Myth: Supply path optimization just means buying through fewer SSPs

Myth: Supply path optimization just means buying through fewer SSPs

The popular shorthand for supply path optimization — choosing the most efficient route to a given impression — is "cut your SSP list down to three and you're done." Collapse the paths, save the fees.

That's the outcome of SPO done badly, not the method.

1. The same impression is often auctioned by multiple SSPs simultaneously (this is resold inventory). Each path has a different take rate, different latency, and different auction mechanics.

2. Naively cutting to "fewer SSPs" optimizes on logo count. Real SPO optimizes on path economics per impression: which route delivers a specific publisher's user at the lowest all-in cost and highest win probability.

3. Counter-example: the cheapest-fee SSP for a publisher may have a higher soft floor or lose more auctions to direct deals, so your effective cost through it is higher despite the lower take rate.

4. Done right, SPO is a per-publisher routing decision built from log-level data — bid request fingerprints, win/loss by path, latency, and observed clearing prices — not a global SSP whitelist.

You might keep eight SSPs and still run tight SPO, because each carries unique paths you can't reach otherwise.

Why it matters: optimizing for fewer partners optimizes the wrong variable. SPO is about the cheapest true path to each user, which sometimes means keeping a partner a logo-count purge would have killed.