Channel Prioritization in Google Ads 2026: Stop Funding Networks You Never Chose
Until October 2026 there was no honest lever for deciding how much of a daily budget went to Search, how much to the Display Network, and how much to YouTube or Discover. Search campaigns with Display opted in distributed spend through logic you could not see; Performance Max and Demand Gen decided the split entirely on their own. Channel prioritization Google Ads changes that with two new controls: a priority tier per channel and a percentage cap on what that channel may spend. Here is what shipped, how to configure it, and why some accounts will see spend jump in February 2027 if nothing is done.
What you get from this article: the setup order for four campaign types, a decision table for choosing priorities, a concrete plan for the February transition, and the metrics that prove whether a cap actually held.
What channel prioritization in Google Ads actually is
The feature has two connected halves.
Channel priority. You label a channel or channel grouping as High priority or Low priority. High-priority channels get funded first — within your target if you run Target CPA or Target ROAS. Whatever budget is left flows to lower-priority channels.
Channel-level budget allocation. This unlocks once a channel is set to Low priority. You specify the maximum percentage of the daily budget that channel may consume. Google’s own caveat matters: allocation is best-effort. The system aims to stay inside your limit but does not guarantee it on any given day.
The difference from older tools is structural. Device and location bid adjustments changed the price you paid, not the way budget was divided. Splitting channels into separate campaigns divided budget hard, but fragmented Smart Bidding’s learning and multiplied entities. Channel prioritization operates inside one campaign on one budget: the algorithm keeps learning on the full data set while you supply the guardrails.
Where the settings live
- Account level — set once, acts as the default for campaigns.
- Campaign level — overrides the account default for that campaign.
- Shared budgets — priorities apply there too, which makes pooled budgets far easier to steer. If you run portfolio strategies, revisit the pool logic at the same time; we covered the mechanics in the guide to portfolio bid strategies and shared budgets.
Which channels can be split apart
| Campaign type | Channels you can tier | What it is good for |
|---|---|---|
| Search | Google Search, search partners, Google Display Network | The main use case: keep Search high priority, give partners and Display a low tier with a hard percentage |
| Shopping | Google surfaces vs partner surfaces | Controls how much spend leaves the core commercial feed |
| Performance Max | Google surfaces bundled as “Google channels” (Search, YouTube, Maps, Gmail, Discover) | The bundle vs other inventory — no split inside the bundle |
| Demand Gen | Same bundling as PMax | Useful for teams that landed here after the Display migration |
One detail kills most of the inflated expectations: inside Performance Max and Demand Gen, Search, YouTube, Gmail and Discover sit in a single grouping. You cannot say “cap YouTube at 20%” — only separate the Google bundle from the rest of the inventory. Splitting traffic inside PMax still relies on the older levers: asset group structure, brand exclusions and channel-level reporting. Our take on controlling where PMax actually lands is in the piece on final URL expansion in Performance Max and AI Max.
The part to diary now: February 2027 for Search plus Display
For Search campaigns with the Display Network turned on, the current spend allocation setup ends in February 2027. After that, budget may redistribute across channels — in most accounts that means the Display share grows, because clicks there are cheaper and volume is easier for the algorithm to buy.
To keep something close to today’s split, Google spells out the configuration: set the Display Network to Low priority and give it an allocation of roughly 50% of the daily budget. Treat that number as the reference point in the recommendation, not as the right answer for your account — for search-led lead generation a sane Display share is usually well below it.
A Search campaign with Display enabled is two different economics sharing one budget. While the split was hidden, you could ignore the problem. From February 2027 either you set the share or the auction sets it, and the auction rarely sets it in your favour.
The honest alternative, which worked long before this announcement: pull Display out of the Search campaign entirely and run it separately — own budget, own creative, own evaluation. Prioritization is the better fit when you deliberately want a blended campaign; separation stays cleaner when the two channels have genuinely different costs per conversion.
Choosing priorities: a decision table
| Goal | High priority | Low priority + cap | Reasoning |
|---|---|---|---|
| Lead gen on existing demand | Google Search | Display 10–20%, partners 10–15% | Network leads look cheap on CPL and convert worse downstream; the cap bounds their influence on Smart Bidding |
| E-commerce with a product feed | Search and Shopping surfaces | Other inventory 20–30% | Most revenue comes from commercial intent surfaces |
| New market entry, building reach | Google channel bundle in PMax or Demand Gen | No cap yet | Early on, collecting signal beats squeezing a channel |
| Brand defence and remarketing | Brand search | Display 5–10% | This budget serves demand that already exists; network expansion adds little |
A caveat straight from the documentation: if a campaign is not budget-limited and runs on Target CPA or Target ROAS, prioritization alone will probably change nothing — the algorithm is already buying everything that fits the target. The effect shows up either when budget is the binding constraint or when you apply a percentage cap. Which is the usual order of operations: model the economics first, touch the settings second. The calculation sits in our piece on budget pacing and daily spend behaviour.
