Google ads testing Iba with privacy preserving signals raw data about ads-privacy HOT 7 OPEN

Hi, thanks for your enquiry.

Unfortunately, we are unable to share the raw data as it contains business sensitive information. If you could share the goals behind your ask, we are happy to partner and provide alternatives that can help accomplish the goal.

from ads-privacy.

Thank you for your prompt response.

On behalf of Alliance Digitale, a trade organization dedicated to assisting the industry in navigating emerging trends like Privacy Sandbox, we greatly appreciate your collaboration. As we approach the general availability phase, it becomes increasingly crucial for us to equip our members with comprehensive information regarding testing results. Therefore, we kindly requested access to raw data to ensure the accuracy of our analysis. Below, you will find the analysis we have conducted thus far, which we are eager to compare and validate against the actual data.

We have reviewed the white papers you provided and would like to address some observations we made regarding the results tables for both Google Display Ads (GAD) and DV360. Specifically, we noticed that while advertiser spend and conversion per dollar metrics exhibited minimal impact (-1.9% and -0.4% for GAD, +0.19% and -1.9% for DV360), the Click Through Rate (CTR) and Conversion Rate were significantly affected (-8% and -1.8% for GAD, -1.16% and -5.64% for DV360).

Initially, we encountered difficulty in reconciling these metrics, as CTR and Conversion Rate are expected to directly influence Conversion per Dollar metrics. However, upon further consideration, it became apparent that the reason advertiser spend and conversion per dollar were only slightly impacted in comparison to CTR and Conversion Rate was due to the disparity in impressions delivered between the test group and the control group. In other words, while the CTR and Conversion Rate may have decreased, the actual number of clicks and conversions could have remained unchanged due to the higher volume of impressions in the test group.

Using the provided Spend, CTR, Conversion Rate, and Conversion per Dollar impact metrics, we were able to deduce that a decrease of 5% to 10% in CPM (Cost per Thousand Impressions) could account for the differences observed between Spend and Conversion per Dollar metrics versus CTR and Conversion Rates. If this deduction holds true, it implies that the implementation of Topics instead of third-party cookies, limited solely to the scope of the experiment (i.e., only iba), had a 5% to 10% impact on CPM. Such an impact could potentially translate into a 5% to 10% reduction in revenue for publishers, which raises concerns considering the relatively narrow scope of the experiment.

In order to validate this assumption, we kindly request your assistance in providing the industry with the raw data. We understand that releasing such information may not be feasible; however, if possible, we would greatly appreciate access to CPM metrics for control groups versus test groups, or alternatively, impressions-level data for test groups versus control groups. Obtaining this additional information would greatly aid our analysis and enable us to reach a more accurate understanding of the observed results.

Thank you in advance for your attention to this matter. We look forward to your response and the opportunity to further explore these findings.

Best regards,
Alliance Digitale France

from ads-privacy.