When it comes to Product 360, most of the conversation around Informatica’s acquisition by Salesforce has been about the wrong question. Teams want to know when they need to migrate off P360. The question I’d ask instead is whether the platform they’re staying on for the next several years is actually built to perform. And whether it’s ready to make use of the AI capability already sitting inside it. I’ve spent over 15 years implementing and running Product 360 for clients, including a recent engagement with a manufacturer that grew its primary ecommerce revenue 65% in a year without migrating off the platform.
Extended Support and the Gap Between On-Prem P360 and P360 SaaS
The Salesforce acquisition of Informatica set off a wave of questions about migration timing and product roadmaps. For on-premises P360 customers, Informatica’s stated extended support with no forced migration path. That’s the headline most people stop at.
Underneath that headline, P360 SaaS is not the same product as the on-prem P360 PIM most enrichment and merchandising teams are running today. On-prem P360 is built as an enrichment and multichannel syndication engine: the workspace where product content gets built, governed, and pushed out to channels. P360 SaaS, as it exists now, is positioned as a master data management (MDM) tool for the product domain. It is focused on deduplication, entity association, and analytics feeds. It serves a different purpose and a different user, and it doesn’t go as deep into enrichment workflows.
For a team whose P360 investment is built around enrichment and channel publishing, that gap is the whole ballgame. Migrating today means moving to a product that shares the name but doesn’t yet do the enrichment and syndication work your business depends on.
The same logic applies to AI. Informatica is building P360’s future around AI agents. The CLAIRE® Product Experience Agent reads supplier files, extracts attributes, and classifies products with far less manual keying. That’s real, and it’s useful. Informatica says the automation can cut time to market by 50%. That figure assumes the data model and governance feeding the agent are already in shape, and an agent performs only as well as the foundation underneath it, on-prem or in the cloud.
The Problems Extended Support Won’t Fix
Extended support buys time to deal with a harder question: whether the platform you’re keeping is set up to perform.
In most P360 environments I look at, it isn’t. The usual causes are attribute models that grew past any governed standard, enrichment rules that were never fully enforced, manual workarounds that quietly became permanent process, and publishing cycles that take days or weeks instead of hours. These are optimization and operating model problems. They were there before the Salesforce acquisition, and they’ll be there after it, whichever platform they end up living on.
They also compound in a way most teams don’t expect once AI enters the picture. An agent working from inconsistent attributes and ungoverned enrichment produces confident, inconsistent output at scale. Fixing the foundation is now a prerequisite for AI readiness as well as good PIM performance, in P360 and in whatever else your AI program touches next.
The fastest way to tell where you stand is to pull your publishing cycle time and your enrichment completion rate. If either number would embarrass you in a board meeting, that’s where I’d start.
What a Well-Run P360 Environment Delivers
A global industrial equipment manufacturer came to us running P360 across more than a million SKUs and multiple global divisions. The environment was familiar: siloed enterprise resource planning (ERP) systems, digital assets scattered across disconnected repositories, enrichment done by hand, and a publishing cycle that ran weeks from content-ready to live.
We didn’t replace the platform. We optimized the one they had (the data model, the workflows, the governance) and moved day-to-day operation onto a managed services model. Their team stopped running the system and started using it.
A year later, the manufacturer’s primary channel ecommerce sales were up 65% year-over-year, and publishing time had dropped from weeks to under 24 hours. Time to identify missing product data fell to 80%, and the number of digital assets linked to ecommerce listings increased 600%.
That’s the work we do with P360 PIM customers: optimize the implementation they already have, then take the day-to-day operation off their team with a managed service. If you’re running on-prem P360 today, I’d spend the time extended support gives you on the data model, the workflows, and the governance. That work pays off on whichever platform you end up on, and that’s what will make the AI agents worth switching on.
Most teams I talk to know something isn’t performing the way it should; they’re just not sure where. We can usually tell you in a single conversation, using the same diagnostic approached we used before any of the results above were possible.
