If you've spent years tuning pinned products, category rules, and seasonal boosts, the last thing you want is an ad system rewriting your page. We get it.

Handing ranking to one decision layer gives merchandising more control, not less. Today your rules live in the search engine, while the ad server fills sponsored tiles with its own logic. With one layer, a single system enforces every rule you've written, across sponsored and organic products alike. One thing does change, and this piece is specific about what.

TL;DR

• Under unified ranking, pinned products, boosts, category rules, fixed ad units, and static tiles can be reproduced one-to-one from day one. Pentaleap describes these as enforced through the same layer that places sponsored products.
• A pin is a rule, and rules are enforced. Sponsored products compete for the positions your rules don't govern.
• What merchandising gives up is the default: unruled positions stop being organic automatically and are decided on combined retail and ad margin instead.
• In a legacy setup, changing the number of sponsored slots usually needs engineering. In Pentaleap's model, ad inventory scales up or down through the UI.
• You can start by mirroring today's page exactly and move toward dynamic placements on your own timeline.
• Before agreeing, merchandising should ask a vendor to demonstrate six specific things, starting with a one-to-one reproduction of your current page.

‍

What happens to pinned products, boosts, and category rules under unified ranking?

They stay, and they're enforced by the same layer that decides where sponsored products go.

Pentaleap's public description is specific on this point: any layout running today can be reproduced one-to-one from day one, including fixed ad units, static tiles, merchandising overrides, and merchant rules, all enforced through the same layer (Pentaleap, "Criteo, Koddi & others are championing a new model," May 2026). The same piece lists retailer rules and product availability among the inputs to the single ranking decision, alongside relevance, predicted click-through and conversion, and margin.

The bigger change is where relevance comes from. In a legacy setup, the ad server ranks sponsored products with its own algorithm, separate from the search engine your team has spent years tuning. In a unified ranking model, sponsored products are ranked using the intelligence of your existing search and personalization system, such as Google Vertex AI or Algolia.

That has a practical consequence merchandising teams tend to like. When your search team improves relevance, the improvement now applies to sponsored products too, without a separate fix on the advertising side (why retail media and merchandising keep fighting each other).

‍

Can a sponsored product ever outrank a product we've pinned?

A pin is a merchandising override, and Pentaleap describes merchandising overrides as enforced by the ranking layer itself. So a pinned product holds its position.

Where a sponsored product can outrank an organic one is in the positions no rule governs. There, every product competes on what Pentaleap calls compound margin: retail margin plus ad margin, computed together. A highly relevant sponsored product with a strong bid can win one of those positions. So can an organic product with strong retail margin, which can outrank a mediocre ad (Pentaleap, "Why Retail Media and Merchandising Keep Fighting Each Other," July 2026).

That second half matters as much as the first. Under bid-led ad serving, a sponsored product fills its tile whether or not it deserves to. Under unified ranking, a weak sponsored product loses to a strong organic one.

This is still a claim worth testing rather than taking on trust, from us or any other vendor. The demonstration list at the end of this piece starts with exactly that.

‍

So what does merchandising actually give up?

Merchandising gives up the default, not the rules.

Today, any position your team hasn't explicitly pinned or boosted is organic by default, ranked by your search engine, and the ad server stays inside its fixed tiles. Under unified ranking, those unruled positions are decided on combined margin across retail and media. A sponsored product can now outrank an organic product your team would have placed higher. We've been direct about this elsewhere: it is a real loss of control over page real estate that ecommerce teams currently own outright, and pretending otherwise in a pitch is what makes teams feel blindsided later (getting ecommerce, search and finance to agree on a ranking change).

Two things are worth setting against that loss.

First, the positions your team cares most about can still be pinned. Choosing which positions to rule becomes the most important merchandising decision on the page, and it stays yours.

Second, compare it with what you control today. In most legacy setups, the sponsored tiles already belong to the ad server, filled by logic your team can't tune and often can't see. Under unified ranking, the competition for unruled positions runs on a calculation that includes your retail margin and your search relevance. You trade positions-by-default for a decision you can inspect.

‍

Who decides how many sponsored products appear on a page?

The retailer does, and under unified ranking that decision moves out of the engineering queue.

In the legacy model, sponsored products are placed into a limited number of fixed, pre-reserved tiles, and scaling that number up or down requires engineering work. In Pentaleap's model, ad inventory can be scaled up and down through the UI (Pentaleap, Why Unified Ranking). For a merchandising team, that changes the practical question from "can we change this?" to "who on our side is allowed to?"

That's a governance decision worth making before go-live rather than after. Agree which team can change sponsored density on which pages, and what has to be true before they do.

It's also worth separating density from performance. A U.S. department store working with Pentaleap added no ad slots and still saw ad revenue rise 78% in its initial A/B test and 23% after scaling across placements and devices. The lift came from a ranking change, with high-relevance sponsored products moving up the grid and less relevant ones moving down (a department store case study with the same ad load). More slots and better-ranked slots are different levers, and only one of them asks merchandising to give up page space.

