Google Ads Target ROAS Not Working: Why Smart Bidding Misses Your Targets

Why is my Google Ads target ROAS not working?

Google Ads target ROAS fails for five main reasons: the target is set too high relative to historical performance; the account does not have enough conversion volume for Smart Bidding to work reliably (Google recommends at least 30-50 conversions per month at campaign level); conversion tracking is inflated or broken; budget constraints are preventing Smart Bidding from bidding competitively; or a recent target change triggered a learning phase that has not yet settled. Fix conversion data quality first, then review whether your target is achievable from where the account sits today.

Google Ads target ROAS not working is one of the most common problems we see when auditing e-commerce accounts. The strategy looks straightforward: tell Google what return on ad spend you need, and let Smart Bidding find the auctions that deliver it. In practice, several things can break it quietly, spending your budget without coming close to the number you set. This post covers the five causes we see most often in the accounts we manage, and the order in which to investigate them.

1. Your Target ROAS Is Set Too High for Your Historical Data

The most common cause (and the one Google rarely surfaces clearly) is a target that is simply too ambitious for where the account is right now. If your last 30 days returned a 5x ROAS and you set a target of 10x, Smart Bidding will restrict bids to a price point where it believes it can hit 10x. In most accounts that means it stops entering a large proportion of eligible auctions: delivery drops, spend falls, and reported ROAS technically improves on a smaller, cherry-picked subset of clicks while overall revenue falls sharply.

The fix is to anchor your target in reality. Check your actual ROAS over the last 30-90 days in the Campaigns view, then set your target 10-20% above that number, not where you eventually want to be. If your historical ROAS is 6x, a starting target of 7x-8x gives Smart Bidding enough room to bid confidently while still pushing performance in the right direction. You can increase the target in 10-15% increments every two weeks as the account demonstrates it can sustain the higher bar.

The mechanics of increasing targets without crashing delivery are covered in more detail in our guide to scaling Google Ads without losing ROAS.

2. Your Account Does Not Have Enough Conversions

Target ROAS is a data-hungry strategy. Google recommends at least 30-50 conversions in the last 30 days at campaign level before enabling it. Below that threshold, the algorithm does not have enough signal to distinguish between high-value and low-value auctions reliably, and its predictions become erratic.

In the accounts we manage, we typically see this play out as a campaign that swings wildly week to week: a strong week followed by a very poor one, with no obvious external cause. That volatility is usually a data problem rather than a targeting or bid problem.

If you are below the 30-conversion floor, the right move is to remove the ROAS target temporarily and switch to Maximise Conversion Value. That strategy still optimises for revenue but without a specific target to anchor bids, it bids more broadly and builds conversion data faster. Once you have consistently 30 or more conversions per month, you can add a target ROAS that reflects your actual historical performance. Our full breakdown of which Google Ads bidding strategy to use at each stage walks through this sequencing in detail.

3. Your Conversion Tracking Is Inflated or Broken

Smart Bidding optimises for the conversions it can see. If your conversion tracking is overcounting (recording duplicate purchases, including micro-conversions like add-to-cart as primary conversions, or pulling in GA4 imports alongside a tag-based purchase event), the algorithm thinks its bids are performing better than they actually are. It bids accordingly into auctions that do not return the value it expects, and your real ROAS falls short of your target despite the platform saying otherwise.

The symptoms are a reported ROAS that looks healthy but does not match your actual revenue from paid search, or a figure that looks reasonable inside Google Ads but is nowhere near that in GA4 or your e-commerce backend.

The check is straightforward: compare your Google Ads reported conversion value for last month against the actual revenue your store received from paid search. If there is a meaningful gap (typically more than 15-20%), you almost certainly have a tracking problem that needs fixing before any bidding changes will stick. Our guide to Google Ads conversion tracking for e-commerce covers the most common errors and how to resolve each one. Always audit tracking before adjusting strategy settings.

