Walk your own aisles for a minute. A screen above the shelf shows an ad. Ten feet away, on a shopper's phone, your app shows a search result for the same category.

One of those two decisions came from a relevance layer you built and can explain. The other came from a media schedule sold months ago, with no idea what that shopper searched for on their phone thirty seconds earlier.

That gap isn't a failure of your team. Onsite retail media has had years to standardize around a shared vocabulary, relevance, RTB, unified ranking, before you started asking harder questions of it. In-store hasn't had that runway yet.

The tools that promise to "connect onsite, offsite, and in-store" are telling the truth about the connection existing. They're just not always specific about what "in-store" means once you look underneath it. 

And here's the stance worth stating up front: you'll probably need both channels, so the real question isn't which one is legitimate. It's which one to build first, and what "connected" should actually cost you once you extend into the other. 

Onsite first, because it's the fastest place to prove a result, and because getting onsite and in-store to agree with each other doesn't require you to rip out either system and buy one all-in-one platform.

TL;DR: Onsite retail media runs on a relevance layer you already own and control, data you've had for years (search intent, inventory, margin), plus a logic for weighing relevance against sponsored bids that's becoming standard practice even though the industry hasn't fully converged on it. In-store retail media is earlier: most in-store demand is still governed by the screen or shelf vendor's own booking and measurement systems, only loosely tied to the relevance logic you already trust onsite. Build onsite first. When you extend into in-store, the fix isn't a bigger platform, it's applying the same relevance and margin logic you already trust to a new demand source.

‍

The result that actually surprises retailers here

The abstract version of this problem, "two systems don't talk to each other," doesn't cost anyone anything on its own. Here's the concrete version: a product your merchandising team deliberately deprioritizes onsite, because the margin doesn't justify the shelf space, can still win the best in-store screen slot next month, because the in-store media plan was sold before anyone checked. Nobody decided that on purpose. It's what happens by default when two systems make decisions about the same product without either one knowing what the other is doing.

That's the cost of treating "connected" as a planning convenience instead of confirming it's a shared decision. It's also exactly the kind of thing worth asking your in-store partner about before you scale the channel, not after.

‍

Why onsite and in-store aren't at the same stage, even though they're selling the same products

You're advertising the same products through both channels, often to the same shopper, sometimes on the same visit. That similarity makes it tempting to assume the two are equally mature. They aren't.

Your onsite ranking decides what a shopper sees using data you've owned for years, search intent, inventory position, margin, and a growing body of shared logic for how relevance and sponsored bids should be weighed against each other. Your in-store ranking, where it exists as an actual decision rather than a fixed media schedule, is newer and less standardized, mostly built by companies whose core business is the screen or the shelf tag, not the relevance logic behind what's on it.

None of that is a knock on your in-store media partners. Screens and shelf tags are a genuinely hard distribution problem to solve well. It's a different problem than the one your onsite relevance layer has already mostly solved, and treating the two as equivalent is how you end up with the margin mismatch above.

‍

What makes onsite structured, and what's still being argued about

Onsite retail media works because three things are true at once: you own the data (search intent, inventory position, margin), you control the page where the decision gets enforced, and there's an increasingly documented logic for weighing relevance against sponsored bids. That last part deserves a caveat: it's not something the whole industry has settled on, plenty of vendors are still converging on their own versions of it, but it's far more developed than anything in-store has today, and specific enough that you can hold a vendor's claims against it.

That head start is also why onsite is usually where the bulk of your retail media revenue already comes from, and why it's almost always the right channel to prove a relevance improvement in first. Proving a result onsite doesn't require anything outside your own systems to change.

‍

What in-store is still missing

In-store doesn't have that same shared vocabulary yet, and it shows up in three specific gaps:

Booking and measurement usually live with the screen or shelf vendor, not with your relevance layer. Most digital screen and shelf-media networks run their own systems for what plays where and how performance gets reported, systems built around media scheduling, not around the same relevance and margin logic you already trust onsite.

"Connected" often means visible in one dashboard, not decided by one logic. Orchestration tools that list onsite, offsite, and in-store as connected placements are usually describing a single planning interface, which is real and useful, not a single ranking decision applied consistently across all three. You can plan a campaign in one place and still have it ranked by three unrelated systems underneath.

There's no in-store equivalent yet of the demand connections onsite already has. Onsite retail media can point to specific, named RTB and API integrations, real-time bidding from programmatic and search sources, API links to advertiser platforms. In-store doesn't have a comparably standardized version of that today. Demand mostly still moves through direct media buys and screen-network sales relationships, not an open, real-time connection.

