Introduction
A pay-per-sale CSS partner sits in an unusual position: it bills on commission through the affiliate team and competes in the Google Shopping auction the PPC team manages. Most organisations have a clear owner for each side and no clear owner for where they meet. This article covers why that gap exists, what it costs the channel, and how to set up both teams to run it as one.
Pay-per-sale CSS touches affiliate and PPC teams
If you are weighing up a CSS partner, one hesitation tends to come up before the commercial terms do: which team owns it. It runs in Google Shopping, where the PPC team works. It can bill on commission through affiliate networks, where the affiliate team works. The concern is reasonable, and it is worth addressing directly, because the answer decides whether the channel grows or stalls.
First, a distinction that the rest of this rests on, because "CSS partner" covers two different products that land with two different teams.
One is Premium CSS. You keep running your own Shopping campaigns and switch the CSS behind your Merchant Center so your bids carry more weight in the auction. It stays with the PPC team, because the campaigns are still theirs. There is no second team involved, so there is nothing to align.
The other is a pay-per-sale CSS partner running managed campaigns. The partner funds the bidding and bills on commission. It enters the business through the affiliate team, gets approved like a publisher, and is managed as a commercial relationship. But it’s often perceived as a channel that competes for the same impressions, even though auction deduplications prevents this from happening.
That pay-per-sale CSS model is the one that touches two teams, and it is the one this article is about. Affiliate owns the commercial side. PPC owns the auction it appears in. Most companies have a clear owner for each, and no clear owner for where they meet.
It usually surfaces with a routine approval. The affiliate manager signs off on the partner the way they would any publisher: a CPA deal, a commission rate, a contract. Weeks later the PPC manager finds it in the Shopping carousel, bidding in the auction they manage, and reasonably wants to know what it is and how it relates to their own results. Both are doing their jobs. The overlap is structural. It comes from a channel that bills like affiliate but competes like Shopping, and runs across two teams the organisation set up to work apart.
Most companies are already at this point: the partner is live, and the way the teams work around it is still catching up. What follows is how that tends to run today, why it holds the channel back, and how to set it up so the two teams reinforce each other instead of working from half a picture each.
PPC and affiliate are different teams with the same goal
PPC and affiliate share the same commercial goal: efficient sales growth against a single revenue target. What differs is the execution path, and that difference comes from economics.
PPC pays for clicks. The team commits media spend up front and carries the risk that a click might not convert, so the work is continuous and hands-on: bids, impression share, feed quality, and query-level performance, all managed against a ROAS the team controls in real time.
Affiliate is built on conversions. Nothing is owed until a sale completes, so the performance risk sits with the publisher, and the team's work is commercial: approving partners, agreeing commissions, and managing a portfolio of relationships. The day-to-day optimisation belongs to the partners. The team sets the terms, reads the results once they settle, and adjusts the strategy.
That independence is what lets each discipline operate at full strength, pushing its own craft as far as it will go with no second model to accommodate. It held for a structural reason: PPC and affiliate rarely acted on the same inventory. What one team did in the search auction did little to move a result in the other team's base of coupon, cashback, and content partners. Coordination stayed limited, and the structure most companies still run is a fair map of how the channels actually work.
A pay-per-sale CSS partner changes that. For the first time, the two disciplines are working on the same inventory.
Both teams now work on the same inventory
A pay-per-sale CSS partner promotes the merchant's own products in Google Shopping, on the merchant's behalf. This is the managed model, not Premium CSS: the partner funds the bidding, and the merchant pays commission only when a sale completes. No media spend, no CPC for the merchant to carry. On the surface, it bills exactly like affiliate.
That commercial surface sits on a specific technical setup. The partner enters Google Shopping through a separate Merchant Center ID, a second entry point into the same inventory.
Because the offers still belong to the same merchant, Google does not treat this as two advertisers competing. When the merchant's own campaigns and the partner's campaigns bid on the same offer, Google deduplicates at merchant level and only one bid moves forward. The merchant is never priced against itself, so adding a partner does not stack bids or push up its own CPC. We covered the auction mechanics in full here: Will adding a CSS partner cannibalise my Google Shopping campaigns?
The value is reach. The partner captures the queries, placements, products, and demand pockets the merchant's own campaigns were not winning, and that is where the incremental sales come from.
What it does not come with is a natural owner. The partner bills through affiliate but performs in PPC's auction, so several teams touch it and none owns how its commercial and performance sides connect. That gap is where the cost starts.
You can’t grow a channel if you have poor visibility
The channel produces two kinds of data, and they sit in two places. Affiliate sees the sales and the commission. PPC sees the Shopping impact: where it shows, what it covers, where it overlaps. Neither team sees the whole channel, and that is the entire problem.
Working from half a picture, the PPC team sets bids and budgets without knowing how much of the merchant's Shopping presence sits outside its own account, so the coverage gaps the partner was brought in to fill can stay open. And when someone proposes growing the channel, the case will not assemble: deciding to scale needs a single view of cost, coverage, and incremental contribution, and that view does not exist when the numbers live in separate reports.
Budget time is where this lands hardest. The channel may be delivering real sales, but if its cost, overlap, and incremental contribution are read through two systems that were never joined up, the case for more budget is hard to prove. So the channel holds at its current size, not because it stopped working, but because no one could see enough of it to back it.
