Does Criteo slow down a store?
In 7IT’s test of 367 US online stores, Criteo ran on 26 of them (7.1%). On the median store that loads it, it downloaded 32 KB and kept the phone busy for 0.06 s before the page could respond.
7.1%of 367 US stores run it
32 KBmedian download on stores that load it
0.06 smedian time the phone spends running it
Stores with it and without it.
| Stores | Median score | Main content | Blocking time | Apps on the page | |
|---|---|---|---|---|---|
| With Criteo | 26 | 31 | 10.3 s | 2337 ms | 32 |
| Without | 341 | 35 | 11.8 s | 1157 ms | 19 |
Stores that run Criteo had a median score of 31, against 35 for stores without it. They also run more apps overall (a median of 32 against 19), so that gap is not all Criteo’s doing. The size and phone time above are Criteo’s own. On this sample there is no score penalty that can be pinned on it.
One in four stores that load it download more than 32 KB of it, and spend more than 0.09 s running it. Median Google phone-lab results, home pages only, September 2026. Method and limits in the study.
If you run Criteo.
- Ad pixels count visits; they do not need to block the visit. Load them after the first interaction, or move them server-side (Conversions API, server-side tagging) so the phone does less.
- Check every pixel still feeds a campaign that is running. Pixels outlive campaigns.
- One container for all pixels, loaded late, beats many tags each loading early.