Your Brand Made a POP Concept With AI. How Do You Get It Built?

A brand marketing manager sends a POP manufacturer an AI image of a counter display. It looks finished. Clean acrylic, the product lit from below, the logo exactly where the brand guidelines want it. The email says "can you quote this, 180 stores, November". And the manufacturer, who has seen this a lot lately, starts by redrawing it.

That is where most AI concepts in POP stall today. The picture did its job, the look is approved, and nobody can price it yet.

I have spent close to four years in point-of-purchase, on the manufacturer side, and this post is for the brand side of that email: the trade, shopper and retail marketing teams, and the agencies working for them, who now make concepts with AI and then have to get them built. What sits between the image and a number, why every manufacturer reads the same image differently, and how Bellto, an AI agent for POP campaigns from brief to shop drawings, closes the gap.

How common is it to arrive at a manufacturer with an AI concept?

Common enough that manufacturers bring it up without being asked. In calls with US POP manufacturers in August 2026, one told me that 4 of about 48 proposals in the last six months had started from a client's AI reference, "and more and more". Another said 5 of about 30 orders in the last month did.

Those are two shops, not a market statistic, and I cite them only as what those two told me. What they describe matches what I see. Brands no longer arrive with a paragraph and a competitor photo. They arrive with a picture of the thing they want, made in an afternoon, and the conversation starts from there.

That is a real improvement. A brand that can show the look it wants saves everyone a round of guessing. It just moves the hard part rather than removing it.

What does an AI concept image actually settle?

It settles the look, and the look matters. A concept tells the brand team and the retailer what the shopper will see: the format, the material family, the finish, where the brand sits, how the product is presented. Getting that agreed internally is often the slowest political step in a campaign, and a good concept image gets it done in one meeting.

What it does not carry is anything a workshop measures. A raster image has no millimeters in it. The proportions can be completely believable and still be wrong for the product, because nothing in the image knows how deep the pack is, how many facings the brand wants on each shelf, or what the retailer's counter space actually is.

So the image answers "do we like this". The manufacturer's question is a different one: "what exactly am I pricing".

What sits between the picture and something a workshop can price?

Six things, and every one of them is a decision somebody has to make before a quote means anything.

Real dimensions, derived from the product. Width, depth and height come from the pack dimensions and the facings the brand wants, not from what looked right in the picture. A shelf sized by eye either wastes material or does not fit the product.

Material and thickness per part. "Acrylic" is not a specification. A back panel, a base and a product rest in the same display can each be a different thickness and a different sheet, and the cost follows those choices part by part.

Joints. Solvent bond, UV bond, screws, slots. The joint changes the parts themselves, their tolerances and their assembly time, and none of it is visible in a render.

Structural checks. Parts that overlap in the model collide on the bench. Somebody has to confirm that nothing interferes and that the piece holds what it was designed to hold.

Dimensioned drawings. Front, side and top views with measurements on them, so the shop can read the piece without guessing what is behind what.

Cut files. The contour of every manufactured part, closed and ready for the cutting table.

When a brand sends only the image, the manufacturer does all six before quoting, usually for free, because that is the cost of being allowed to bid. I went through the manufacturer's side of this step by step in another post. This one stays on the brand's side of the desk.

Why does every manufacturer read the same image differently?

Because each of those six decisions is open, and each shop closes it its own way.

Send one AI concept to three manufacturers and you do not get three prices for one display. You get three different displays, each with a price. One assumes thicker sheet and a bonded joint. One sizes the shelves to the packs and another to the picture. One adds a reinforcement it thinks the structure needs and another leaves it out. Every one of those readings is reasonable, and every one of them moves the number.

The brand then compares quotes that are not comparable, picks one, and discovers the differences at the sample stage, when changing anything is expensive. The picture was never the problem. The problem is that the picture was the specification.

If your team is comparing quotes on the same concept right now and the numbers do not line up, the Bellto waitlist is where that campaign should go.

How does Bellto turn a concept into a package every manufacturer quotes the same way?

Bellto is an AI agent, and it works from the campaign rather than from the image. You tell it what you are launching: brand, product, channel, stores, budget, date. It asks for what is missing before proposing anything. It proposes the campaign mix and the materials, floor stand, glorifier, shelf strip, stopper, with the budget split per store.

Then it designs every piece with you at real scale in parametric 3D, with proportions derived from the product dimensions and the facings. That is the step where the six open decisions get closed, once, in the model, instead of three times in three workshops. A dimension change recomputes the whole piece and re-runs the manufacturing checks in seconds, so asking for a taller header or one more facing does not mean waiting on a revised proposal.

Every approved piece comes out in five formats: a 3D model, a part-by-part cutlist, dimensioned drawings, a STEP of the assembly and a cut DXF per part. That is a package a workshop can quote without redrawing, and every model goes through engineering review before delivery. Bellto then suggests manufacturers that fit by material, format and volume. The quote comes from the manufacturer, and now every manufacturer is quoting the same piece.

Your AI concept still has a place in that flow. It is the look the piece is designed against. If you do not have one yet, AI POP Displays is where the concept image comes from.

What does that package look like on a real piece?

The Bellto landing page publishes the files for a worked example rather than describing them.

It is a four-panel acrylic counter glorifier with an envelope of 200 × 200 × 250 mm. Four parts and four cut DXF files, one per part. Across its 6 pairs of parts the measured overlap is 0.0 mm³, which is the structural check that nothing collides. The STEP was reimported with no solids lost, and the drawings carry hidden lines on their own layer. After a dimension change, the whole piece recomputes in under a second.

At floor stand scale, the example is a four-shelf unit of 1000 × 650 × 1820 mm with a header and casters: 241 parts and 250 cutlist rows, 89 manufactured and 152 bought in. Going from 16 to 20 facings recomputed the entire unit from the parametric geometry alone in 1 min 18 s, with the checks green afterwards.

Put those next to the email at the top of this post. The glorifier in that email is the same kind of piece. The difference is whether it arrives at the manufacturer as a picture or as four cut files, a cutlist and a set of drawings. I covered each of the five formats in more detail in what Bellto is, and the wider question of where a concept render stops in can AI design a POP display.

What changes for the brand?

Three things.

The quotes become comparable. When every manufacturer prices the same dimensioned piece, the difference between two numbers is the difference between two shops, not between two interpretations.

The changes get cheap. A new facing count or a different height is a recompute with the checks re-run, not an email and a week.

And the brand keeps the design. The package belongs to the campaign, not to whichever workshop did the redrawing, so moving a piece to another manufacturer or another market does not start the work over.

If the next launch on your calendar is one you would rather send out as a package than as a picture, put it on the Bellto waitlist.

None of that replaces the manufacturer. The shop still quotes, samples and builds, and the manufacturers I talk to would rather receive a package than a picture, because the free redrawing is the part of the job they like least. POP manufacturers can sign up to Bellto too.

How do I get a campaign into Bellto?

Through the waitlist. We are building Bellto now, the list is open to any brand or agency, and we are contacting the first ones soon to run the first real campaigns end to end. It opens by invitation, in small groups, and the campaign you describe when you sign up sets your place, so a concrete launch with a retailer, a store count and a date moves faster than general interest. Pricing goes first to the people on the list, and there is no card.

If you have an AI concept on your desk that nobody can quote yet, tell us about that campaign on the Bellto waitlist.