How Many Facebook Pixels Should I Have?

The standard advice is one pixel

Almost every guide gives the same answer: run a single pixel, because more data means faster learning. That is right as far as it goes.

It stops being right when your offers do not share a buyer, and it ignores what happens to that pixel when the account holding it goes down.

One per niche, not one per product

Per product is too many. Each pixel then learns slowly, and none of them collect enough events to optimise properly.

Per niche is the unit we run. Offers that share a buyer share a pixel:

• Clothing and outdoor clothing on one pixel, beachwear on its own, shoes on another
• In jewellery, every bracelet offer on one bracelet pixel

The test is simple. If someone who bought one offer is a plausible buyer of the other, they belong on the same pixel.

A pixel does not survive the Business Manager that owns it

This is the part the pixel guides leave out. A pixel cannot be moved between Business Managers, and sharing it out grants access rather than ownership.

If the Business Manager holding your pixel is restricted, the history goes with it. Everything running on that one pixel loses its data at the same moment.

That is the real argument against putting every offer on a single pixel.

Keep the pixels away from the ads

The structure we use is a host Business Manager that owns the pixels and never runs ads, sharing them out to the Business Managers that do.

A host BM has to be warmed first. A new Business Manager cannot share a pixel out.

Our Scale plan is built around this, with a host BM for pixels alongside four Business Managers that run the ads. You keep that host BM, and its pixel history, even if you stop using us.