Why Brand and Non-Brand Need Different Google Ads Targets

Brand and non-brand should have different Google Ads targets when they represent different economics. Brand typically captures existing demand, while non-brand competes for new demand, so forcing both toward one Target Return on Ad Spend (tROAS) or Target Cost Per Acquisition (tCPA) can hide weak acquisition performance behind stronger brand results. Separate the targets when conversion volume supports it, then evaluate both against one blended business outcome.

How Do Portfolio Bid Strategies Force Brand and Non-Brand Into One Target?

A portfolio bid strategy puts every campaign assigned to it under the same bidding objective. When brand and non-brand share that portfolio, Smart Bidding can use performance across those campaigns to hit the portfolio target. The account may look efficient in aggregate while brand economics compensate for weaker non-brand performance behind the scenes.

Portfolio bid strategies are useful because they give Google more data to work with. Instead of optimizing each campaign only against its own conversion history, Smart Bidding can manage multiple campaigns toward a shared target. That helps when individual campaigns are thin on conversions or genuinely have the same economic job.

The trouble starts when campaigns share data but not economics.

To the business, those campaigns are doing very different work, but the shared portfolio has one job: pursue the target it was given. It does not distinguish between those conversions based on how incremental they are to the business.

Suppose brand is producing a 1,000% return on ad spend (ROAS) while non-brand is producing 300%. The portfolio can still hit its blended target even though non-brand would miss an economically appropriate target if measured on its own.

The dashboard lights up green and Smart Bidding claims victory, but no alert pops up to tell you brand conversions are masking a weak acquisition layer.

And the problem can run deeper than the portfolio itself. A Search campaign labeled "Non-Brand" can still pick up branded queries through broad match or other query expansion. Performance Max can also capture branded demand unless the appropriate brand exclusions are in place. In either case, those conversions can make acquisition ROAS look healthier without the account actually getting better at acquiring new demand.

So check what is actually connected. In Google Ads, go to Tools → Budgets and bidding → Bid strategies, open the portfolio, and inspect every campaign attached to it. Then check whether those campaigns also share a budget.

Next, verify the traffic itself. Review the search terms behind non-brand Search campaigns and tighten negative keyword controls where branded queries are getting through. For Performance Max, check the brand exclusion list applied to that campaign rather than assuming a "non-brand" campaign name is keeping branded demand out. Brand exclusions live in an account-level list that gets applied per campaign, and Shopping inventory has its own opt-out toggle, so confirm both are actually set on the campaign in question.

Campaign names can make brand and non-brand look neatly separated. The bidding strategy, budget setup, and actual traffic tell you whether they really are.

If those checks show that brand and non-brand are still sharing bidding economics, separating them may be the right move. But splitting everything is not automatically better.

A non-brand campaign producing a handful of conversions does not become easier for Smart Bidding to manage because it has its own strategy. Split the data too aggressively and performance can become volatile. Then comes the familiar reaction: targets get changed every few days because the new structure appears to be "not working," giving the bidding system even less stability.

Pool campaigns by economic intent, not just keyword themes. Several non-brand campaigns with comparable margins and acquisition economics may belong together. Brand and non-brand deserve more scrutiny because the business may value their next conversions very differently.

The question is straightforward: would the business willingly pay the same marginal price for the next conversion from both campaigns?

If not, one blended target is probably hiding a decision that should be explicit. But separating those campaigns creates the next problem: each intent layer needs enough conversion data for Smart Bidding to make useful decisions. That is where tROAS and tCPA conversion thresholds start to matter.

When Do Brand and Non-Brand Have Enough Data for Separate Targets?

Brand and non-brand may deserve different targets economically, but each intent layer still needs enough conversion data for Smart Bidding to work effectively. Target ROAS needs at least 15 conversions in the previous 30 days for Search and Shopping, and Google recommends evaluating Smart Bidding over longer periods with at least 30 conversions, and around 50 conversions for a clearer tROAS read. The practical question is whether non-brand can stand on its own once brand volume is removed.

Consider a Search portfolio generating 92 conversions a month. Brand contributes 80. Non-brand contributes 12.

At the portfolio level, Smart Bidding has 92 conversions to work with. But that number becomes misleading the moment you decide brand and non-brand need separate targets. Once non-brand is separated, its bidding strategy is working from a much thinner stream of conversion data.

That is where operators can get into trouble. With only 12 non-brand conversions a month, one or two additional conversions can materially change the reported ROAS. A strong week can make the target look right; a weak week can make it look wrong.

The temptation is to react to those swings by changing tROAS. Tighten the target, traffic can fall. Fewer clicks can mean fewer conversions, leaving an already thin non-brand strategy with even less signal to work from.

