Ad blocker performance study in response to Manifest V3 finds that Ghostery’s ad blocker beats the competition
A recent study by Cliqz analyzed the effectiveness of popular content-blockers, finding that overall they are very efficient, having sub-millisecond median decision time per each request.
Additionally, the study found that the ad blocker used by Ghostery and Cliqz in many of their privacy products performs as well or better than the other popular content blockers assessed (uBlock Origin, Adblock Plus, Brave, DuckDuckGo).
This study came about in response to recent concerns over Chrome’s Manifest V3 proposal. One of the proposed changes threatens the functionality of ad blockers for Chrome and Chromium-based browsers that utilize the dynamic blocking capabilities of the webRequest API. If this change is implemented, developers will be required to use the declarativeNetRequest API—which uses static filtering and limits the number of filters to 30K. Google claims this API is optimal in terms of performance and security.
Cliqz’s study took up this first issue of «optimal performance» and found that «the most popular content-blockers are already very efficient… and should not result in any over-head noticeable by users.»
Furthermore, the study revealed that Ghostery outperforms the competition in terms of time per request, time to load the blocking engine, and memory consumption:
- All content-blockers except DuckDuckGo have sub-millisecond median decision time per request.
- Time to Process a Request in Ghostery (median): 0.007. This is faster than all other ad blockers considered (anywhere from 2.7x to 1258.4x faster!)
- Loading Ghostery’s Blocking Engine (from cache): 0.03 ms. This was faster than Brave, uBlock Origin
And Adblock Plus (anywhere from 368x to 3575x faster).
- Memory Consumption of Ghostery’s Blocking Engine (at startup, in Chrome): 1.8 MB. Here Ghostery uses less memory than uBlock Origin, Adblock Plus and DuckDuckGo’s ad blocker.
The full study with additional findings and analyses can be found here.