‍

Can we start by mirroring today's page exactly?

Yes. Pentaleap's stated approach is to start static, mirror your existing page exactly, and move toward dynamic placements gradually, on your own timeline (Pentaleap, "Criteo, Koddi & others are championing a new model," May 2026).

Starting static doesn't mean starting with nothing. Even when sponsored products stay in the same fixed positions, the logic for which sponsored product fills each position can change from bid-first to relevance-first. A relevant product buried deep in the organic results can be pulled forward into a visible slot instead of the highest bidder taking it automatically. We've explained that mechanism in how injected ad slots decide which sponsored product wins. Most retailers keep strict positioning rules for sponsored products, and that piece covers why the mechanism works within them.

The other half of starting safely is measuring the change on a limited share of traffic before it touches the whole site. A holdout that tracks organic conversion on the same pages, not just sponsored clicks, tells merchandising whether its pages got better or worse (testing a ranking change on a slice of traffic). Pentaleap reports that implementing its unified ranking layer typically takes about three weeks (Pentaleap, Omnichannel Orchestration).

‍

What should merchandising ask a vendor to demonstrate before agreeing?

Ask to see it, not hear it. These six demonstrations separate a vendor that enforces your rules from one that promises to:

1. Reproduce our current page one-to-one. Every pin, boost, fixed ad unit, and static tile, on a real category page, before anything else changes.
2. Show what happens to a pinned product when a high bid targets its position. The pin should hold. If it doesn't, you've learned what "enforced" means to that vendor.
3. Show where each rule lives and who can edit it. Rules split across two systems, with no single view of both, are how pages drift.
4. Change the number of sponsored slots on one category page during the demo. Through the UI, without an engineering ticket.
5. Show where relevance comes from. Is sponsored ranking using our search engine's relevance, or the vendor's own model? If it's the vendor's, ask how our search team's tuning reaches it.
6. Show how to roll back. If the test result disappoints, what does it take to return to today's page, and does anything else in our stack change?

For the broader check on whether a vendor's unified ranking is really one decision rather than two systems in sequence, we've also published five questions to ask a vendor that says it ranks sponsored and organic together.

‍

Three Diagnostic Questions

#1 Which of your merchandising rules live in your search engine today, and which live in your ad server?

#2 When your ad system and your search engine disagree about a product's position, who wins right now?

#3 Could your team change the number of sponsored slots on a category page this week without an engineering ticket?

‍

Key Takeaways

• Unified ranking doesn't remove merchandising rules. Pentaleap describes pins, overrides, merchant rules, fixed ad units, and static tiles as reproducible one-to-one and enforced through the same layer.
• A pinned product holds its position. Sponsored products compete for the positions no rule governs.
• What merchandising gives up is the default. Unruled positions are decided on compound margin, retail margin plus ad margin, rather than being organic automatically.
• Sponsored ranking uses your existing search relevance, so your team's tuning applies to sponsored products too.
• Ad inventory scales through the UI rather than through engineering, which makes "who is allowed to change it" a decision to settle before go-live.
• Start by mirroring today's page, test on a slice of traffic, and ask the vendor to demonstrate rule enforcement live before agreeing.

‍

Frequently Asked Questions

Does unified ranking override merchandising rules?

No. In the model Pentaleap describes, merchandising overrides, merchant rules, fixed ad units, and static tiles are enforced through the same layer that places sponsored products. Any layout running today can be reproduced one-to-one from day one.

Can a sponsored product outrank a pinned product under unified ranking?

A pin is a merchandising override, which Pentaleap describes as enforced by the ranking layer, so the pinned product holds its position. Sponsored products compete for positions that no rule governs, where they're ranked on retail margin plus ad margin alongside organic products.

Who controls how many sponsored products appear on a page?

The retailer. In legacy setups, changing the number of fixed sponsored tiles typically requires engineering work. In Pentaleap's unified ranking model, ad inventory can be scaled up or down through the UI.

Can we keep our current page layout when we move to unified ranking?

Yes. Pentaleap's approach is to start static, mirror the existing page exactly, and move toward dynamic placements gradually. Even with fixed ad positions, which sponsored product fills each position can shift from bid-first to relevance-first.

What does merchandising lose with unified ranking?

Merchandising loses the default that every unruled position is organic. Those positions are decided on combined retail and ad margin, so a sponsored product can outrank an organic product the team would have placed higher. Positions the team wants to protect can still be governed by rules.

What should we ask a retail media vendor to demonstrate about merchandising control?

Ask the vendor to reproduce your current page one-to-one, show a pinned product holding its position against a high bid, show where each rule lives and who can edit it, change sponsored slot counts through the UI, explain where sponsored relevance comes from, and walk through a rollback.

‍