4. Budget Constraints Are Throttling Delivery

Target ROAS needs room to bid. If your daily budget is consistently hitting its cap before the day ends (look for the "Limited by budget" status in the Budget column of the Campaigns view), Smart Bidding cannot enter the auctions it would choose. It ends up rationing bids across the day, often missing the highest-converting windows (typically mornings and early evenings for UK retail), and reported ROAS suffers as a result because it is only reaching a fraction of the relevant traffic.

The fix depends on context. If you are hitting budget limits and your current ROAS is strong, increasing the daily budget to give Smart Bidding the headroom it needs is usually the correct call. If ROAS is already marginal, raising the target ROAS slightly to make the algorithm more selective and reduce overall spend makes more sense. Neither is universally right.

Note that Google's automated budget recommendations inside the interface are designed to increase spend, not necessarily to maximise your efficiency. Treat them as directional information rather than instructions.

5. A Recent Change Triggered a Learning Phase Reset

Every significant change to a target ROAS setting, a budget increase of more than 20%, or a campaign restructure pushes Smart Bidding back into a learning phase. During that period (typically 7-14 days for established accounts, occasionally longer) performance will be inconsistent and often worse than before the change. That does not mean the change was wrong; it means Smart Bidding is recalibrating its auction predictions.

The mistake we see repeatedly is making a second change during the learning phase because the first one appeared not to work. Stacking changes resets the clock and can keep a campaign in an indefinite learning loop with no stable baseline to assess.

The rule in the accounts we manage: make one meaningful change at a time, hold for at least 14 days before drawing conclusions, and document every change with a date so you can attribute cause and effect accurately. If you have made multiple changes recently and performance is unstable, stop making changes and give Smart Bidding two full weeks of consistent conditions. Check the "Learning" status label in your Campaigns view to confirm the campaign has exited that phase.

When Target ROAS Is the Wrong Strategy Altogether

Not every account is a candidate for target ROAS, even with good data. If your product range spans a very wide value gap (for example, a cycling store selling both chain lube at £8 and complete bikes at £2,500+), target ROAS can behave unpredictably because one high-value sale can distort the algorithm's conversion value model for that campaign. In cases like that, separating product tiers into distinct campaigns with their own strategies often gives better results than a single target ROAS across everything.

If your conversion volume is consistently below 30 per month, target CPA on lower-AOV product lines (where every transaction has broadly similar value) can be a useful intermediate step while you build volume on the higher-value side. The broader decision framework is covered in our bidding strategy guide for e-commerce.

If you have worked through all five areas above and target ROAS is still consistently underperforming, a full account audit is the quickest route to a clear answer. There is usually a single root cause rather than several compounding ones, but it often takes a systematic review of the account structure, feed quality, tracking setup and campaign settings to isolate it. You can book a free PPC audit with Social Surge and we will identify the specific issue in your account rather than working from general principles.

Frequently Asked Questions

How long does it take for target ROAS to start working in Google Ads?

Target ROAS typically needs 4-6 weeks of consistent data to stabilise. If your account has fewer than 30-50 conversions in the last 30 days, the strategy will struggle to gather enough signals and will often throttle delivery or underspend. In the accounts we manage, we usually give a freshly set target ROAS at least 14-21 days without changes before assessing performance.

How do I set a realistic target ROAS in Google Ads?

Start from your actual historical ROAS over the last 30-90 days, then set your target 10-20% above that. Do not set a target based on where you want to end up; set it based on where Smart Bidding can plausibly get to from where you are now. If you have no historical data, use Maximise Conversion Value first and let the account build at least 30 conversions before switching to a fixed target ROAS.

What is the minimum conversion volume for target ROAS to work?

Google recommends at least 30-50 conversions per month at campaign level. Below that threshold, Smart Bidding does not have enough data to find patterns reliably, and performance becomes erratic. For newer accounts, Maximise Conversion Value (with no target) is a safer starting point.

Should I use target ROAS or target CPA for e-commerce Google Ads?

For e-commerce, target ROAS is usually the better fit because it optimises for revenue value, not just transaction count. If your products all sell at a similar price point, target CPA can work well. If your product range spans a wide AOV (average order value), target ROAS is better because it weights bids toward higher-value purchases.