Generating in-store creative from Shopify data

On Shopify, changing what a customer sees takes seconds. Swap imagery, update pricing, push a promotion live. It’s all part of the same system, and it all happens in real time. On the shop floor however, that level of immediacy doesn’t really exist.

The marketing collateral that customers see in-store is still tied to assets that have to be designed, exported, printed and rolled out. Even small changes can carry sizable overhead, so updates happen less often, and creative signage tends to stay fixed for longer than it should.

That not only creates a gap between the online and in-store experience, but also means in-store creative doesn’t reflect what’s in stock, what’s being pushed, or what’s performing. What’s more, it eats up a lot of time.

Replacing assets with a system

Some retailers have already moved to digital screens to speed up how in-store signage is managed. That’s a step forward. Content can be updated without reprinting materials or manually swapping displays, but those screens still need to be fed.

Every change still depends on producing new imagery, which usually means briefing an agency, working through revisions, exporting assets, and uploading them through a separate system. Even with faster distribution, creatives are still slow to produce.

We’ve been working on a way to remove that step entirely.

Broadcast, our latest Shopify app, is built as a Shopify-native layer that sits directly on top of the Admin API and Storefront API. It pulls product data, variants, pricing, and metafields in real time, and uses that as the input for creative generation.

Image generation is handled through Google’s Gemini models (Nano Banana), which lets us combine structured product data with prompts and reference inputs. That means we can generate context-aware visuals - not just generic outputs - using the same data that powers the storefront.

On top of that, we’ve built a composition layer that applies layout rules, brand constraints, and aspect ratios automatically. This avoids the need for manual design passes and keeps outputs consistent across formats.

Once generated, assets are cached and versioned, then delivered to in-store screens via a lightweight rendering layer. Updates can be triggered by changes in Shopify data, scheduled campaigns, or manual inputs.

The entire flow - data, generation, and delivery - runs off the same system.

What changes on the shop floor

That changes how in-store campaigns are handled day to day. They’re no longer tied to a fixed set of assets, and output can vary by location, by product, or by what’s being pushed.

Changes don’t need to go back through a full production cycle either, which cuts operational overhead. Creative can move with stock, promotions, and seasonal focus without starting from scratch each time.

It also opens the door to personalising physical spaces, in the same way retailers already tailor their online storefronts. Different products, messages, or creative can be shown depending on the store or context, so each store can show what’s most relevant to its customers.

We’re currently packaging up Broadcast for release, with launch planned for June 2026.


Liam Quinn

by Liam Quinn

Director of Innovation

Share :
More from R&D

Related articles

Generating in-store creative from Shopify data

Generating in-store creative from Shopify data

Magic Mirror POC

Magic Mirror POC

Keeping shelf pricing in sync

Keeping shelf pricing in sync