So what does that actually mean for you right now? It means treating in-store as a demand source you evaluate, not a ranking decision you trust by default, the same way you'd want proof before trusting any new onsite demand partner.

‍

The commercial question: do you need one platform, or one shared logic?

This is the part worth deciding before you sign anything. When a vendor pitches "onsite, offsite, and in-store, all connected," the pitch usually implies you need to buy their platform to get that alignment. You don't.

What you actually need is for in-store visibility to be decided by the same relevance and margin logic that already governs your onsite ranking, not a separate scoring system bolted on next to it. That's an argument for extending the logic you already trust into a new demand source, using whichever best-of-breed screen or shelf partner does that job well, rather than replacing your whole stack with a single vendor's version of "all-in-one." A shared decision beats a shared dashboard, and you can get a shared decision without consolidating every vendor relationship into one contract.

Ask a vendor to be specific here. "Connected" claims survive a real conversation only if the vendor can answer these without reaching for the word "platform":

• What specific data decides what appears on an in-store screen or shelf tag, and where does that data come from? If the honest answer is "the media plan," that's a booking system, not a relevance decision.
• Is in-store visibility tied to the same margin and merchandising priorities that govern your onsite ranking, or is it a separate system entirely? A separate system means you're running two sets of rules for the same products.
• What would it take to extend your existing onsite relevance logic to this in-store partner, instead of adopting a new scoring system? If there's no clear answer, the in-store connection is likely still standalone.
• Does getting these two channels to agree require replacing your ad server, your search platform, or your in-store partner, or does it just require a connection between the systems you already have? The second answer scales. The first is a rebuild wearing an "in-store" label.

‍

What to do about it now

None of this means you should ignore in-store, or wait for the ecosystem to standardize before doing anything. It means sequencing correctly:

Prove onsite first. It's the fastest, lowest-risk place to show a relevance improvement, because nothing outside your own systems has to change.

Treat in-store as a demand source to evaluate, not a ranking decision to trust. Ask what data actually decides in-store placement, and whether it connects to your onsite margin logic at all.

Use orchestration tools like Vantage for what they're genuinely good at today, planning and reporting across channels in one workspace, without assuming that also means unified ranking underneath.

Watch for yourself extending your own onsite relevance logic outward, rather than adopting an in-store vendor's separate scoring system as if it were equivalent. That's the same pattern that made unified ranking work onsite in the first place, and it doesn't require consolidating vendors to do it.

‍

Key takeaways

• Onsite and in-store are advertising the same products but sit at very different stages of maturity. Onsite has years of increasingly shared logic behind its relevance decisions; in-store mostly doesn't yet.
• The real cost of skipping this distinction isn't abstract: a product you deprioritize onsite for margin reasons can still win the best in-store slot, because the two systems never compared notes.
• Most digital screen and shelf-media networks still run their own booking and measurement systems, loosely tied at best to the relevance logic that already governs the same products online.
• "Connected" in most orchestration tool pitches means a shared planning workspace, not a shared ranking decision. Ask a vendor directly which one they mean.
• You don't need to buy an all-in-one platform to get onsite and in-store working under the same logic. You need the logic you already trust onsite extended to whichever in-store partner does the screen or shelf job well.

Frequently Asked Questions

Should I prioritize onsite or in-store retail media first?

Onsite, in almost every case, not because in-store doesn't matter, but because onsite is the channel with the most standardized relevance logic, the most direct data ownership, and the fastest path to a provable result. You'll likely need both eventually. The question is sequencing, not choosing one over the other permanently.

‍

If an orchestration tool says it connects onsite, offsite, and in-store, what does that actually mean?

Usually that you can plan and report on all three from one workspace. It doesn't automatically mean a product's visibility across those channels is decided by the same relevance and margin logic. Ask the vendor to be specific about which one they're offering.

‍

Do I need to buy an all-in-one platform to make onsite and in-store agree with each other?

No. What you need is your existing onsite relevance and margin logic extended to your in-store demand, not a single vendor's bundled system replacing everything you already have. Best-of-breed screen and shelf partners can plug into that logic without you consolidating every contract into one platform.

‍

Does in-store retail media have the same kind of RTB or API demand connections that onsite has?

Not to the same degree. Onsite retail media has named, standardized real-time bidding and API integrations with major demand sources. In-store demand today mostly still moves through direct media buys and screen-network sales relationships rather than an open, real-time connection.

‍

Should I wait to build in-store retail media until the ecosystem matures?

Not necessarily. It means evaluating in-store partners as demand sources to question carefully rather than ranking decisions to trust automatically, and using orchestration tools for the planning and reporting value they already provide today.

‍

Get in touch with Pentaleap to learn more.

‍