Black Friday is less than four weeks away, and most advertisers are in the same uncomfortable position. Campaigns are finally stable, performance looks decent, and nobody wants to be the person who touches a setting and watches results fall apart right before the biggest shopping weekend of the year. That fear keeps a lot of budgets frozen exactly when they should be moving.
Google Ads product manager Tal Akabas addressed that fear directly in a recent episode of the Ads Decoded podcast, hosted by Google’s Ads Product Liaison Ginny Marvin. The conversation covered new Channel Prioritization and Budget Allocation controls, but the most useful parts were the smaller details that rarely make it into launch announcements. Here is what matters before peak season.
The learning myth, finally put to rest
Ask any PPC manager what they worry about, and near the top of the list is the idea that changing a target CPA, a target ROAS, or a daily budget by more than 20 percent wipes out everything Smart Bidding has learned. Akabas said plainly that this is not how it works on a campaign that is already running steadily.
Changing a target or a budget on a steady-state campaign does not erase historical data, and it does not push the campaign back into “Learning” status. The system already holds bid landscape data, so it can scale bids in proportion to the new target instead of starting from scratch. The “Learning” label in the interface is not the same thing as the continuous adjustment happening in the background.
There are exceptions. A brand new campaign still needs time to ramp up for its mix of assets, targeting, and goals. A new or reactivated bid strategy, a major change to campaign composition, or a significant change to conversion goals can still trigger Learning status. For new conversion-focused campaigns, Google’s advice from the episode is to start with Maximize Conversions or Maximize Conversion Value rather than treating Maximize Clicks as a mandatory first step.
The practical point is simpler: stop leaving targets and budgets untouched out of fear. If business conditions change, change the settings. Just allow for conversion delay afterward, and avoid stacking repeated target changes inside a single conversion cycle.
Skip seasonality adjustments for ordinary holiday spikes
Seasonality adjustments let you tell Smart Bidding to expect a short-term change in conversion rates, for example during a sale. With Black Friday coming, switching one on feels like the responsible move. Akabas said you do not necessarily need one just because it is Black Friday and Cyber Monday. Google’s year-over-year models already anticipate standard holiday spikes.
There is also evidence that reaching for the adjustment can backfire. A study of Black Friday accounts from 2022 to 2024 found that accounts using seasonality adjustments saw cost per click rise about twice as much as accounts that did not use them. The adjustment group also saw return on ad spend fall in each year studied, even though its revenue grew faster. Those are observational comparisons, not proof that the adjustment caused the difference, but they are a good reason to be selective. Save adjustments for anomalies the system cannot know about: an off-cycle flash sale, a sitewide discount much deeper than last year’s, or a promotion timed in a way no historical pattern would predict.
Promotion Mode and exploration need room to work
Promotion Mode is rolling out to Search and Performance Max. It lets you schedule a promotional window with dedicated budget and bid changes, and it combines a faster ramp-up with Smart Bidding Exploration during that window so the campaign can capture additional traffic. It works on existing steady-state campaigns, not just new ones.
Smart Bidding Exploration, now rolling out to retail Performance Max and coming to Shopping, gives the algorithm a small tolerance, typically around 10 percent effective ROAS flexibility, to bid into new queries and incremental auctions it would otherwise skip. Here is the catch: it requires an eligible target ROAS campaign with a budget that does not constrain spend. The budget has to be set high enough that it never limits spending, so the ROAS target does the guiding. The moment the budget becomes the constraint, exploration switches off. If you plan to use either tool during Black Friday, check the budget headroom first. A capped budget quietly disables the exact feature you turned on to find extra demand.
Channel Prioritization only works when budgets bite
Channel Prioritization is the headline feature from the episode. It sorts supported channels into High Priority and Low Priority tiers across Search, Shopping, Performance Max, and Demand Gen, and Budget Allocation adds an optional daily spend limit for Low Priority channels. The idea is a middle ground between opting into a channel fully and switching it off entirely.
The mechanics depend on something many advertisers overlook: whether the campaign is limited by budget. When a budget limit exists, spend flows to High Priority channels first, then shifts to Low Priority channels once additional High Priority spend would miss the target, and only when those channels are predicted to deliver conversions at or better than the target.
When there is no budget limit, which is Google’s recommended setup for target CPA and target ROAS campaigns, Channel Prioritization alone changes nothing. Smart Bidding already chases every eligible conversion at your target across every network you opted into. That creates an awkward tension: Google recommends uncapped, demand-led budgets, but a budget that never limits spend leaves Channel Prioritization with nothing to do. On a truly uncapped campaign you need Budget Allocation to restrict a Low Priority channel, since a 0 percent allocation means no spend at all.
