Case study: the conversions of a comparison site on Google Ads, server-side
Published on by the Oriva team.
On zonscope.com, the comparison site of one of our customers, Oriva counted 3,221 outbound clicks from Google Ads from October 1 to 9, 2026, where the server tag already in place counted 3,375, i.e. 95%. Day by day, the gap ranges from −9% to +6%. Over the period, Google Ads showed 6,353 clicks for €677.71. What this shows: server-side tracking, with no container to maintain, finds almost what a server tag already in place counts. What it does not show: that Oriva counts better or earns money; that is not measured here.
- 95%
- of the server tag's conversions found by Oriva, from −9% to +6% depending on the day
- 6,353
- Google Ads clicks over the 9 days
- €677.71
- of Google Ads spend, October 1 to 9
The site and what is counted
zonscope.com is a comparison site: it sends its visitors to partner sites. The conversion tracked is therefore an outbound click to a partner site, not a purchase on the site: there is no cart and no amount, which is why no ROAS is calculated here.
zonscope.com is the site of one of our customers. It gave us access to its Google Ads account: that is what lets us read both sides of the figures, and publish only what we can verify.
Two measures run in parallel on the same site, without knowing each other:
- Server tag already in place3,375
- Oriva3,221
- A server tag already in place, which sends each outbound click to Google Ads, in a conversion action of its own. Its figure is read in the “Conversions” column of Google Ads, by conversion date.
- Oriva, installed on the same site: it counts the outbound clicks whose originating visit carries a Google click ID (
gclid), by Paris day.
What the two measures give, day by day
| Day (Paris) | Oriva | Server tag | Oriva as % |
|---|---|---|---|
| October 1 | 242 | 259 | 93% |
| October 2 | 239 | 256 | 93% |
| October 3 | 199 | 217 | 92% |
| October 4 | 362 | 361 | 100% |
| October 5 | 389 | 427 | 91% |
| October 6 | 646 | 698 | 93% |
| October 7 | 456 | 478 | 95% |
| October 8 | 429 | 434 | 99% |
| October 9 | 259 | 245 | 106% |
| Total | 3,221 | 3,375 | 95% |
Oriva: outbound clicks whose originating visit carries a gclid. Server tag: the “Conversions” column of the tag's action in Google Ads, by conversion date. Data read on October 10, 2026.
Three things to know when reading this table:
- The two do not measure exactly the same thing. Oriva ties each outbound click to the latest visit carrying a
gclidwithin 30 days; Google dates the conversion to its own time; and the days do not cut at the same place. Oriva also only sees the visitors who accepted advertising cookies. - The cause of the gaps is not isolated. We know they are of the same order every day, not why.
- Nine days, one site. It is a measurement, not a rule.
What it means, and what it does not
What it shows
- It means that a reader who already has a server tag can check Oriva against it: the figure is readable on both sides, in Google Ads and in Oriva's log.
What it does not show
- It does not mean that Oriva counts better. On this site it counts a little less than the tag, around 5%.
- It says nothing about the advertising result: no amount, no ROAS, no before/after comparison.
The same check on ChatGPT Ads
- 98%
- of the 233 clicks counted by OpenAI found by Oriva (228 visits with oppref)
Same test on ChatGPT Ads clicks, the only days with clicks counted (September 25, then October 1 to 3): OpenAI counted 233 (€108.24), Oriva captured 228 visits carrying OpenAI's click ID (oppref), i.e. 98%. Day by day: 13 vs 10, 63 vs 61, 61 vs 62, 96 vs 95. The days do not cut at the same place (the OpenAI account's day on one side, the UTC day on the other).
As of today, conversions are not counted on OpenAI's side: this check covers clicks only.
Who does not need Oriva?
- You have a server tag running and someone to maintain it, and you only have one platform. It counts almost the same; switching tools has a cost.
- You do not run paid advertising. There is nothing to compare.
- Your conversions are purchases with an amount, which your tag already reports with their value. This study is not about value.
Oriva is for when you want the same figure without a container to maintain, across several platforms at once, with a log that says where each send went and why it failed. The detail is on the Oriva and Google Ads page, and the Google-side walkthrough is in the guide Create a Google Ads conversion action.
Start free trialSources
Data from zonscope.com's Google Ads account, read through the API on October 10, 2026, and Oriva's data for the same site, same date.