Google's New Target-Based Bidding Update: Are ROAS Targets Becoming Performance Ceilings?

Google's latest update to target-based bidding should make campaign performance more predictable. On paper, that's a good thing.

Campaigns that are Limited by Budget and using Target ROAS (tROAS) or Target CPA (tCPA) now optimise much more closely towards the target you've set, even when budgets change.

Google's guidance is simple: if you're happy with your current business target, leave it as it is. If you want to maintain your recent performance, adjust your target to match your last 28 days. Alternatively, switch to Maximise Conversion Value or Maximise Conversions if your goal is to maximise results from a fixed budget.

I understand the reasoning behind the change. In fact, Meta has worked in a similar way for years.

What I'm still trying to work out is what this means for the way we scale campaigns in the real world.

Why this update matters

Historically, campaigns using target-based bidding didn't always hit the target exactly.

In many cases, they exceeded it.

For example, you might have a campaign with a target ROAS of 20 that consistently delivered a ROAS of 24 or 25.

That overperformance wasn't just a pleasant surprise. It was useful.

It told me there was probably room to scale. I could increase budgets, test slightly higher targets and make informed decisions based on how much headroom the campaign appeared to have.

With this update, Google expects campaigns to optimise much closer to the target you've given them.

The upside is greater predictability.

The downside?

We may lose one of the signals we've traditionally used to understand when campaigns still had room to grow.

Should you simply match your last 28 days?

Google recommends adjusting your targets to reflect your recent performance if you want to maintain similar results.

For many advertisers, that's sensible advice.

For Australian ecommerce businesses, though, there's another consideration.

The last 28 days included EOFY.

For many retailers, July wasn't a normal trading period. It was heavily influenced by promotions, discounts and unusually high demand.

If I simply copy that performance into my bidding target, am I setting a realistic long-term business objective?

Or am I basing future optimisation on a temporary spike in demand?

That's why I'm taking a more measured approach rather than updating every account immediately.

The question I can't answer yet

The biggest question I have isn't whether this update is good or bad.

It's this: How do we identify the next scaling opportunity?

If campaigns are no longer expected to outperform their targets, how do we know when it's appropriate to increase a target ROAS from 20 to 22?

Previously, overperformance gave us confidence that there was still room to grow.

Going forward, what becomes that signal?

Should we rely on:

• Smart Bidding Exploration?

• Ads Advisor?

• Recommendations?

• Experiments?

• Or simply trust Google's automation?

At the moment, I don't think there's a clear answer.

My concern with Smart Bidding Exploration

Google has positioned Smart Bidding Exploration as the way to uncover additional growth.

Conceptually, I like the idea.

Practically, I still have questions.

Let's say I have a Performance Max campaign that captures both branded and non-branded searches.

If Smart Bidding Exploration starts looking for additional opportunities, how do I influence where that exploration happens?

Because there's a significant difference between:

• finding genuinely new customers, and

• simply capturing more branded traffic that was likely to convert anyway.

For some of my clients, separating those two isn't even straightforward.

Their product name is their brand.

People search directly for the product itself.

So how do I know Smart Bidding Exploration is creating incremental growth rather than simply leaning further into the easiest conversions available?

Without more visibility or controls, I think that's a genuine gap.

As advertisers, we don't just care about more conversions.

We care about more incremental business.

What I'm doing for my clients

Rather than applying one rule across every account, I'm taking a different approach depending on the type of business.

Stable ecommerce businesses

For businesses with relatively consistent demand, I'm waiting.

I'm collecting around 30 days of post-update data and gradually aligning targets with what campaigns are genuinely delivering.

I'm treating this as a calibration exercise rather than reacting immediately.

Businesses with unpredictable demand

Some accounts simply don't behave like the average ecommerce business.

These include:

• businesses with recent AI-focused restructures

• brands with products that can suddenly go viral

• businesses receiving large wholesale orders

• accounts where demand changes dramatically throughout the year

For these clients, I'm moving away from target-based bidding and switching to Maximise Conversion Value.

My reasoning is simple.

I don't want to tell AI what I think the business is capable of achieving.

I want AI to show me what's possible.

If I lock the system into a specific ROAS target too early, I worry I'm creating an artificial ceiling before the algorithm has had the opportunity to discover new profitable demand.

A funny thought I had...

When I first heard about this update, my immediate reaction was:

"Fantastic. I'll just set every campaign to a 20x ROAS and let Google work its magic."

Of course, that's not how Smart Bidding works.

A target doesn't create demand.

It simply defines the outcome the system should optimise towards.

And that's exactly why this update matters.

Targets are becoming much more literal than they were before.

My final thoughts

I don't think this update is a bad thing.

More predictable performance is valuable.

Better forecasting is valuable.

Consistent optimisation is valuable.

What I don't think we've fully answered yet is how advertisers should identify their next scaling opportunity once campaigns stop outperforming their targets.

Most ecommerce accounts I manage today are limited by budget.

If target ROAS is becoming a much firmer instruction to Google's AI, then choosing the right target may become one of the most important strategic decisions we make.

For now, I'm not rushing to update every campaign. I'm testing. I'm watching.

And I'm trying to understand whether we're setting business goals... or unintentionally setting performance ceilings.

I'd genuinely love to hear how other PPC practitioners are approaching this change. If you've started testing already, have you seen anything that changes the way you're thinking about scaling?