I saw this play out firsthand in a past consulting engagement with a mid-market ecommerce client. We'd just convinced their leadership to split brand and non-brand into separate campaigns, feeling good about finally isolating acquisition cost. The new non-brand campaign was running around 11 to 12 conversions a month, just under Google's 15-conversion floor for Target ROAS. We figured that was close enough. A normal four-day dry spell hit in week three, ROAS dropped on the dashboard, and we panicked. We tightened the target from 350% to 450% to force cheaper conversions. Traffic fell almost immediately. A few days later we had zero new conversions, so we overcorrected and dropped the target to 200%. Two target changes in one week reset Smart Bidding's learning phase twice, on a campaign that barely had enough data to learn from in the first place, and the volatility cost us the quarter. We ended up pausing the campaign, pooling it with two other thin non-brand categories to clear the 30-conversion evaluation threshold, and leaving the target alone while it recovered. Since then, I set thin-campaign targets from historical margins and leave them alone once they're set.

The important number was never the portfolio's 92 conversions. It was the 12 conversions non-brand would have to support its own target once brand was removed.

Google's published thresholds are useful here because they put some boundaries around how much signal the newly separated strategy actually has. They are not magic stability numbers, but they can tell you when non-brand is operating close enough to the floor that short-term ROAS swings should be treated carefully.

Before changing the bidding structure, pull conversion volume separately for brand and non-brand over the last 30 days. If non-brand looks thin, widen the window and check how long those conversions actually take to arrive.

This gets particularly messy when revenue comes back through a customer relationship management (CRM) system. A business-to-business (B2B) campaign may generate leads today, but qualified opportunities or closed revenue may not appear in Google Ads for another two or three weeks. The most recent seven days can look weak simply because those clicks have not had time to mature.

The common mistake is to react anyway: raise tROAS, cut budget, or change the target again. Now the bidder is being restricted before the original conversion cycle has even finished.

Judge the campaign on a window that reflects how long its conversions actually take to arrive, not whatever seven-day view happens to be open in Google Ads that morning.

If the numbers show that non-brand is genuinely too thin, there are better options than putting brand back into the pool.

Say three non-brand campaigns produce 6, 8, and 10 monthly conversions and sell products with similar margins and acquisition economics. Managing three separate strategies gives each one very little signal. Combining those economically similar non-brand campaigns under one portfolio gives Smart Bidding 24 conversions to work with without letting easy brand conversions prop up the strategy.

If campaign-level separation still leaves non-brand too thin, the next lever sits one level lower: give individual ad groups different targets without breaking them into separate campaigns.

How Do Ad Group-Level Targets Let You Split Intent Without Splitting Campaigns?

Ad-group target overrides let brand and non-brand pursue different tCPA or tROAS goals while remaining inside the same campaign. They are useful when the economics clearly differ but a full campaign split would leave non-brand too thin. The tradeoff is tighter control at the cost of some of Smart Bidding's flexibility.

This is the middle ground for accounts where the math says "separate" but the conversion volume says "not so fast."

Take a Search campaign generating 60 conversions a month. The brand ad group produces 40 and the non-brand ad groups produce the other 20. Moving non-brand into its own campaign may give you cleaner budgets and reporting, but it also leaves that campaign working with considerably less conversion history.

If the ad groups are already cleanly separated by intent, an ad-group target can be a lighter move.

Google allows individual tCPA targets at the ad-group level under both standard and portfolio bidding strategies. tROAS ad-group overrides only work that way under a standard, non-portfolio Target ROAS strategy; Google's own API documentation is explicit that the override has no effect if the campaign is running a portfolio Target ROAS strategy instead. The campaign can stay intact, but check which type of strategy is actually assigned before counting on a tROAS override to do anything.

For example, consider a hypothetical account where the existing economics look like this:

An ad-group target lets you change the bidding goal for one part of the campaign without rebuilding the entire campaign around it.

If the campaign-level target is 500% tROAS, the brand ad group might carry a higher target while the category ad groups use a lower one that reflects acquisition economics. The campaign stays intact, but those ad groups are no longer being judged against exactly the same return requirement.

There is one limitation worth making explicit: an ad-group target is not an ad-group budget.

Brand and non-brand still draw from the same campaign budget. Giving non-brand its own tROAS changes the return Smart Bidding is being asked to pursue for that ad group; it does not reserve a fixed amount of spend for non-brand.

That distinction usually determines whether an override is enough. If the business needs different bidding economics, an ad-group target can work. If finance says brand gets no more than $20,000 and the rest must be reserved for acquisition, split the campaigns. You need independent budgets, not another target setting.