Rollout in seven steps
- Capture a baseline. Export 28 days of spend, conversions and CPA segmented by network — in the campaigns report add the “Network (with search partners)” segment. Without it you cannot separate the effect of the setting from ordinary noise.
- Compute CPA per channel separately. If Display CPA is two or more times your allowable cost, you have found your cap candidate.
- Set account-level priorities as your default posture.
- Override at campaign level where the logic differs: brand, non-brand, regional.
- Introduce the cap gradually. Start near the current actual share and reduce in 10-percentage-point steps every 7–10 days. Halving it in one move behaves like a budget cut: volume drops and Smart Bidding re-solves.
- Allow a recalibration window. Seven to fourteen days as a reference, but match it to your conversion lag — if a lead takes a week to mature, two weeks is the minimum before conclusions.
- Read results by network, not by campaign total. Compare equal-length periods and account for seasonality.
If caps change at the same time as ad schedules or bid targets, the effects blend and become impossible to untangle. Change one thing at a time — the same discipline we apply in campaign experiments in Google Ads.
What to measure
- Channel share of spend — the direct verification: the cap either holds or it does not. Check it daily for the first week.
- CPA and ROAS by network — answers “did it get better”, not merely “did something move”.
- Search impression share — if you are redirecting budget into Search, it should rise. If it does not, the constraint was never budget; it was bid or ad quality. The mechanics are in our breakdown of Quality Score and Ad Rank.
- Lead quality — a separate column. Network traffic often produces leads that never reach a deal; filtering approaches are in fake leads and form checks.
- Incrementality — the highest standard of proof. Where volume allows, run a geo split: capped in some regions, uncapped in others. Method in Conversion Lift for Search and Performance Max.
Five mistakes that already cost money
- Capping without a baseline. A month later you cannot prove the change helped. Any budget-mechanics change starts by recording the “before”.
- Treating the cap as a hard wall. It is best-effort. If overspend on a channel would genuinely hurt the business, turn the channel off instead of capping it.
- Expecting movement where budget was never the limit. A tROAS campaign with unspent budget will barely react to priorities alone.
- Re-tiering during a promo peak. Any change to allocation resets Smart Bidding’s equilibrium. Doing it three days before a sale means paying for learning at the most expensive moment of the year.
- Ignoring February 2027. The transition touches every Search campaign with Display enabled. Configure it deliberately now rather than reactively during a spend spike.
Who benefits most
Three groups. First, accounts that have run Display inside Search campaigns for years “because that was the default” — they gain both control and an honest read on what the network actually cost. Second, teams pushed out of Display and into Demand Gen who lost their familiar controls; priorities return part of that steering, as discussed in the Display to Demand Gen migration. Third, large accounts on shared budgets, where constraining a channel used to require splitting structure and running into Google Ads account limits.
One pairing worth stating plainly: channel priority decides where an ad appears, while match types and close variants decide which queries trigger it. Configure both layers together, otherwise the budget you just reclaimed goes to irrelevant searches — see today’s companion piece on keyword match types and close variants in 2026. And if reallocation sends more traffic to one landing page, confirm the page can convert it: correct testing procedure is in landing page A/B tests for paid traffic.
On the access side: in a standalone client account these controls appear as the rollout reaches you; in a manager-account structure they can be configured centrally and inherited. If you operate through Google Ads agency accounts, set priorities once at account level and override only where the logic genuinely differs. For a broader view of how we run paid search programmes, see the Google PPC agency overview.
Priorities and Smart Bidding: what normal looks like
An account behaves predictably after a priority change, as long as you know what counts as normal. For the first two or three days total conversions usually dip while spend holds: the strategy has stopped buying cheap volume in the constrained channel and has not yet replaced it in the prioritized one. That dip is the transition, not a fault.
From there, three outcomes. The one you want: spend share in the priority channel rises, account CPA holds or improves, conversion volume returns to baseline within 7–10 days. The second: spend does not flow into the priority channel at all, because its constraint was never budget but available auction volume — in which case you simply cut traffic, and the cap should be loosened. The third: spend did move, but CPA rose. That means the constrained channel was delivering genuinely cheap conversions you had assumed were junk on instinct rather than evidence; go back to reconciling lead quality with your CRM.
