The number everyone quotes but nobody has published
When you use Google's own CSS, a share is taken out of every bid before it enters the auction. The industry says 20%. Google does not publish the figure, not in its documentation and not in the account, and we have not found a primary source that does. Every number you see on this subject, ours included, is therefore measured from the outside or copied from somebody else who measured from the outside.
That is why this page exists. Not because our measurement is large, but because there are almost no public measurements at all.
What we measured
The Danish webshop Danmarks-Posen moved from Google's own CSS to ours on 16 March 2026, and the switch took effect from 17 to 19 March. The numbers compare the 30 days before with the 30 days after on their Performance Max campaign, with the three switch days left out. Cost per click, clicks and spend are also measured for the Shopping ads from the product feed alone, because those are the ads the CSS works on. Everything comes from the Google Ads API, one query per period.
| Before | After | Change | |
|---|---|---|---|
| Impression share, whole campaign | 23.8% | 28.8% | +21% |
| Impression share lost to rank, whole campaign | 72.5% | 65.6% | 6.9 percentage points better |
| Cost per click, whole campaign, index | 100 | 86 | -14% |
| Clicks, whole campaign, index | 100 | 111 | +11% |
| Spend, whole campaign, index | 100 | 96 | -4% |
| Cost per click, Shopping ads from the feed, index | 100 | 82 | -18% |
| Clicks, Shopping ads from the feed, index | 100 | 90 | -10% |
| Spend, Shopping ads from the feed, index | 100 | 74 | -26% |
Clicks, cost per click and spend are shown as an index where the period before is 100, because the amounts are the customer's own. Across the whole campaign, all four performance metrics improved at the same time while spend fell by 4%. The gain was not bought with a larger budget. On the Shopping ads from the feed, cost per click fell 18% and clicks fell 10%, because Performance Max moved part of the budget from them to other formats.
Read the Danmarks-Posen case study
Why this is not proof
If the March switch were a real effect, it should stand out from every other point in the series. So we ran every possible switch week from April 2025 to March 2026 through exactly the same method on the whole campaign, the 30 days before compared with the 30 days after, 52 weeks in all.
| Metric | The March switch ranks |
|---|---|
| Impression share | 17th of 52 |
| Impression share lost to rank | 18th of 52 |
| Cost per click | 12th of 52 |
The switch sits around the top third, but not at the top. Summer 2025 showed far larger swings with nothing happening on the CSS side. The account also started a new campaign in the period after the switch. And there is no control group: a CSS switch applies to the entire Merchant Center account at once, so no part of the account carried on unchanged for comparison.
The honest conclusion: it is an indication, not proof. We do not present it as anything else.
Measure it yourself, and measure the right thing
The most common mistake is to look at cost per click alone. Smart Bidding sees that it is reaching its target and buys more volume, so cost per click (CPC) can stay flat while you get more for the same budget. Measure impression share and impression share lost to rank as well.
In Performance Max the advantage applies to Shopping inventory only. Search text ads, Display, YouTube and Gmail are unaffected, so a combined campaign report dilutes the effect. You can see it here: cost per click fell 14% on the whole campaign and 18% on the ads from the feed.
The fields we used: search_impression_share, search_rank_lost_impression_share, average_cpc, clicks and cost_micros on the campaign resource. Pull each period as one query over the whole date range, so Google calculates the weighted impression share itself. An average of daily figures gives a different and wrong number.
And write down the day, not the month. The first version of this page assumed 1 March, because only the month had been noted. The real day, 16 March, was in the CSS Center log, and we have recalculated every number from it.
What we do from here
One account is an observation. We now measure every new customer before and after with the same method and the same fields, and we publish the numbers here as they come in. With around ten measurements the noise starts to average out, and the result can be quoted without caveats.
Measured 28 October 2026. Source: Google Ads API, Danmarks-Posen's Performance Max campaign, 15 February to 16 March 2026 compared with 20 March to 18 April 2026. First version published 21 September 2026. Write to info@thecsspartner.com for the full method, or to have your own account measured.