Users Pay Twice: The Hidden Energy Cost of Web Advertising

[SamuelPelissier]

Hello everyone,

First of all, thanks for your invaluable work on uBlock Origin.

I am one of the authors of a paper measuring the energy consumption of ads & tracking on end-user devices, with and without ad blockers, including uBlock Origin. I thought it might be of interest to the community.

TL;DR:

• We tested on Linux systems, with the latest (at the time) version of Chrome, with uBlock Origin (MV2 and Lite with MV3) installed. We also tested a network-based adblocker for comparison.
• We measured CPU (with integrated GPUs) energy consumption of Chrome during thousands of website crawls with and without adblockers.
• Blocking ads is generally worth it energy-wise.
• MV2 uBlock Origin shows greater energy-consumption reductions compared to MV3 uBlock Origin Lite.
• Some specific websites show energy-consuming behaviors when blocking ads.

You can find the full paper here (published @ The Web Conference 2026, open access): https://dl.acm.org/doi/10.1145/3774904.3792414.

Other resources:

• A slightly more public-friendly summary version: https://pelissier.eu/users-pay-twice/
• Presentation slides: https://pelissier.eu/docs/UsersPayTwice_TheWebConf2026_slides.pdf
• Source code: https://zenodo.org/records/18309442

Of course, as with all studies, there are several limitations and lots of questions remain regarding ads & tracking w.r.t. energy consumption: is it the same on other browsers (e.g., Firefox) or operating systems (e.g., Windows, macOS)? What about GPU energy consumption? What about server-side consumption?

Don't hesitate if you have any questions!

[Yuki2718]

BTW