Parametric POP Display Design: Change the Facings, the Whole Stand Recomputes

A brand marketing manager approves a floor stand in October, and in the first week of November the category team comes back wanting four more facings on each shelf. On a normal campaign that email costs days. Somebody at the manufacturer redraws the shelf, then discovers the uprights no longer sit where they did, then the header overhangs, then the drawings and the cut files have to be exported again, and the structural check that was done on the first version quietly does not get re-run on the second.

I have spent close to four years in point-of-purchase, and that sequence is the one I have watched swallow the last two weeks of more campaigns than I can count. The change itself is trivial. What makes it expensive is that the display was drawn rather than derived. This post is about the alternative, which is a display whose geometry comes out of the product and the facings, and what actually happens when you move one of those numbers.

What does parametric actually mean for a display?

A parametric model is one where the geometry is computed from a set of inputs instead of being drawn by hand and then measured. Move an input and everything downstream of it moves with it, because the relationships are part of the model rather than something a person remembers to update.

For a point-of-purchase display, the inputs that decide the piece are not abstract. They are the dimensions of the pack going on it, how many of those packs sit side by side on a shelf, how many shelves there are, how much clearance the shopper needs above each one to lift a unit out, what material every part is made of and how thick it is. Those numbers decide the shelf width, the shelf pitch, the height of the uprights, the footprint, the header position and the load the base carries.

Draw the display first and those numbers become something you check afterwards. Derive the display from them and the checking is the drawing.

Bellto, an AI agent for POP campaigns that takes a launch from brief to shop drawings, designs every piece this way. Proportions come from the product dimensions and the facings, at real scale, from the first concept rather than after an engineering pass.

Why doesn't the word already mean this in display CAD?

Because in packaging CAD, parametric has settled into a narrower and older meaning, and the vendors are straightforward about what it is.

Esko's ArtiosCAD Display Store, which is the page that ranks for almost every parametric display search, describes itself as hundreds of resizable, production ready POP designs, and says ArtiosCAD makes it easy to resize the parametric POP designs. It is aimed at sign and display converters who want to move into display production without building structural expertise first, and it has been published in that form since 2016. ArtiosCAD itself is very much current, on release 26.07 from 7 July 2026.

EngView sells the same idea with more of it exposed. Its Package and Display Designer shipped a 2026 version on 30 January, its subscription pages list parametric drafting and more than 2,500 parametric templates, and it holds searchable libraries of parametric multipart designs for corrugated and rigid board. Fill in the parameters and every drawing in a multipart structure resizes together, including a recalculation when the material thickness changes.

Both are real, both do what they say, and a structural designer working in a board plant has no reason to give up either of them. What they parameterise is the display. Width, height, shelf count, board thickness. The product that the display exists to hold is not one of the parameters, and the word facings does not appear in that vocabulary at all. I went looking for anyone on the open web who derives display geometry from the SKU and the facing count and found one in-house patent and nothing readable. The merchandising word and the CAD word have never been on the same page.

Outside packaging the term belongs to architecture, where parametric means Rhino and Grasshopper and rule-based geometry for facades. Also real, also not a display.

If a campaign on your desk is heading for a resize that nobody can price, the Bellto waitlist is where I would put it.

What changes when the product drives the geometry?

The first thing that changes is that the piece stops being plausible and starts being correct.

A display drawn by eye can look completely convincing and still be wrong for the product, because nothing in the drawing knows how deep the pack is. The shelf ends up 30 mm too shallow and the front face of the product sits proud of the edge, or it ends up 40 mm too deep and the brand pays for material that holds air across 180 stores. Neither shows up in a picture. Both show up at the sample.

The second thing is that the argument with the category team becomes answerable. When somebody asks whether five facings fit instead of four, that is a question with an arithmetic answer rather than a design opinion, and the model can answer it in the meeting.

The third is that the manufacturing package stops drifting. The cutlist, the dimensioned drawings, the STEP of the assembly and the cut DXF files all come out of the same geometry, so they cannot disagree with each other. That is the part I went through in detail in what a shop drawing package for a POP display actually contains, and the reason five files assembled separately go wrong is exactly that each one has its own last-updated date.

What happens when the facings change from 16 to 20?

Here is the worked case, and it is published rather than described, which is the only version of this claim worth making.

The piece is a four-shelf floor stand, 1000 by 650 by 1820 mm, with a header and casters. It comes to 241 parts and 250 cutlist rows, 89 of them manufactured and 152 bought in, each row carrying its material, profile and joint. Going from 16 facings to 20 recomputed the entire unit from the parametric geometry alone in 1 min 18 s, and the manufacturing checks came back green afterwards.

Four more facings is not a small change on a piece like that. The shelves widen, so the uprights move, so the base footprint changes, so the header spans further, so the load path through the casters changes, so 89 manufactured parts have new contours and 152 bought-in parts may need different quantities or lengths. Every one of those consequences is a line in the cutlist and a file on the cutting table.

At the other end of the size range, the published counter piece is a four-panel acrylic glorifier with an exact envelope of 200 by 200 by 250 mm, four parts and four cut DXF files. There, a whole-piece recompute after a dimension change lands in under a second.

Set that against the version of this that a brand lives through today, which is an email to the manufacturer and several days of waiting for a revised proposal that may or may not have had anything re-checked.

Why does the recompute have to re-run the checks?

Because a fast wrong answer is worse than a slow one, and the failure mode of parametric geometry is specific. Push a model far enough from the shape it was built around and parts start to interfere, contours stop closing, and a joint that worked at one thickness stops working at another. None of that is visible in a render of the new version.

So the checks run on the model, on every recompute. On the acrylic glorifier the published numbers are 0.0 mm³ of overlap measured as volume across the six pairs of parts, a STEP that reimports with no solids lost, four closed cut contours, and drawings that keep hidden lines on their own layer, 13 visible and 55 hidden edges on the front elevation. Those are geometric facts about the model rather than opinions about the design, which is why they can be re-run automatically. Every model still goes through engineering review before it is delivered.

That combination, a change that takes seconds and checks that re-run themselves, is what makes a late facing change a conversation instead of a crisis.

Where the concept image fits

None of this replaces the picture. A campaign still has to get a look approved internally before anything gets engineered, and that is what a concept render is for. AI POP Displays is where the concept image in a POP campaign comes from, with the format, the material and the product load right, and the piece is then designed against it. What changes is the render's job. It stops being the thing sent out to quote and becomes the reference the model is built to, which is the shift I wrote about in how a brand gets an AI concept built.

What we're building next: Bellto

Bellto starts from the campaign rather than from the piece. You tell it what you are launching, brand, product, channel, stores, budget and date, and it asks for whatever is missing before it proposes anything. The landing page shows the illustrative version of that exchange, a premium cutlery range going into 180 stores on a 300,000 € budget, where the agent comes back with 1,667 € per store and two questions it needs answered first. It proposes the campaign mix and the materials, floor stand, glorifier, shelf strip and stopper, with the budget split per store.

Then it designs every piece with you at real scale in parametric 3D, which is the part this whole post is about. 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, a package a workshop can quote without redrawing. It suggests manufacturers that fit by material, format and volume, and the quote comes from the manufacturer. The full reference page is what Bellto is.

Brands are already arriving at manufacturers with a picture. In calls with US POP manufacturers in August 2026, one told me that 4 of roughly 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. Two shops, not a market statistic, and what they describe is the same gap from the other side of the desk.

We are building it now. The waitlist is open to any brand or agency, and POP manufacturers can sign up too. 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 the next campaign on your calendar is one where somebody is going to ask for more facings in week three, tell us about it on the Bellto waitlist.