Post Black Friday analysis: what to compare with last year and how to read the numbers

Aligned periods, one source of truth, traffic, conversion and order value, margin after discounts, ads and returns: how to run an honest analysis after Black Friday 2026.

Vasile Dingă2026-11-16T11:50:09.821394+00:00
Post Black Friday analysis: what to compare with last year and how to read the numbers

In the week after Black Friday, the first report I see at almost every online store is a single number: sales went up or down against last year. It's a useful number, but it says very little. This year Black Friday fell on Friday 27 November 2026 and last year on 28 November 2025, so the calendar moved as well. Below I show what I compare after the campaign, how I line up the periods, how I split revenue into traffic, conversion and order value, how I get to the real profit after discounts, ads and returns, and how I write the conclusions so they're useful next year. At the end there's a comparator where you enter your own figures and get the interpretation.

1. Pick the periods carefully: the calendar moved

Black Friday always falls on the fourth Friday of November, so the date changes from year to year. In 2025, Black Friday was on 28 November and Cyber Monday on 1 December. In 2026, Black Friday was on 27 November and Cyber Monday on 30 November. That has two practical consequences:

  • Don't compare November with November. November 2026 contains Cyber Monday, November 2025 doesn't. The monthly report shows growth that comes from the calendar alone.
  • Compare days by their position relative to Black Friday, not by date: Friday with Friday, weekend with weekend, Monday with Monday. I use Friday to Monday and, separately, the 14 days before and the 14 days after.

Watch out in Google Analytics 4: the previous year comparison in the date picker lines up the same calendar dates, not the same weekdays. In Google's documentation I saw that the "previous year, match day of week" option has been removed. So for Black Friday I pull the two periods separately and compare them myself.

2. Take the figures from one source, with one definition

For orders and revenue, my main source is the store platform, not the Google Ads or Meta reports. Each ad platform attributes orders by its own rules, and the same order can appear in both. I use GA4 for traffic and conversion rate, and the ad platforms only for cost.

Make sure both years are calculated the same way: revenue with or without VAT, with or without shipping, cancelled orders removed or not. If you changed tracking or consent settings in the meantime, GA4 traffic may not be comparable with last year. Note that next to the figures rather than hiding it.

3. The figures I compare

  • Revenue and orders, from the store platform.
  • Sessions and conversion rate: how many people came and what share bought.
  • Average order value: revenue divided by orders.
  • New and returning customers: what share of orders came from people who hadn't bought from you before.
  • Margin after discounts: revenue minus the cost of the goods sold.
  • Ad spend, in total and per order.
  • Returns, as a share of revenue.
  • Contribution after ads and returns: the money you actually kept.

An example with made-up figures, for the Friday to Monday period:

Metric Sessions Last year 42,000 This year 51,000 Change +21.4% Metric Conversion rate Last year 2.00% This year 1.82% Change -0.18 pp Metric Average order value Last year €50 This year €52 Change +4.0% Metric Revenue excl. VAT Last year €42,000 This year €48,360 Change +15.1% Metric Ad cost per order Last year €9.05 This year €11.18 Change +23.6% Metric Contribution after ads and returns Last year €9,328 This year €7,924 Change -15.1%

Revenue grew by 15%, and the money kept fell by 15%. That's exactly the situation a single sales number hides.

4. Break revenue down: traffic, conversion, order value

Revenue is the product of three things: sessions times conversion rate times average order value. Once you know which of them moved, you know where to look for the cause:

  • More traffic usually comes from ad budget, the email list or mentions. The question is what it cost.
  • Lower conversion on higher traffic often means colder traffic or a weaker offer than competitors had on the same days. Sometimes it means problems on the site: check speed and errors on the peak days.
  • Higher order value comes from prices, bundles or the free shipping threshold. If it rose only because you sold pricier products at a smaller discount, it isn't necessarily a volume win.

5. Margin after discounts, ads and returns

This is where the conclusion usually changes. I start from revenue excluding VAT, take off returns, take off the cost of the goods customers kept and take off ad spend. What's left is the campaign's contribution, before fixed costs. I explained why I look at it and not only at ROAS in the article on contribution margin vs ROAS.

Returns have a timing trap. For online purchases in the EU, consumers have 14 days from delivery to withdraw from the contract, and Black Friday parcels often arrive only the following week. An analysis done on 2 December doesn't have the real returns yet. I do a first version straight away, using last year's return rate, and a final version in mid December.

6. New customers, or customers who would have bought anyway

A campaign that sells mostly to existing customers can look great in reports and still bring nothing new: some of those people would have bought in December anyway, at full price. I count new customers by order email in the store platform, not in GA4, which counts devices and browsers, not people.

The share of new customers tells you how much you invested in the future. Its real value only shows after a few months, once you know how many ordered a second time. I wrote about that in Black Week customers and the second order.

7. What changed apart from the numbers

Before drawing conclusions, I write down next to the table everything that was different between the years. My short list:

  • How big the discounts were and on which products.
  • Whether the campaign started earlier. A campaign that starts a week before moves sales from the main weekend into the days before.
  • Which products sold out, and when.
  • Ad budget and new channels.
  • Changes to the site, tracking or shipping prices.

I also look at the surrounding weeks. Sales usually dip before Black Friday, because people wait for the deals, and dip again afterwards. If the weekend grew but you lost the same amount in the weeks around it, you moved sales, you didn't create them. I explained what happens to results after the peak in why ROAS drops after Black Friday.

8. Write the conclusions as decisions for next year

A good analysis ends with three or four sentences of the "next year we'll do this differently" kind. For example: stop ads on low margin products once they hit the cost per order ceiling, prepare double stock for the top five products, start the email list two weeks earlier. Keep the table and the notes in one document so it's at hand next October. The GA4 reports I use for this are described in GA4 for online stores: key reports.

How I work

For my clients I run the analysis in two stages: a first version in the week after Cyber Monday, with estimated returns, and a final one in mid December. I take orders and costs from the store, traffic from GA4 and spend from the ad platforms, line up the days around Black Friday and finish with a short list of decisions for the following year.

Compare Black Friday with last year

Enter the figures for the same period in both years; the comparator shows the changes, what moved revenue and what happened to the money you kept.

Frequently asked questions

Conclusion

An honest post Black Friday analysis compares aligned days, takes figures from one source, breaks revenue into traffic, conversion and order value, and ends at contribution after ads and returns, not at sales. If you'd like to do the analysis and reporting for your store together, see the Data Analytics page or write to me via contact.