Four weeks have passed since Index revealed that they were using bid caching – or the act of keeping a buyer’s bid when it fails to win an impression and applying that bid to a later ad request – in order to “improve their user experience.” There were many reactions from different parties following the revelation.

Most of the issue stemmed from the fact that the practice was not fully transparent, and some buyers even complained that the bids “submitted” via the bid caching process were bids that they would not have made or would have done differently, had they known. We’ve come to understand that the practice was live for several months before it was shut down mid-August, a period during which Index Exchange experienced an increase in market share.
To dig deeper… Following the reactions of different parties on the programmatic landscape, we looked at real data to understand how the controversy really impacted Index. We’re going to focus on the short-term impact for Index.
In order to make our analysis as useful as possible, we have selected a group of publishers in different markets (United Kingdom, United States, France, Belgium, Spain, Netherlands and Belgium) that are using at least AppNexus, Index, Rubicon, and GAM (Google AdX). This way, we were able to detect an overall drop in revenue and impressions and/or more isolated drops.
Since we consider August 14th as the day when everything started, we looked at 7 days before and 7 days after.
At first glance, it is clear that all the SSPs had a slight drop in revenue and/or impressions during this period, though some more than others. The overall change is probably due to seasonality. Nevertheless, Index had taken a bigger hit than the rest of the SSPs.
While the revenue drop for AppNexus and GAM varied from 0 to 5%, and slightly increased for Rubicon by nearly 4%, Index took a big hit with more than a 20% decrease.
Analysis-of-the-impact-of-Index-bid-caching

Let’s focus on some buyers to dig in to what happened:

Index witnessed a revenue drop, and as a consequence, buyers slightly compensated in other SSPs.

Google Adwords

The overall trend on Google Adwords was a decrease in spend.
For GAM and AppNexus, we noticed roughly a 5% decrease in spend whereas on Index, we saw an 11% decrease. Rubicon is the SSP whose Adwords revenue increased by 24%.
From the day we started the analysis, Google Adwords was roughly spending the same amount on Index and Rubicon. On the last day of our analysis, Adwords was almost spending double on Rubicon than on Index.

Accuen OMG

Overall during the period, Accuen OMG’s spend during the period dropped by 1%.
What was interesting to notice was a budget transfer from Index to AppNexus.
While revenue dropped by 20% on Index, it increased by 26% on AppNexus and 12% on Rubicon.

Cadreon

They dropped their revenue by 22% on Index while making a 134% increase on their spend on Rubicon.
Overall, they increased their spend everywhere else except on Index.

There were few others cases on a smaller scale such as Sizmek, Vivaki, and Quantcast. 

Conclusion and Recommendation for Publishers

The bid caching had a measurable impact on Index. There was a revenue drop for the publishers working with Index SSPs and it is noticeable in buyers’ spends. Overall, we see that there was a slight drop in the market due to seasonality, however, buyers that decreased their spend on Index compensated by transferring their spend to other SSPs.
Looking at the data, what we recommend for the industry and specifically for publishers is to advocate for more transparency by being more demanding with their partners. Transparency is beneficial for everyone, it builds trust, increases revenue overall and helps both parties avoid problems in the future.

Tags :

adtech bid caching programmatic market trends publishers