Black Friday change freeze: what to lock, what can still change and how to log it

What is a Black Friday change freeze and why do you need one?

A Black Friday change freeze is a rule that says which changes your team may and may not make in the weeks around peak season. It protects the days that matter most from avoidable breakage, and it keeps the results readable: when few things change, you can tell what moved the numbers.

In 2026, Black Friday falls on Friday 27 November and Cyber Monday on Monday 30 November. Most teams freeze for one to three weeks around those dates, depending on how much of the year's revenue sits in that window.

There are two reasons to freeze, and both are about cause and effect:

• Smart Bidding needs time to recalibrate. Google says a bid strategy enters a learning period when it is created or reactivated, when one of its settings changes, or when campaigns, ad groups or keywords are added to or removed from it. Calibration typically takes 1–2 conversion cycles. A target change on 25 November may still be settling on Cyber Monday.
• Too many simultaneous changes make the peak impossible to explain. If a new checkout template, a consent banner update and a tROAS change all ship the same week, nobody can say which one cut conversions. A freeze keeps the list of suspects short.

A freeze is not a ban on work. It moves risky work earlier and turns every late change into a logged, approved exception.

When should the freeze start and end?

Start the freeze early enough that the last structural changes have finished learning before peak traffic, and end it after the peak has been measured. Your conversion cycle (the time from click to conversion) sets the minimum gap.

These dates are an example. If your typical conversion cycle is 7 days, Google's 1–2 cycles means a bid strategy change can need up to two weeks to calibrate, so the build phase should end at least two weeks before the hard freeze. Longer B2B or high-ticket cycles need an earlier cut-off.

What should you lock, and what can still change?

Lock anything that changes how results are measured or how the bidding system learns; allow changes that were planned, tested and are easy to reverse. Agree the list in writing before the freeze starts, so nobody negotiates it under pressure.

Two notes on budgets. Google Ads may spend up to twice your average daily budget on a given day, and over a month it will not charge more than 30.4 times the average daily budget. Raise budgets with that in mind, and log every increase with the reason, because a budget change is one of the first things people ask about when CPA moves.

For Meta campaigns, check which edits restart learning before the freeze; see which significant edits reset the Meta Ads learning phase.

How do you use seasonality adjustments instead of changing targets?

If you expect conversion rates to jump during the sale, a seasonality adjustment tells Smart Bidding about it without touching your targets. When the event ends, campaigns return to their usual behaviour; Google says no negative adjustment is needed afterwards.

Google's own guidance sets the limits:

• Use them only when you expect major changes in conversion rate, because Smart Bidding already accounts for normal seasonality.
• They suit short events of 1–7 days and may work less well for periods longer than 14 days.
• They apply to Search, Shopping and Display campaigns with Target CPA or Target ROAS, and to Performance Max with any bid strategy. Travel campaigns are not supported.

To create one, follow the steps in Create a seasonality adjustment:

1. Click the Tools icon, open "Budgets and bidding" and click "Adjustments".
2. Select "Seasonal" and click the plus button.
3. Under adjustment type, choose "Conversion rate", then add a name, description and start and end dates.
4. Choose the "Scope" (campaign types or specific campaigns) and enter the expected change in conversion rate.
5. Click "Create seasonality adjustment".

Write the expected uplift and the reason in the description. If results surprise you later, you will want to know whether the adjustment was set to +20% or +80%.

How do you log exceptions during the freeze?

Every change made during the freeze should be an approved exception with a written reason. That record is what lets you explain the peak afterwards.

1. Request: the person proposing the change writes what will change, why, and what they expect to happen.
2. Approve: one named owner per account says yes or no. No approval, no change.
3. Log: record the date and time, owner, platform, before and after values, and the reason on a shared timeline.
4. Check: set a time to look at the result, for example 48 hours later, and note what you saw.

Platform logs only cover part of this. Google Ads change history shows the past 2 years in the interface, but the API's change event data only covers the past 30 days, and neither shows your website release, a tag publish or the reason behind an edit. For tags, use Google Tag Manager version history alongside your log.

A marketing change log puts all of it in one place. In Changeline, a connected Google Ads account imports its changes automatically, with who made them and the before and after values, and your team logs the website, tracking and CRM changes next to them. You can start a change log for your peak season for free.

How do you isolate the cause if results move during peak week?

Start from the list of changes, not from the chart. Put every change from the days before the shift on one timeline, rule out tracking, then test one explanation at a time.

1. Build the timeline: your approved exceptions, Google Ads change history (including changes from automated rules, scripts and recommendations), Meta activity history, site releases, tag publishes and external events such as a competitor sale or a delivery cut-off.
2. Rule out tracking first: compare orders in your shop or CRM with platform conversions. If orders are steady and conversions fell, the problem is measurement, not demand.
3. Compare with the same weeks last year: what changed then, and how did the curve look? See year over year PPC analysis and how to tell seasonality from your own changes.
4. Test one hypothesis at a time: check whether the shift starts at the same hour as a change, in the same campaigns, devices or countries the change touched.

Example: ROAS drops 30% on Black Friday morning. The timeline shows no bid edits, but a checkout release went live at 07:00 and orders in the shop fell at the same time. The cause is the site, and the fix is a rollback, not a bid change.

If you need to present the finding, the guide to explaining a conversion drop to a client shows how to structure it.

With exceptions, bid edits and site releases logged on one timeline, the 07:00 checkout release sits right next to the ROAS drop, and you can show the client that no bid was touched. Start a free change log in Changeline before peak week begins.

What should you do after the freeze ends?

Measure first, then release held-back changes one at a time. Releasing everything on 2 December recreates the problem the freeze solved.

• Write a short peak review: results against plan, every exception and its effect, and what broke.
• Check that seasonality adjustments ended on the dates you set and that temporary budgets went back down.
• Queue the held changes by risk and release them with a few days between each, so each one can be read on its own.
• Keep the log. Next year, the same weeks will be your best reference.

To share the period with a client or manager, Changeline's Report button builds a two-page change report for two chosen dates, which you can save as PDF.