Key takeaways
How far has ROAS really fallen?
It depends on who measured. The four sources we read range from practically unchanged to minus 46%. They look at different campaign types, periods and customers, and they count ROAS differently. So none of them tells you where your account stands.
| Source | What was measured | Value |
|---|---|---|
| Channable, software for product data | ROAS of merchants in the EU and EEA, June 2025 to June 2026 | Performance Max minus 46%, Standard Shopping minus 43% |
| Optmyzr, tools for Google Ads | 21,425 Google Ads accounts, Q1 2025 against Q1 2026 | All accounts 428.5% to 430.9%, Performance Max minus 4.53% year on year |
| smec, software for budget and targets | Median ROAS in European e-commerce, latest weekly value (November 2, 2026) | 5.6, a year earlier 5.4 |
| adtribute, analytics for online brands | D2C brands, non-brand, last touch from shop orders, September 2026 | Minus 14% against the previous year |
All four publish for their own audience. Compare the rows with caution, they are not one data set.
Why do they not fit together? Four reasons stand out. The period differs: a full year, one quarter, one month and one week are different questions. The campaign type differs, since Channable looks only at Shopping and Performance Max while Optmyzr looks at all accounts. The customer base differs, because each vendor sees its own users, not the market. And the source of ROAS differs: adtribute points out that the ROAS Google reports and the ROAS attributed from shop orders are two different numbers.
The conclusion is plain. A headline gives you a direction, not a diagnosis. How your account stands, only your numbers can say, measured the same way over the same periods.
Why is your ROAS dropping?
ROAS is revenue divided by cost. It falls when the revenue side shrinks or the cost side grows, and often only your measurement changed. Work through the causes below in this order, because the first one decides whether the rest is worth your time.
- Measurement and tracking. A broken tag, a changed cookie banner or a changed attribution setting means you no longer see sales that still happen. Check this first, otherwise you repair something that is not broken.
- Basket value, prices, promotions and stock. Lower order values or fewer items on offer lower revenue per click. A promotion that ended can look like a drop in performance.
- Click prices. If every click costs more, the same revenue is divided by a bigger number. We cover this in our article on rising click prices.
- Product data. Weak titles, categories or identifiers send your products to the wrong searches. You can change that in your feed.
- Distribution in Performance Max. Budget can flow to channels outside search that convert worse. Look at where your spend lands before you change targets.
- Bid target and budget. A target that is too tight starves good auctions, one that is too loose buys weak clicks. You set both, so you can revisit both.
- Clicks without value. Some paid clicks never had a chance to buy, for example invalid ones. This is only one item on the list, and the one the next section is about.
How much ROAS do you gain when invalid clicks drop out?
On paper, a little. The ROAS after the drop-out equals the ROAS before, divided by one minus the share of invalid clicks. This holds if those clicks were charged and brought no revenue. It is arithmetic, not a promise about your account.
An example with round numbers. You make €400 in revenue and spend €100, so ROAS is 4.00. Of the €100, €5 went to invalid clicks without revenue. If those €5 drop out, your costs are €95, the revenue stays at €400, and ROAS is 4.21.
Lunio's retail report gives the shares. Lunio sells click protection itself. The report (status August 14, 2026) covers 414 million clicks from 88 retail accounts in monitor-only mode, so these are clicks that got past Google's own filters. Applied to the formula, the shares give these results:
| Lunio average | Invalid clicks | ROAS rises on paper by |
|---|---|---|
| Search | 4.73% | 5.0% |
| All channels | 5.0% | 5.3% |
| Shopping | 6.33% | 6.8% |
Calculation from Lunio's averages in these accounts, not a measurement from your account.
Read these values as a vendor's average, not a limit. Your account may look better, for example because you already use click protection, or worse. Credits that Google issues later shrink the effect. They appear on your invoice, while Google says its campaign metrics are not adjusted retroactively for them. And the figures come from a vendor of exactly this kind of tool. We do not set them against the 46% above. We only say that next to the large swings in the table they are small, and that this small piece is the one that can drop out.
How can you tell whether invalid clicks are pushing your ROAS down?
Look for several signs together. One alone proves nothing. A cluster of them gives you a reason to look closer. The question is never answered from outside, only your account and your logs show it.
- Clicks far above visits. Clicks in Google Ads sit clearly above sessions in your analytics. A part of this gap is normal: visitors leave before the page loads, or they do not consent to measurement.
- Regions you did not target. Clicks arrive from places you never advertised in.
- Bounce rate close to 100% on single campaigns or placements, while the rest of the account looks normal.
- Click spikes without orders. Traffic jumps and the orders stay flat.
Be careful with the conclusion. Google itself says that a rise in clicks often has other reasons, such as budget, bids, new keywords, targeting or season. The column Invalid clicks in Google Ads shows only what Google already filters out, and those clicks are not charged. It is not the number you paid for in error. Outside tools can also flag clicks that Google has already filtered, so the same click may be counted twice. Treat the sum with caution.
What can you do now?
Go in this order: check measurement, work through the causes, read what Google already shows you, and only then think about a review request. Start recording logs today, because proof exists only from the day you begin.
- Check measurement. Test your tag and your cookie banner, and compare orders in your shop with conversions in Google Ads.
- Work through the cause list from the second section and write down what you changed and when.
- Open the Google Ads reports. Look at the column Invalid clicks and the report on credits for invalid activity. Credits appear on your invoice as the line "Invalid activity" under adjustments, given when appropriate and possible, as described on Google's help page on invalid traffic.
- Request a review if you have a pattern and proof. You can ask for a review of the last 60 days through the click quality form. Google decides, and a request is no promise of a credit. How to prepare is in our section on the review request.
- Record logs now. You only have evidence for the days you stored it, and the 60 days keep running.
How does seacashback help you with this?
seacashback checks the clicks on your landing pages, excludes detected bots at account level and documents invalid clicks, so you can report them to Google. You see up front how much of your budget is affected, and then you decide.
It does not change the auction and it does not raise your ROAS there. It lets detected invalid clicks drop out, and that dampens costs without revenue, which improves ROAS on paper by a little. As the arithmetic above shows, that is a small share of everything that can happen when ROAS falls. Whether invalid clicks play a part for you, only your account can answer. Read how it works under How it works. For data on invalid clicks in shops, see the Click Protection Report 2026. It measures German shops.
Frequently asked questions
Does the ROAS column in Google Ads already exclude invalid clicks?
Partly. Google does not charge for invalid traffic it detects before billing, so those clicks do not enter your costs. Clicks it detects later can be credited when appropriate and possible. What slips through unnoticed stays in your costs.
Can invalid clicks be the reason my ROAS dropped?
They can play a part, but on paper only a small one. In Lunio's figures, dropping them lifts ROAS by 5.0% to 6.8%. Tracking, order values, product data and bid targets usually explain more. Only your own account can tell you.
Does click protection raise my ROAS in the auction?
No. Click protection does not change the auction. It lets detected invalid clicks drop out, and that improves ROAS on paper a little, because costs without revenue fall away.
How do I find out if I paid for invalid clicks?
Look at the column Invalid clicks, the report on credits for invalid activity and the adjustments on your invoice. Then compare Google Ads clicks with visits in your analytics and check the regions and campaigns that stand out.
