Decoding Google Shopping Target ROAS

Why a high ROAS is not automatically better – and what you really need to understand about Target ROAS.  

"We want to increase our ROAS." A phrase that comes up in almost every e-commerce setup. And at first glance, it seems completely logical. However, in the context of Google Shopping or Performance Max campaigns, this way of thinking often leads to misunderstandings. Because what many underestimate: As soon as you activate a Target ROAS strategy in your campaigns, the entire control logic changes. 

What sounds efficient in theory often has different implications in practice: 

• Decrease in impressions

• Fewer clicks

• Lower revenues

  … even though the ROAS looks "better" mathematically.

 Because: Target ROAS is not a "performance booster," but a controlling bidding objective that strongly depends on your data situation, the margin per product, and the actual conversion behavior. And this is often ignored.

That’s why you will learn in the following article:

• what Target ROAS really does (and what it doesn't),

• why blanket target values are almost always the wrong approach,

• how to build smarter bidding logics with margins, price bands, and custom labels.

So that in the end, you not only have a "nice" ROAS – but also more profit.

Setting Target ROAS in Google Ads

Once you define a Target ROAS for a shopping or PMAX campaign in Google Ads, you tell the system: "Please only show my ads when the expected conversion justifies this revenue value."

This means concretely:

Google analyzes each potential click in real-time – and calculates whether that click is expected to bring in enough revenue to meet your ROAS target value.

If the click has a revenue potential that is too low according to Google's forecast? → No bid.

Looks good? → Google automatically increases the CPC to secure the click.

This logic leads to Google: 

• Increasing CPCs for search queries with high conversion probability or strong purchase intentions (e.g., brand-related keywords, bottom-funnel signals)

• Lowering CPCs or not bidding at all on generic terms or unclear purchase intent 

Target ROAS can strategically steer your campaign – or completely stall it if you set the levers wrong.

Target ROAS in Google Shopping: Why a blanket value can stunt your growth

A single ROAS value does not work for an entire assortment. Because: Not all products are equally profitable.

A high-priced niche product with a fat margin can run super profitable at 300% ROAS. A low-ticket bestseller with a thin margin may need 800% to break even. Other products are more about bringing traffic and have strategic value, even if the initial ROAS is "below target."

And what happens when you still specify a blanket ROAS value – e.g., 600% across all products?

• Google automatically suppresses everything that does not reach this value according to the forecast.

• Your visibility decreases – even for products that are actually performing well (but just have slightly deviating margins).

• Revenue potentials are capped – at the expense of purely mathematical efficiency.

Using Margins & Custom Labels for Intelligent Bidding in Google Shopping

If you want to use Target ROAS really meaningfully, you need a product logic that does not rely on gut feeling – but on hard numbers. The basis for this: Margin.

And this is where many advertisers miss out on potential. Because even though Google has long supported the attribute Cost of Goods Sold (COGS) in the product feed, it is rarely used. Yet it is a game changer.

Why margins should be your actual decision-making basis

Not every product needs the same ROAS to be profitable.

What counts is: How much do you have left in the end – after ad spend, purchasing, logistics, etc.?

• A product with a 70% margin is already in the green area at ROAS 200%.

• A product with only a 10% margin? You might need ROAS 900% not to incur a loss.

  
  
  

Implementation with Custom Labels & Price Bands

• Tools like Label Up help you here: COGS & margin are made directly available in the feed

• Products are automatically labeled according to the margin logic (e.g., "40%+ margin," "low margin," "strategic")

• ROAS target values are assigned per label group instead of universally across the entire campaign

• Additional lever: Price bands

  • Products under €50 usually perform differently than those over €300. You can also target differently here

  • Your assortment is not homogeneous, so your ROAS targeting shouldn’t be either.

By using margin-based custom labels, you finally bring economic intelligence to your campaigns. 

Too high a Target ROAS can kill your sales

A wrongly set ROAS value can cost you more revenue than any poor creative test.

Especially in e-commerce, we see it repeatedly: Shops define their desired ROAS under the motto "The higher, the better." So the Target ROAS is set at 700%, 800%, or even 1000% – in hopes of maximum profitability.

And what happens then? -Visibility disappears. Clicks disappear. Revenue disappears.

If Google determines from your historical campaign data that specific products cannot reach this target value, the CPC is lowered so much that you simply no longer participate in the auction. Products that previously generated 80% of your revenue are removed, and impression share decreases.  Your "top sellers" lose their performance foundation, and your campaign crashes, even though your ROAS has mathematically increased.

Setting the Right Target ROAS

A good Target ROAS is not wishful thinking. It is based on data, history, and reality. If you define any target value in your campaign, Google will quickly slow you down.

Google recommends: Use the historical metric "Conversion Value / Costs" as a basis. Look at what your campaign has actually achieved in the last few weeks and set your ROAS not above it but slightly below or at the same level if you want stability.

Three Google Ads Mistakes You Must Avoid: 

1. Setting ROAS without a data basis:

   If you start a new campaign and immediately specify a ROAS of 600%, Google lacks the signals → your campaign starts blind.

   
   
   

2. Defining too high values arbitrarily:

   Sure, 800% sounds good – but if you've only achieved 450%, you will lose visibility.

   
   
   

3. Ignoring conversion delays:

   Depending on the attribution and tracking setup (Google Ads vs. GA4), conversions may only appear after days or weeks. Your actual ROAS values are always a bit delayed.

Read more here about the top 10 Google Shopping mistakes that you should definitely avoid: Google Shopping Tips and Tricks

At a Glance: Best Practice for Setting Target ROAS

• Start with a realistic value from your last 30 days

• Only adjust ROAS targets gradually

• Work with custom labels per margin or price band to optimize more distinctly

• Test new target values in separate campaigns or with portfolio strategies

  
  
  

Target ROAS in Google Ads: How often should you really change the target value?

Every change to the Target ROAS triggers a new learning phase. And this can take up to 2 weeks – during which your campaign may initially perform less well.

Rule of thumb: As rarely as possible, as thoughtfully as necessary.

• Don't change your Target ROAS more frequently than every 14 days

• Observe the effects over a longer period (at least 1-2 weeks)

• If you make a change, document it and do not change multiple things at once (otherwise you won't know what caused the effect)

 Google Ads campaigns with Target ROAS work with machine learning. The system needs: Reliable signals, time for evaluation, and stability

If you constantly tweak it, you disrupt the learning process.

Conclusion: Target ROAS is not a performance hack but a strategic tool

A high ROAS looks good in reporting. But if you sacrifice visibility, scaling, and revenue for it, you won nothing in e-commerce. Target ROAS is not a magic switch, but a sensitive control mechanism. When used correctly, you can grow more profitably, use your budget more efficiently, and scale your campaigns purposefully.

If you misuse it, you risk stalling your entire setup.

The most important learnings again at a glance:

• ROAS needs context: A blanket target value does not work for a diversified assortment.

• Margin first: Work with actual product data (COGS, price bands) and use custom labels for control.

• Data-based instead of wishful thinking: Set realistic target values based on history – not your revenue expectations.

• Avoid frequent changes: Give the algorithm time to learn and optimize your strategy.

  
  
  

If you need help optimizing your Google Ads – for example, with segmenting your feed by margin, price, or ROAS potential: Then let the Labelizer from Label Up assist you.

You can read more about it here.