On timing: for a tCPA or tROAS campaign, any change to spend structure is a new equilibrium, and recalibration takes longer the fewer weekly conversions you have. At 10–15 conversions a week, expecting clarity before two weeks is wishful; at 100+ a week usually suffices. What the strategy’s normal exploratory behaviour looks like in that period is covered in the Smart Bidding learning period.
One habit that settles half the internal arguments: log the priority change as a dated entry with the expected outcome written next to it. Two weeks later you compare against a record rather than against memory.
Reading the reports after priorities are live
The most common evaluation mistake is reading the campaign total. Priorities change the composition of spend, so any aggregate metric hides the effect. What to open:
- The “Network (with search partners)” segment in the campaigns report — the baseline tool for Search and Shopping. It exposes spend share, conversions and CPA per channel.
- Performance Max channel reporting — for PMax and Demand Gen a dedicated view shows distribution across surfaces. Remember that priorities in these types act on the surface grouping, so you will see the internal split without being able to cap it.
- Daily, not weekly, granularity for the first week after the change. The cap is best-effort, and single-day overshoots only show up at day level.
- The impression share column — the test of the “budget was the constraint” hypothesis. If share does not move once budget frees up, the limiting factor is bid or quality.
- CRM reconciliation on quality. Without that layer you are optimising CPA rather than revenue; the gap between those two jobs is discussed in Google Ads and GA4 conversion discrepancies.
For recurring reporting, promote channel share of spend into its own dashboard column. It is the single metric that reveals whether the configuration is still in force or was quietly overridden at campaign level by someone else’s edit.
If the controls have not reached your account yet
The rollout is gradual, so some accounts will see these settings later. Everything below is worth doing regardless.
- Capture channel baselines for 28 and 90 days. When the controls appear you will have a reference point instead of a guess.
- Compute CPA and lead quality separately for Search, partners and Display. That is the answer to which channel earns the low tier.
- Find every Search campaign where Display is on “by default”. In many accounts that checkbox has been set since creation and nobody revisited it — those are exactly the campaigns the February transition touches.
- Decide the structural question: blended or separate. If the data says Search and Display have different economics, separation remains cleaner than a percentage cap.
- Name an owner for the setting. In accounts where several people make edits, priorities are easy to break with a stray campaign-level override.
FAQ
What is channel prioritization in Google Ads in plain terms?
It lets you tell the system which channels to fund first. A channel is labelled High or Low priority; high-priority channels are bought first within your Target CPA or ROAS, and the remaining budget flows to the rest.
How is a channel budget cap different from a priority?
The priority sets the funding order; the cap sets the maximum share of the daily budget in percent. The cap is only available for a channel marked Low priority.
Is the cap strictly enforced?
No. Google describes it as best-effort: the system aims to stay inside the limit but does not guarantee it. If channel overspend is unacceptable, switching the channel off is the reliable option.
Which campaign types support it?
Search, Shopping, Performance Max and Demand Gen. The rollout is gradual, so some accounts will see the controls later than others.
Can I cap YouTube alone inside Performance Max?
No. In PMax and Demand Gen the Google surfaces — Search, YouTube, Maps, Gmail, Discover — are treated as one grouping, and you cannot isolate a single channel inside it with a percentage.
What happens in February 2027 to Search campaigns with Display?
The existing spend allocation setup stops working, and budget may redistribute across channels. To approximate today’s split, set the Display Network to Low priority and apply a percentage allocation.
What percentage should Display get?
Google’s guidance cites roughly 50% as the setting that preserves current behaviour. For search-led lead generation that is usually too high: start from your actual current share and step down by 10 percentage points at a time while watching CPA and lead quality.
Do priorities work at account or campaign level?
Both, and they also apply to shared budgets. Account values act as the default; a campaign can override them.
Will this disturb Smart Bidding?
Yes. Changing the spend distribution changes which auctions the campaign enters, so the strategy looks for a new equilibrium. Plan a 7–14 day window and avoid stacking other changes on top. See the Smart Bidding learning period for what normal recalibration looks like.
Why bother if my campaign never spends its full budget?
In that case prioritization alone will likely change nothing, because the algorithm is already buying all available volume within target. It becomes meaningful under budget constraint or when you add a percentage cap.
How do I prove the change worked?
Compare by network rather than campaign total: share of spend, CPA and ROAS per channel, before and after, over equal-length periods. The most reliable proof is a geo experiment with the cap on in some regions and off in others.
Does this replace splitting channels into separate campaigns?
No, it is an alternative. Splitting gives hard budget control and clean reporting but fragments learning data. Prioritization keeps one budget and shared learning, with softer control.
