Google Ends Search's Hidden 50% Display Cap, and 50% Was Always Too Much
Google will retire the hidden 50% budget allocation that has capped Display Expansion on Search campaigns, moving every Search campaign with the Display Network switched on to channel prioritization starting in February 2027. Google's help page says to set Display to Low priority with a 50% daily allocation to keep things as they are. For most accounts, 50% is far more than Display has earned.
The change itself isn't brand new. Google announced channel prioritization and budget allocation on October 14, and the February line was in that post. What's new is that Google's Display Expansion help page now says, in writing, that the old setup ran "with a 50% budget allocation applied in the backend." PPC Land spotted the rewrite on October 24. So for years there was a number sitting behind the "Include Google Display Network" checkbox, and most of us never saw it.
Honestly, that part bugs me more than the migration does.
What actually changes in February
Right now, Display Expansion is supposed to live off scraps. Google's help page describes it as serving Display ads with unspent Search budget, and that 50% backend allocation was the ceiling. You didn't set it, and you couldn't change it.
Under the new model you get two controls instead. Each channel (Google Search, Search partners, the Display Network) gets a High or Low priority tier, and Low priority channels can get an optional daily cap expressed as a percentage of the campaign's average daily budget. Google's channel prioritization help page spells out the default: "By default, all channels are set to High priority." No cap applies unless you set one.
That's the bit worth sitting with. If Google migrates your campaign and doesn't stamp a Low/50% setting on it (and the help pages don't promise it will), Display shows up in February as a High priority channel with no percentage limit. PPC Land flagged the same gap: the Display Expansion page never says what happens to campaigns that take no action, and the channel prioritization docs point to High with no cap unless an account default says otherwise.
Two other details from the help pages are easy to miss:
- Target Impression Share campaigns lose Display entirely. The channel prioritization page says "GDN will no longer serve on Search campaigns utilizing the Target Impression Share bid strategy." If you run brand defense on Target IS with the Display box ticked, that spend just disappears. Probably fine. Possibly a surprise in a client report.
- Campaigns that aren't budget-limited mostly ignore the tiers. Per Google, prioritization "doesn't change how your budget is spent when your campaign isn't limited by budget." For a Target CPA or Target ROAS campaign with room in its budget, the percentage is the only lever that does anything.
Why Google's suggested fix is too generous
Google says Low priority plus 50% keeps your current configuration. Maybe. The old 50% was a backend allocation applied to unspent budget, and the new one is 50% of average daily budget on a best-effort basis. Google hasn't explained whether those two 50s measure the same thing, and PPC Land noted the page never says why 50% is the matching figure. I'd treat "equivalent" as a rough approximation, not a promise.
But even if it's a perfect match, so what? Matching the old ceiling just preserves a setting nobody chose on purpose.
The evidence on Display Expansion has never been flattering. When Seer Interactive tested Display Expansion on budget-constrained tCPA Search campaigns, US campaigns saw cost rise 100% and conversions rise 32%, which pushed cost per lead up 51%. UK campaigns did worse: cost up 89%, conversions down 33%, cost per lead up 184%. That test is from 2021 and the network has changed a lot since, so I wouldn't treat it as gospel. It's still a decent picture of the downside when Display shares a Search budget.
Google's own pitch is "up to 15% more conversions at similar CPA" and an average of 14% more conversion value for Target ROAS and Maximize Conversion Value campaigns. The same help page says Display cost per conversion "may vary," and then a boxed note says it "should be similar" to Search. Pick one, Google.
A cap you inherit is a decision you didn't make. February is when that stops being a technicality.
The 20-minute audit before Display gets a vote
This is the part you can do Monday. It doesn't need the API, which is lucky, because Google Ads API and Editor support is still listed as "coming soon," according to Relevant Audience's breakdown of the documentation.
- Find every Search campaign with Display on. Filter the campaigns table by network setting. In older accounts this tends to be more campaigns than you'd guess, since the box was easy to leave ticked at setup.
- Segment the last 90 days by network. Compare Display CPA (or ROAS) against Search CPA for the same campaign. Use your primary conversion action, not "all conversions," because view-throughs and micro-conversions flatter Display a lot.
- Apply a simple threshold. My rough rule: if Display CPA is more than 1.5x Search CPA, set Display to Low with a 0% allocation (Google says 0% stops spend on that channel). If it's within 1.2x, Low with a 10% to 20% cap is a reasonable test. Only campaigns where Display is genuinely at parity have any business near 50%.
- Set an account-level default. Google lets you set priorities once at the account level, and they apply to every campaign of that type unless overridden. Put Display at Low with your chosen cap there, so the campaigns you build next spring don't quietly inherit High.
- Flag Target IS campaigns. Note any Display spend on them now, so the February drop doesn't get read as a performance change.
One catch for agencies: channel prioritization can't be set from a manager account, per Relevant Audience. It's per client account or per campaign. If you run 40 accounts, that's 40 sessions of clicking, so put it on the calendar now instead of the last week of January.
And remember the cap is best-effort. Google says daily spend on a channel "can occasionally go over its allocation," allocations reset daily, and nothing carries over. A 10% cap on a $500 daily budget works out to about $50, and some days it'll be $60. Check weekly for the first month, not daily, or you'll drive yourself a little nuts.
Where this sits in Google's bigger channel shuffle
This is at least the third time this year Google has moved the walls around Display. The network moved under Demand Gen in May and standalone Display campaigns were retired, per PPC Land's coverage. Then the October controls arrived, which we covered when Google first let advertisers cap Search partners at 0% but not YouTube. Ginny Marvin's note that PMax advertisers can set Search partner spend to 0% followed two days later.
To be fair to Google, the new system is better. Explicit tiers and a visible percentage beat a hidden 50% you couldn't touch. I just think the transition is set up so the easy path (doing nothing) keeps Display money flowing, and the default leans the same way. It fits with what Brainlabs found earlier this month: Search still takes most of the budget while other surfaces pick up a growing share of clicks. Google pretty clearly wants more of the mix on the non-Search side.
My prediction, and I could be off on the size: at least a third of active Search campaigns with Display Expansion will reach March 2027 with no explicit channel setting at all, and most advertisers who notice will find out from a CPA spike, not from a Google notification.
Give Display a test budget and a review date
None of this means Display is useless on Search. For some lead-gen accounts with tight budgets and decent image assets, it can probably fill real gaps (the help page also says responsive display ads "may create" automatically from uploaded images, which is worth checking separately). The difference now is that you get to decide how much it gets.
So decide. Pick a number from your own 90 days of data, write it down somewhere, and look at it again in April. If Display earns a bigger share, give it one. It just shouldn't get half your budget because a backend setting nobody showed you said so.
