Most vendor-led roundtables are designed to produce consensus. Everyone agrees on the problem, everyone agrees on the solution, the conversation wraps up clean and the recording becomes a piece of content that says nothing that could unsettle anyone.
That’s not what happened when Eduardo Frias from Shopify and Todd Dean from Psycho Bunny got to the legacy systems question.
They disagreed. Directly, respectfully, on a live call with hundreds of attendees. And the disagreement was more useful than any consensus would have been — because it reflects the actual debate that technology leaders are having inside their organizations right now.
The Frias position: stop putting lipstick on old technology
Eduardo Frias didn’t equivocate. Legacy systems, in his view, memorialize outdated ways of working. They got brands to where they are. They served a purpose. But porting old processes into new technology is a missed opportunity, and trying to extend legacy systems with modern capabilities is a trap.
“Stop putting lipstick on old pieces of technology,” he said. “It’s time to throw them their retirement party, buy the cake, print the t-shirt, get the balloons, and send them their merry way.”
His specific concern: brands that implement AI on top of old architecture will feel good for about eight months. Then they’ll realize they haven’t started doing the hard work. The AI is running on a foundation that can’t support what they’re trying to build, and the accumulated technical debt makes every subsequent investment more expensive and more fragile.
The Frias position is essentially one of executive courage. Make the hard call. Don’t delay it. The longer you wait, the more expensive the transition becomes — and the more ground you lose to competitors who made the call earlier.
The Dean position: operational reality requires a different playbook
Todd Dean’s pushback wasn’t a defense of legacy systems. It was a defense of operational pragmatism.
“Legacy systems are hard to move. That’s just reality.”
Dean’s position: for most brands, a full rip-and-replace isn’t on the table — not because leadership lacks courage, but because the operational complexity, cost, and risk of wholesale modernization is genuinely prohibitive for many organizations. The smarter play, in those cases, is to push legacy systems further down the stack: use them for what they’re actually good at, use AI and integration layers to build around them, and deploy new capabilities without waiting for a full modernization.
“Don’t try to make them into something they’re not,” he said. “Peel back the functionality, use it for something core, and then augment.”
Dean’s framework: identify what the legacy system does well — usually some form of system-of-record function — and let it do only that. Then build modern capabilities on top of and around it, rather than through it. The ERP stays the ERP. The WMS stays the WMS. Neither gets asked to do things it was never designed for.
Where they actually agree
The disagreement is real, but it’s narrower than it might appear. Both Frias and Dean agree on the core principle: the mistake is using legacy systems to do things they were never designed for. Both agree that trying to make old systems behave like modern ones — rather than replacing them or building around them — is a trap.
The difference is in the prescription. Frias is arguing for the bold move: make the hard call, modernize, don’t accept the legacy constraint as permanent. Dean is arguing for the pragmatic move: work intelligently around the constraint while you build toward modernization.
They’re not contradictory positions. They’re positions calibrated to different organizational realities.
A framework for figuring out where you land
Two questions help locate where your organization sits in this debate:
How much runway do you have? Brands with strong growth trajectories and investor backing can absorb the short-term disruption of a full modernization. Brands operating on tighter margins, in more constrained environments, may not have that option. Dean’s augmentation approach is often better suited to the latter.
How much technical debt are you already carrying? If the legacy systems are genuinely limiting your ability to compete — if they’re the reason you can’t offer ship-from-store, can’t give customers a unified view of their orders, can’t give your team real-time inventory visibility — Frias’s position is probably right. The cost of staying is already higher than the cost of leaving.
Whether you’re ready for a full modernization or building strategically around the systems you have, Deposco was designed for both — modern foundation, pragmatic deployment, no legacy constraints baked in.
