Cookieless World

Will Your Bidding Efficiency Drop without 3rd Party Cookies?

Accurate valuations

We have previously presented what the user will see in a cookieless future, what products will be displayed, and the end results – the visual consequences of the proposed changes to 3rd-party cookies. In this article, we would like to focus on what is happening behind the scenes.

In this article you will learn about:

  • Does the elimination of 3rd party cookies affect accurate valuations?
  • What is the solution for the challenges ahead in a cookieless future
  • Is it possible to maintain the current accuracy of valuations when user data security is significantly improved?

Table of Contents:

Accurate valuations in the cookieless future

We use powerful algorithms to find the right user at the right time to display an ad at an optimal price to distribute the customer advertising budgets effectively. These algorithms are fed a lot of data, especially for impression valuation. The more information we deliver to the bidder – the more willing the bidder will be to buy.

Ultimately, the idea is to match the price that is paid for the impression to the likelihood to the purchasing parameters of the user. Today’s algorithms can accurately estimate this parameter and specify the price accordingly which assists with keeping an advertising budget in line.

By default, an impression served to the person clicking on many products, adding them to the basket, and frequently returning to the website is more expensive than the casual visitor. Such a pattern should be adequately mirrored in the page view price.

All the changes resulting from a cookieless future and the difference in the approach to targeting individuals as opposed to using 3rd party cookies appear to limit the possibilities of an accurate valuation. The apparent necessity of group valuations could leave several inaccuracies as compared to individual valuations.

The inaccuracy of impression valuations, especially in the context of first-price auctions, would increase the ineffectiveness of advertising budget allocations. A valuation that is too low would result in the inability to buy an impression, while a valuation that is too high would cause suboptimal budget spend.

Outcome-based TURTLEDOVE (OBTD) is the direct answer to these potential challenges.

EMEA Automotive Guide | See our report
The auto industry is on the brink of the biggest changes in a century. This guide gives an overview of the current state, as well as practical advice and big ideas for auto industry leaders who want to use digital advertising formats to drive business.

What does Outcome-based TURTLEDOVE do?

The main idea of Outcome-based TURTLEDOVE is to separate the mechanism of impression price valuation for a given user from the means of microtargeting prevention. As the authors of this proposal, we would like to show you how it works.

Thanks to OBTD, when a user visits the advertisers’ website, we can save additional information in the user’s browser as part of the so-called UserSignals. This data will be stored directly on the user’s device and will not be accessible to external companies.

The diagram below is designed to help you understand the process and show you how helpful this information can be:

Valuation process

= On the advertiser’s website, the user is added to an interest group: “t-shirts_buyers”
= Visits only premium t-shirts of an average conversion value of 100 PLN vs. just 30 PLN for the whole category

How to improve the accuracy of this valuation compared to other members of the interest group for the auction?

= RTB House sends what the value of the “prod_conv_value” variable should be for this user, based on 1st party data from the advertiser
= We define the bidding function for the whole interest group, for example:
BID = 0.01 * prod_conv_value
= The variable, stored in the user’s browser, is then used in the on-device auction for displaying an ad on the publisher’s website while keeping everything privacy-friendly

Based on the results of our simulations, the accuracy of bidding is as high as 99.98% (measured by AUC) of the current precision.

See also: How Will Retargeting Work without 3rd Party Cookies?

Wrapping up

Since the industry needs to maintain the accuracy of valuations at current levels, RTB House has proposed a solution that makes this possible while protecting the identity of individual users.

The separation of the microtargeting prevention mechanism from the bidding mechanism allows for a flexibility increase – it can be adjusted without modifying the bidding module.

In addition, we have developed a proof of concept ad validation algorithm and analyzed its microtargeting prevention aspects. We carried out an extensive analysis and proved that the current separate mechanisms meet the highest mathematical guarantees of privacy and security.

If you have any questions or comments regarding the article, please let us know.

Into the future of online ads?
Subscribe to receive email notifications about our latest blog posts
By subscribing you agree that RTB House S.A. (data controller) will process your personal data to send the newsletter and – if you agree by checking the box below – to contact you to discuss potential cooperation. You may revoke your consent by "Unsubscribing" or at For more information, see Privacy Policy.