Watch out for the ghost budget trap
Budget Allocation percentages are calculated against your average daily budget setting, not your actual spend. That single detail can quietly wreck a cap. Imagine a campaign with a 5,000 dollar “uncapped” budget setting that typically spends 500 dollars a day. A 10 percent Display cap still permits 500 dollars of daily Display spend, which is as much as the whole campaign normally spends in a day. The cap looks small on paper and behaves like no cap at all in practice.
Two more fine-print details deserve a check. Budget Allocation caps operate on a best-effort basis and reset daily with no rollover, so during volatile traffic periods Low Priority spend can exceed the cap. And Channel Prioritization works with shared budgets only when every campaign on the budget is the same type. Because enforcement is best-effort, verify the result in cost reporting rather than trusting the setting.
Mark February 2027 for Display Expansion
Search campaigns opted into Display through Display Expansion currently treat Display as a low-priority outlet for unused Search budget. Starting in February 2027, Google replaces that allocation behavior with the new controls, and advertisers may see spending shift between Search, Search Partners, and Display. Google plans to add a one-click migration button, or you can manually set Display to Low Priority with a 50 percent budget allocation, which Google has confirmed keeps the current configuration, subject to the same best-effort enforcement. Audit your account now and flag every legacy Search campaign using Display Expansion so the change does not catch you off guard. Also check which of those campaigns use Target Impression Share, because the Display Network will no longer serve on Search campaigns using that bid strategy.
Lifecycle goals get easier to test
Feeding better customer data into Google Ads remains one of the highest-leverage things a PPC team can do, and three updates make lifecycle goals easier to run and audit. A new reporting-only mode for New Customer Acquisition lets you measure new versus returning customers and verify your data strength before putting bidding dollars behind it. A re-engagement only mode, just launched in Performance Max and coming to Search and Shopping, focuses the campaign budget on a churned or disengaged customer list. And Loyalty mode, now generally available for Performance Max and Shopping, pairs loyalty audience lists with differential bidding and member pricing or specialized promotional callouts on Shopping ads.
Closing the gap between finance and the auction
The episode’s bigger theme was demand-led budgeting: pairing uncapped or highly flexible budgets with strict efficiency targets so campaigns expand automatically when demand spikes. The logic is sound. If you are willing to pay 50 dollars per acquisition, you should buy every 50 dollar conversion available instead of shutting off at 2 p.m. because a daily budget ran out. Missing efficient conversions early can mean paying more for them later.
In practice, PPC teams answer to finance departments and clients who run on fixed monthly budgets by business unit. Three steps bridge that gap. First, anchor conversion values and targets in real economics, using profit-based values or revenue targets calibrated to margins and customer lifetime value. If a target overstates what an acquisition is worth, extra budget headroom just scales unprofitable spend. Second, use planning tools with filters, so budget reallocation suggestions stay inside one brand’s profit and loss instead of drifting across business units. Third, if you cannot uncap a monthly budget, give campaigns enough daily headroom that exploration features are not throttled on peak days, then layer pacing monitors or automated rules on top to alert you, dial back daily caps, or raise target ROAS when spend runs ahead of plan.
The throughline is reassuring: the system is more forgiving of changes than most advertisers believe. What it is not forgiving of is settings that quietly contradict each other, a cap calculated on a budget you never spend, or an exploration feature throttled by the budget meant to support it. Fix those contradictions now, and Black Friday becomes a demand problem, which is the good kind of problem to have.

FAQ
Does changing my target CPA reset Smart Bidding learning?
No, not on a steady-state campaign. Changing a target or budget does not erase historical data or trigger Learning status, though performance can still shift as bids and traffic mix adjust. New campaigns, new bid strategies, and major conversion goal changes are the cases that can still trigger Learning.
Should I use a seasonality adjustment for Black Friday?
Only if your promotion is unusual in a way Google cannot predict, such as an off-cycle flash sale or a much deeper discount than last year. Standard holiday spikes are already built into Google’s year-over-year models.
Why is Smart Bidding Exploration not working in my campaign?
It requires an eligible target ROAS campaign with an unconstrained budget. If your budget limits spend, exploration is disabled, so raise the budget headroom or accept that the feature will not run.
Does Channel Prioritization do anything on an uncapped campaign?
On its own, no. Without a budget limit, Smart Bidding already pursues every eligible conversion at your target. Use Budget Allocation to cap Low Priority channels, keeping in mind the cap is a percentage of your budget setting, not actual spend.
What should I do about Display Expansion before February 2027?
Audit your account for Search campaigns using Display Expansion, then either wait for Google’s one-click migration button or manually set Display to Low Priority with a 50 percent allocation. Also note that Display will stop serving on Search campaigns using Target Impression Share.