Your Q4 is decided in August: the 2026 golden quarter planning timeline

That sounds like plenty. It isn't. Take out August holidays, the two weeks in September everyone spends catching up from August holidays, and the fact that your creative needs signing off six weeks before it runs, and the working time available to actually build a Q4 plan is closer to eight weeks.

Every year we watch the same pattern. Brands that had a good golden quarter did the boring work in August. Brands that had a stressful one started in October, spent November firefighting, and finished the year unable to explain what actually worked.

Here's the timeline we run, and what belongs in each month.

The 2026 dates that anchor everything

Two quirks are worth planning around this year.

Christmas Day falls on a Friday, so the final full shopping weekend is 19 and 20 December and your last posting dates land midweek. That compresses the panic-buying window that usually rescues underperforming December campaigns.

And Boxing Day is a Saturday running into a bank holiday Monday, which gives the post-Christmas sale a proper three-day run. If you sell anything giftable or discountable, that weekend deserves a real budget line rather than whatever is left over.

What last year actually told us

UK shoppers spent £3.8 billion online across the four days from Black Friday to Cyber Monday in 2025, up 4.6% year on year, on an average discount of 16.7% across all retail categories (Adobe Analytics).

But the more useful number is the shape of it. Worldline recorded transaction volumes up 32% on Black Friday itself and 39% on Cyber Monday, even though most retailers had been discounting for weeks beforehand. IMRG also found only 46% of consumers planned to take part, down from 53% the year before (IMRG).

Read those together and you get the current state of UK peak trading. Fewer people are participating, the promotional window keeps stretching earlier, and yet purchases still concentrate hard on the two named days. Running a six-week sale doesn't move demand forward as much as retailers hope. It mostly discounts the people who were going to buy on the day anyway.

2026 forecasts aren't published yet, so plan against last year's shape rather than a number nobody has produced.

August: build the foundations nobody notices

August work is invisible if you do it right and catastrophic if you skip it.

Fix the feed. For anyone running Shopping or Performance Max, the product feed is the campaign. Audit it now: missing GTINs, weak titles, wrong availability, blank product types, images that fail on mobile. A feed fixed in August has three months to earn its performance history. A feed fixed in November is a feed you're testing during peak.

Rebuild your audiences. Remarketing lists, customer match files, exclusion lists. Anyone who bought last Christmas is a different prospect to anyone who has never bought at all, and if you're still running the same audience structure you built in March, you're paying to reach people you already own.

Get measurement straight while the stakes are low. Conversion tracking, enhanced conversions, consent mode, server-side tagging if you're using it. Check that your platform numbers and your back-end numbers agree in August, when a discrepancy is an interesting question. In November it's an argument.

Write the creative brief, not the creative. You need to know what you're saying before you book the shoot. What's the offer? What's the hook? Which formats, which placements, how many variants for testing? Brief in August, produce in September, and you'll have assets in hand before the platforms get congested.

Model the budget three ways. Flat, plus 20%, plus 50%. Decide now what you'd spend more on if things go well, because the decision you make in a good week in November is always worse than the one you made calmly in August.

September: the platform deadline nobody has diarised

This is the one to pay attention to this year.

From September 2026, Google Ads is auto-upgrading automatically created assets and campaign-level broad match settings to AI Max, and you will no longer be able to create new Dynamic Search Ads campaigns in the interface, Editor or API (Google).

The full DSA migration was pushed back to February 2027 specifically to keep it out of Q4, which was a sensible call by Google and has also given a lot of advertisers the false impression that nothing happens until next year. It does. The September changes land right in the middle of your Q4 build.

So in September: audit which campaigns are affected, migrate deliberately rather than being migrated, and give the new setup six to eight weeks of learning before November. Note that AI Max has also started appearing in Standard Shopping campaigns, so retail accounts should expect the same direction of travel there.

The rest of September:

• Produce creative against the August brief
• Launch upper-funnel and prospecting activity so November's remarketing pools aren't empty
• Run your tests now, on a test budget, while a bad week costs you very little
• Sense-check landing pages and site speed under load

October: the month everyone underestimates

October is when the plan meets reality, and the single biggest mistake is treating it as a quiet run-up.

Get everything live and stable by mid-October. Campaigns built, budgets loaded, creative approved and through policy review, promotions configured in the platforms rather than pencilled in. Ad disapprovals on promotional creative are a well-worn way to lose the first day of a sale, and reviews take longer when every advertiser in the country is submitting at once.

Late October is also when the sale actually starts for a growing number of UK retailers. If your competitors go early and you don't, decide that deliberately rather than discovering it on 25 October.

One thing worth being honest about: if a campaign type is new to your account, October is your last chance to learn it. Anything still in learning phase on 27 November is a liability.

November: trade, don't build

By November the job changes completely. You are no longer building, you are trading.

Daily budget checks against pacing. Watching where the money wants to go and letting it. Creative rotation when frequency climbs. Stock and margin checks so you don't scale spend into something you can't fulfil or can't afford to sell.

Resist the urge to restructure. Every year someone rebuilds a campaign on the Tuesday before Black Friday because performance looked soft, resets the learning, and turns a slow week into a bad one. If something is genuinely broken, fix it. If it's just not exciting yet, leave it alone.

And plan for the concentration. Given how hard purchases still cluster on the two named days, your budget shape should look less like a plateau and more like two spikes with a supported middle.

December: the second half nobody plans

December splits into three distinct phases, and most plans only cover the first.

Up to around 19 December is gifting: last posting dates, delivery promises, gift guides, urgency messaging that's true rather than manufactured. From 20 to 24 December, physical delivery is off the table for most retailers, so the money moves to gift cards, digital products, click and collect, and local stock availability.

Then Boxing Day onwards is its own event, and this year it gets a Saturday plus a bank holiday Monday. Sale traffic, high intent, self-purchase rather than gifting, and much less competition for attention than November. It routinely delivers some of the best return of the whole quarter and routinely gets planned in the last week of December by someone on holiday.

The measurement questions to settle now

Before any of this goes live, agree the answers to these:

• What is a conversion, and does everyone in the business agree?
• Which attribution model are we judging Q4 on, and did we pick it before we saw the results?
• How do we handle the gap between platform-reported and back-end revenue?
• What's the incrementality question we want to answer, and what would we need to hold back to answer it?
• What does the January report look like, and do we currently collect everything it needs?

That last one catches people out constantly. Deciding in January what you wish you'd measured in October is an expensive way to learn.

The short version

Black Friday is 27 November 2026. Cyber Monday is 30 November. Christmas Day is a Friday, which shortens the final gifting weekend, and Boxing Day runs into a bank holiday, which lengthens the sale.

August is for feeds, audiences, measurement and briefs. September is for the AI Max migration, creative production and testing on a small budget. October is for getting live and stable. November is for trading, not building. December is three campaigns, not one.

None of this is complicated. It's just work that has to happen in a specific order, and the order starts now.

If you'd like a second pair of eyes on your Q4 plan before it's locked, get in touch. We'd rather look at it in August than rescue it in November.