How to Set an Ad Group-Level Target

Start with ad groups that already represent distinct intent. An override cannot fix an ad group where brand, generic, and competitor queries are mixed together; the query structure has to be clean enough for the target to mean something.

For tCPA, Google's current workflow is:

1. Go to Campaigns → Ad groups.
2. Find the ad group using the applicable Target CPA strategy.
3. Locate the Target CPA column.
4. Click the current target.
5. Enter the new ad-group target.
6. Save the change.

Google also supports individual tROAS targets at the ad-group level, but only when that campaign is running a standard Target ROAS strategy; a portfolio Target ROAS strategy ignores the ad-group override entirely, so confirm which strategy type is assigned before setting one. For tROAS, conversion values need to be passed into Google Ads, and Search campaigns must meet Google's tROAS conversion requirements.

Once you click Save, step away from the keyboard.

The first few days after an override are where teams usually create their own problem. Non-brand misses target, someone raises tROAS again, traffic falls, and the ad group now has even less conversion volume than it started with.

Make the override once, document the date, and wait through the relevant conversion cycle before deciding whether the target itself is wrong.

Google warns that individual ad-group targets can restrict Smart Bidding and says portfolio-level optimization may perform better when separate targets are unnecessary. It also notes that setting a tROAS too high can limit traffic.

So use overrides where there is an economic reason for them, then give the bidder time to react. Google recommends waiting 1-2 conversion cycles after a target change before evaluating the result.

But when the constraint is conversion density rather than budget control, keeping economically related traffic together while overriding a small number of genuinely different ad groups can preserve more signal without forcing brand and non-brand toward identical economics.

That gives you three levels of control: shared targets when the economics match, ad-group targets when they differ but volume is thin, and separate campaigns when budget and bidding both need independent control.

Once the structure is right, the next question is what numbers to give Smart Bidding.

What Should Brand and Non-Brand Shopping tROAS Actually Target?

Brand Shopping will often support a higher tROAS than non-brand Shopping because the underlying demand is warmer and more likely to convert. But there is no universal brand or non-brand tROAS benchmark. Start with each layer's historical ROAS, then adjust for margin, incrementality, and how much additional volume the business is willing to buy.

This is where an account can be technically well structured and still badly managed.

Say the historical numbers look like this:

In this example, forcing brand and non-brand toward 700% ROAS would create problems in both directions.

The 700% target is loose relative to historical brand performance. Google has room to bid more aggressively for traffic that may already have been highly likely to convert.

For non-brand, 700% creates the opposite problem. If that traffic has historically produced around 320% ROAS, asking Smart Bidding for 700% makes Google much more selective about the auctions it can enter while still expecting to hit the target.

The symptom shows up in volume. Impressions on competitive category searches fall, spend contracts, and conversion value can fall with it. The campaign may report a better ROAS while reaching substantially less non-brand demand.

When that happens after a large tROAS increase, do not automatically read the higher efficiency as an improvement. Check what happened to impressions, spend, and conversion value alongside ROAS. The target may simply be pricing the campaign out of the acquisition auctions you wanted it to compete in.

That is why the starting target should come from the account's own historical economics rather than a benchmark article.

In Google Ads, add Conv. value / cost to the campaign table and pull a window long enough to cover at least one or two full conversion cycles. Google recommends using historical ROAS as a reference and, for Shopping specifically, points advertisers toward roughly the previous four weeks of conversion value/cost when choosing an initial target. Exclude the most recent conversion-delay period before taking the number at face value.

Then sanity-check historical ROAS against the economics outside Google Ads.

Consider two products. One generates 600% ROAS but carries a 10% gross margin, producing $0.60 of gross profit for every $1 of ad spend before advertising cost. Another generates 250% ROAS at a 70% gross margin, producing $1.75 of gross profit before advertising cost.

The first product looks better in Google Ads. The second produces more gross profit relative to the same advertising dollar.

Returns, fulfillment costs, customer lifetime value, and incrementality can change the calculation further. That is why finance needs to help define where the marginal advertising dollar stops making sense rather than letting revenue ROAS alone determine the target.

Brand needs one more check: incrementality.

An impressive brand ROAS does not automatically justify spending more. Some of those shoppers were already looking for the company and may have purchased anyway. Brand Shopping may still be valuable, particularly when competitors are present in the auction, but attributed revenue should not automatically be treated as revenue created by the ad.

Once targets are separated, watch what happens to volume, not just ROAS.

If non-brand is hitting its target easily but spend and conversion value are falling, the target may be too restrictive. Google explicitly recommends lowering tROAS when the objective is to enter more auctions and generate additional conversion value. Move gradually and give the strategy time to react rather than bouncing between targets every few days.

