I mean, there are a lot input/output apps, and there will be more. What if Shopify aligns its architecture in a way that allows merchants to create their logic/chain and connect inside Shopify interface in a way that let merchant use any of available apps on Shopify store and connect it in his own chain, so Save would not directly write to Shopify but send info to another app that communicates to another app - simple example would be - merchant want to enrich product data, he goes into app that enrich, after that he put that same data through SEO corrective app, then data flows to app that derive metrics, optimize sales, and through production pipeline send changes? Custom apps will overflood the market and Shopify could benefit from becoming the orchestration layer for these workflows.
Save would no longer have to mean publish. It could mean: submit this change into the merchant’s configured production chain. This feels like the logical next step.
I like the idea, especially for merchants who already use several apps to manage and improve product data.
The main challenge would probably be keeping the workflow clear and manageable. Merchants would need to see which app changed what, what order the apps run in, and what happens if one step fails.
If Shopify added a simple workflow history, permissions, and rollback option, this could be genuinely useful instead of making store management more complicated.
I’d also be interested to see how Shopify could standardise the way different apps send and receive product data.
That is exactly the key point. AI powered apps will eventually need shared protocol for communicating. From that perspective Shopify is in a strong position to provide trusted infrastructure with standardized inputs and outputs, and control plane for connecting apps.
Custom apps would then work like a Lego blocks - each with specialized functions, remaining interoperable, replaceable and governed within the merchants own workflow and logic.
Hello there @CommerceGov
That’s a fun direction. Shopify Flow does already automate some proceses, but a merchant-defined product change pipeline would be a lot further, allowing apps to handle changes before they go live. That might lead to less manual labor, more consistent data, and merchants getting more flexibility without custom integrations. A visual orchestration layer with approval steps and rollback options would make complex product workflows easier to manage and everything would remain in the Shopify ecosystem.
Hey, @CommerceGov
Hope you are doing well!
That’s an interesting idea. A built in workflow orchestration layer could make app integrations much more seamless and reduce the need for custom middleware. It would also give merchants more control over how data flows between apps before changes are published.
Yes. Shopify Flow already orchestrates events and actions, but changes still usually move directly from an app into production.
A merchant-defined change pipeline could introduce an entirely new way of operating. It would enable new business logic, new systems, and even new categories of apps capable of working on almost any type of change. Instead of writing directly to production, those apps could operate on the same proposed changes within and intermediate governed state before any writeback ocurs.
That line of thinking has gotten us to build a @CommerceGov. Apps or AI systems can propose changes, teams can review, edit and save, while policy enforcement, human approval, audit evidence, controlled writeback, and rollback remain part of the production boundary.
Very well, thank you! I hope you’re doing well too.
Exactly, orchestration layer would define how data moves between apps, while a governance layer controls when and what data is allowed to become production fact.For example, merchants could route proposed changes through enrichment, validation, SEO, and approval before they are written to Shopify - with rollback available if needed.
This is the idea that we are exploring with CommerceGov: proposals could eventually come from Any AI system or application, but production writeback remains separate process which is policy controlled, reviewable, auditable and reversible.