Businesses on Stripe took more than $40 billion between Black Friday and Cyber Monday in 2025 — 578 million transactions, peaking above 152,000 a minute. More than 150,000 of those businesses had their best day ever. On Shopify itself, more than 94,900 merchants had their highest-selling day. Somewhere in those numbers are thousands of small companies for whom a single weekend paid for a quarter.

That is the case for taking the week seriously. The case against is that most of the advice is written for retailers, and a founder selling software or a digital product has a different problem: a small list, a recurring-revenue model that discounts can quietly damage, and very few hands when the checkout breaks at midnight. This is how to prepare your startup for Black Friday. It is 21 October; you have five weeks. (If you have eight, start with our eight-week plan; this is the compressed version.)

Stage one: decide whether a discount is your tool at all

Research published by Paddle on subscription businesses found that customers acquired with a discount showed lower willingness to pay, higher price sensitivity and churned at a much higher rate than the core customer base — the figure most often quoted is a lifetime value about 30% lower, though sources disagree about which firm produced it. The direction is consistent across every source that has looked. A 50%-off monthly plan sold in November is, on average, a customer who leaves in February.

So the first decision is honest arithmetic. If your product is a one-off purchase — a template, a course, a licence — a discount costs you margin once and nothing after. If it is a subscription, a discount on the monthly price buys you the wrong customers. The instruments that suit a subscription are different: a discount on the annual plan; an upgrade offer to existing customers; or a bundle that adds something rather than subtracting from the price.

If none of those fits, skip the week. "No Black Friday sale, same fair price all year" is itself a position some audiences reward.

Stage two: build one offer, and protect the people who already pay

Pick a single offer you can describe in one line. The convention for annual plans is two months free — about 16.7% — and the Black Friday version typically stretches that to 25% or 30% for new annual sign-ups.

Lifetime deals are the small founder's temptation, because the cash arrives at once. Treat them as what they are: selling the future at a discount to people who, by definition, never pay again. If you run one, cap it the way the 2025 deal lists show others doing — at 1,000 claims, or even the first twenty buyers — and understand that marketplaces such as AppSumo come with a 60-day refund window and a support obligation that outlasts the weekend.

Then write the email to existing customers before you write the one to strangers. The person who paid full price in September will see your 30% banner. Offer them something — an upgrade at the sale price, a free month, early access to a feature — or expect the cancellation emails to arrive with the orders.

Stage three: the list is the campaign, and it is built in October

Adobe's September forecast expects $95.8 billion to be spent online in the US in October alone, up 8% on last year, before the $47.5 billion it predicts for Cyber Week. Survey figures quoted in trade coverage put the share of shoppers starting before Black Friday at around 71%, and the share starting before November at 46%. Either way, the people who will buy from you on 27 November are deciding now.

For a founder with no ad budget, the list is the whole channel. Put a one-line "Black Friday offer — get it first" capture on your site this week, and tell your existing subscribers the date without the price. Klaviyo's benchmark data shows why the automation matters more than the blast: automated flows are 5.3% of the emails its customers send and 41% of the revenue those emails produce. A welcome sequence sent before the sale outsells a discount code sent cold.

Stage four: get listed where buyers are already looking

The developer and designer communities maintain open, free Black Friday deal lists every year — the Awesome Black Friday and Cyber Monday repository on GitHub is the best known. A pull request costs nothing and puts you in front of buyers who are there precisely to spend. Submit in the first half of November, before the lists are long. And list your deal here, where readers come looking for exactly this.

The newer channel is the shopping assistant. Adobe expects traffic to US retail sites from AI tools to rise 130% this holiday season. The volume is still modest, Adobe says, but the shoppers it does send are unusually decided — 77% of those who have used AI to shop say they feel more confident in the purchase. An assistant can only recommend what it can read, so your pricing page needs the offer, the dates and the conditions in plain text, not in an image.

Stage five: harden the plumbing you will not have time to fix

Shopify's 2025 traffic peaked at 12:01pm Eastern on Black Friday, at $5.1 million of sales a minute. In the week before, run through the whole purchase on a phone — 76% of Shopify's Black Friday traffic was mobile — with the coupon code, from an incognito window, using a real card. Check the receipt email and whatever provisions the account.

Write the three support replies you will send most — "the code did not work", "I bought last week, can I have the discount", "can I switch my monthly to the annual deal" — and decide the answers now, generously. Payment providers held up last year; Stripe reported better than 99.9999% uptime across the weekend. The failures that reach a small team are the ones in their own stack.

And rest before, not after. A founder who ships a new feature on 25 November is testing it in production on 27 November.

Stage six: run the week on a schedule you set in advance

Write every email and post before the week begins, with send times. A sensible cadence for a small list is an announcement on the Monday, a reminder on Black Friday morning, a "last day" on Cyber Monday and nothing else. Sinch measured open rates on Black Friday 2025 at about 16%, excluding bot opens — below a normal week — because every brand sent more. The emails that get read are the ones from senders people already open.

Keep a one-line log: orders, revenue, refunds, support tickets, by day. That record is the plan for next year. Then end on the date you announced. Extending a deadline teaches your list that your deadlines are fiction.

What to do with this

  • This week: decide discount or no discount; put the "get it first" capture on the site; email existing customers the date.

  • By 10 November: offer page live in plain text; pull request to the deal lists; deal submitted here.

  • By 20 November: full checkout tested on a phone with a real card; support replies written; all emails scheduled.

  • 27–30 November: send on schedule, log daily, end on the date. Then sleep.

What this costs. Almost no money: a landing page, an email tool you probably have, and your time. The real cost is margin on the offer and the attention of the following week, when the support arrives.

The honest caveat

Most of the figures here describe retail and large platforms: Stripe's and Shopify's totals, Adobe's forecast, Klaviyo's customers. They prove the week is large; they say nothing about whether your product's buyers shop by the calendar. A developer tool with a hundred paying customers may do better with a quiet October price rise than a November discount, and the churn research cuts hardest for exactly the small subscription businesses this guide is for.

If you prepare your startup for Black Friday at all, prepare it as a list-building exercise with a single, honest offer at the end, run on plumbing you have tested and a schedule you wrote in advance. The founders in Stripe's best-day-ever list did not have more hands than you. They had fewer things left to decide on the day. The dates for this year and the next few are in our Black Friday calendar; what last year actually looked like is in Black Friday 2025 by the numbers.