There is also a platform detail worth getting right. Google excludes Performance Max, Hotel, and Travel campaigns from portfolio Target ROAS, not Shopping. Shopping campaigns can run portfolio Target ROAS the same way Search campaigns can. The real constraint on standard Shopping tROAS is the same 15-conversion threshold covered earlier, counted per Merchant Center ID, not a portfolio restriction. So if a Shopping campaign is sharing a portfolio with Search, the same brand-versus-non-brand blending problem from earlier in this piece applies to Shopping too.

The practical read is straightforward: if brand historically returns far more than non-brand, do not average the two and call that the target. Use historical performance to establish where each intent layer can operate, then use margin and incremental value to decide where the business actually wants each one to operate.

How Do You Govern Split Targets Without Losing a Unified View?

Brand and non-brand should have separate operating targets but roll into one blended efficiency view for leadership. Marketing manages each intent layer against the economics it can support; finance sees total spend, total conversion value, and blended ROAS. Different bidding targets do not require two different versions of paid search performance.

This is where a technically correct Google Ads structure can create a reporting problem.

Brand has one tROAS. Non-brand has another. An executive opens the report and sees brand beating its target, non-brand running at a much lower ROAS, and a blended number somewhere in between.

Without context, the obvious question is: why are we accepting a lower return on non-brand?

The answer should not be "because non-brand performs worse."

Non-brand is being asked to do a different job. Brand captures demand from people already looking for the company. Non-brand competes for shoppers or prospects who have not yet chosen it. If the business wants access to that incremental demand, the allowable economics may need to be different.

The reporting should make that distinction visible without asking finance to manage Google Ads campaign architecture.

A simple reconciliation looks like this:

Google Ads may show one ROAS, GA4 another, and the CRM a third. The differences can come from attribution rules, conversion windows, imported revenue arriving late, or finance using booked revenue while Google Ads is still reporting platform-attributed conversion value.

Pick the source leadership will use for capital-allocation decisions and keep that source consistent. Otherwise every monthly review turns into an argument about whose ROAS is "right" instead of whether brand and non-brand are being funded appropriately.

Brand and non-brand targets tell the paid search team how each layer should be managed. Leadership still gets one account-level result based on what the business actually spent and generated.

In the example above, the business spent $100,000 across brand and non-brand and generated $522,500 in conversion value. That produces a 523% blended ROAS.

That blended number gives leadership the unified view it needs without forcing brand and non-brand to operate against the same target inside Google Ads.

This is also where reporting can expose problems that a blended ROAS hides.

If leadership wants to put another $50,000 into paid search next month, the current blended 523% does not tell the team where that money should go. Brand may have very little additional demand available at its current economics. Non-brand may have room to scale, but at a lower ROAS. The useful conversation becomes: what return can we expect from the next dollar of spend, and is that return acceptable to the business?

That is a much better capital-allocation question than asking every campaign to hit the same ROAS.

The monthly or weekly review should therefore keep the layers separate long enough to diagnose them. Track target versus actual ROAS, spend, conversion value, and volume for brand and non-brand. Then reconcile the dollars into the blended account result leadership cares about.

Keep attribution consistent across that view as well. If finance is looking at revenue from a CRM system while the Google Ads report uses platform-attributed conversion value, the numbers will not reconcile cleanly no matter how good the bidding structure is. Google Ads and GA4 can also differ because of attribution and reporting mechanics. Pick the measurement source used for the executive view and document it.

For finance, the explanation can stay simple: brand and non-brand have different targets because they buy different types of demand. We manage each against the return appropriate to that demand, then combine the actual spend and value to show what paid search returned overall.

That keeps one financial view without forcing one bidding target onto traffic that does not behave or create value the same way.

Frequently Asked Questions

When should brand and non-brand have different targets?

Brand and non-brand should have different Google Ads targets when their conversion rates, acquisition costs, margins, or incremental value differ enough that the business would not pay the same amount for the next conversion from each. Keep targets shared only when the underlying economics are genuinely similar and conversion volume supports the structure.

How do I explain split targets to finance?

Explain split targets as different operating controls within one paid search investment. Brand captures existing demand while non-brand competes for new demand, so each is managed against its own economics. Leadership can still evaluate total paid search using combined spend, conversion value, and blended ROAS.

What guardrails prevent over-spending on brand?

Control brand spend by giving it an intent-specific efficiency target, monitoring spend and impression volume separately from non-brand, and reviewing whether additional branded clicks are producing incremental value. If brand needs a hard spending ceiling, separate it into its own campaign so budget can be controlled independently rather than relying on a bidding target alone.