb these trade-offs? A 30-minute emailing audit is enough to spot what would break at peak.

FAQ

When should you start preparing Black Friday for an e-commerce brand?

The structuring decisions belong to August or September: format and length of the operation, the single offer, planned reinforcements, whether the objective is expressed in revenue or in margin, and the governance rules. October is the upstream execution window: deliverability testing, list-building campaigns, creative testing, and algorithm learning on the paid side. November is for executing what was prepared, not for deciding.

Should you target bargain hunters during Black Friday?

It depends on the objective set for the period. With a revenue objective, these profiles bring volume and including them is coherent. With a margin objective, excluding them is coherent, since they only come back at degraded margin. The logic described in the webinar is an exclusion logic more than an inclusion one, and it assumes you flagged those profiles in advance.

How long should a Black Friday operation last?

The pattern the team observes most often is two weeks, with the preceding weekend as early access, sometimes three. Some highly competitive verticals go up to a full month because certain players start very early. The decision criterion is not catalogue size but whether your teams can sustain the cadence for the full length announced.

Can you launch a brand new promotional mechanic for Black Friday?

It is not recommended. An untried mechanic demands heavy implementation, stretches teams, and has no performance history on your audience. The advised approach is to take an offer that has already been tested, adjust the percentages, and rebrand it in the colours of the period. Tests belong on earlier commercial windows: French Days, private sales, or an event specific to the brand.

What should you do with customers acquired during Black Friday?

Do not treat them like customers acquired in a normal period. Flag during the operation whether the purchase was made under promotion, then let a dedicated post-purchase flow sort out who buys again at full price in the following weeks. The opposite bet, accepting a loss on the first order while counting on future lifetime value, is described in these sources as too uncertain, with one reported case of close to 200,000 euros lost on a single operation.

Sources

The three pieces this guide draws on are public on the MS4D group's YouTube channel, in French:

• E-commerce: turning your Black Friday into a ROI machine, https://www.youtube.com/watch?v=453tQ709NsM
• How the best brands generate millions for Black Friday, https://www.youtube.com/watch?v=_KX-dZEcHlQ
• The underrated advice of four specialists, SEA, Social Ads, CRM and SEO, https://www.youtube.com/watch?v=RLQEjJJsOWU

→ Have your CRM set-up audited before the November peak

Further reading

• The 7 Klaviyo flows to fix before Black Friday
• BFCM: 9 Klaviyo segments to prepare
• Black Friday email checklist: 60 days out to Day+7
• E-commerce CRM strategy: the full framework
• France e-commerce marketing calendar 2026

Charlotte Rodrigues, CRM Lead at Deliver.