Deposco and Shopify brought together the executives who’ve actually built unified commerce — for a conversation with no pitches and no filter. Missed it live? Watch the full replay.
The gap between commerce and fulfillment is where retailers win or lose
Key insights from The Unified Retail Stack roundtable featuring EVEREVE, Psycho Bunny, Shopify, and Deposco
Everyone in retail is talking about unified commerce. Far fewer have actually built it. On May 5, 2026, Deposco and Shopify brought together the executives who have — for a roundtable with no pitches and no filter.
Joining Deposco CEO Bill Gibson and VP of Alliances Jack Margeson were Eduardo Frias, Field CTO Enterprise at Shopify; Tamer Selim, CTO of EVEREVE; and Todd Dean, Strategic Lead of Technology and Transformation at Psycho Bunny. If you’re a CTO, VP of Operations, or senior technology leader responsible for how commerce and fulfillment connect at your company, here’s what they said.
The foundation isn’t technology. It’s data.
Every panelist came back to the same place: data quality is the precondition for everything else. Not just for unified commerce — for AI, agentic commerce, any next-generation capability a retailer wants to unlock. Get this wrong and everything built on top of it is unstable.
Tamer Selim put it plainly. EVEREVE’s three-year unification journey started with a hard look at what was actually happening between their systems. Multiple platforms syncing imperfect data was creating cascading problems — in store returns, inventory accuracy, customer recognition. Fixing it wasn’t glamorous. It was the prerequisite for everything else.
“When you place an order in stores, that same order shows up online immediately — same customer record, same transaction history,” Selim said.
Having that validity of data makes it that much easier to integrate other systems without worrying about data quality issues.
Eduardo Frias drew a sharp line between integrated commerce and truly unified commerce. Integrated means the bits and bytes are talking to each other. Unified means a single data layer powers every channel and every decision. The gap between those two things is where most brands are stuck — and where the competitive advantage is for the ones who close it.
Organizational alignment matters as much as the technology.
Here’s the thing nobody puts in the brochure: the technology is rarely the hard part. The hard part is getting your organization to operate as one.
Frias was blunt about it. E-commerce and store operations teams that work independently — with different metrics, different priorities, different definitions of success — will undermine even the best technology stack. “If you’re singing out of two different songbooks, it’s not going to work.”
Todd Dean described how Psycho Bunny uses customer experience design as an organizational alignment tool. Start with how customers actually move across channels — not how you want them to.
When the customer journey is at the center of the conversation, the internal debates about technology and roadmap priorities tend to resolve themselves.
“It helps clear the air,” he said, “and that bleeds into the technology roadmap and the data structure.” Do the CX work before you pick the systems. Not after.
Buy what doesn’t differentiate you. Build what does.
The build versus buy question came up naturally, and the panel’s answer was consistent: the brands winning at unified commerce have made disciplined decisions about where proprietary investment creates competitive advantage — and where it just creates debt.
Frias didn’t soften it. “Why would you reinvent order management, POS, checkout? I wanna meet the person who thinks they can build a checkout that converts better than Shopify’s. Actually, I wanna invest in that company.” The point isn’t that Shopify is the only answer. The point is that building commodity technology when proven platforms exist is a distraction from what actually makes a brand worth choosing.
Selim described EVEREVE’s approach: partner with companies innovating on your behalf so your own team can focus on what makes you distinctly EVEREVE. Dean added the practical corollary for brands dealing with legacy systems — which is most brands. You don’t always have to replace them. Push them further down the stack, use them for what they’re genuinely built for, and build around them. The mistake isn’t legacy technology. It’s trying to make legacy technology do things it was never designed for.
Bill Gibson from Deposco framed the principle simply:
We’re an outcomes company disguised as a software company. Technology is a means to an end. Full stop.
The debate worth having: What do you actually do with legacy systems?
Most roundtables smooth over disagreement. This one didn’t. When the conversation turned to legacy systems, Eduardo Frias and Todd Dean landed in genuinely different places — and both positions are worth understanding.
Eduardo Frias, Shopify
Legacy systems memorialize outdated ways of working. They served us well — but porting old processes into new technology is a missed opportunity. Brands implementing AI on top of old architecture will feel good for eight months. Then they’ll realize they haven’t started the hard work. “Stop putting lipstick on old pieces of technology. It’s time to throw them their retirement party.”
Todd Dean, Psycho Bunny:
Legacy systems are hard to move — that’s just reality. The smarter play is pushing them further down the stack, using AI and integration layers to build around them. Don’t try to make them do things they weren’t built for. Peel back the functionality, use them for what they’re good at, and augment. “Things are getting easier and easier to build. So where does that bolt in? That’s really the art in moving the pieces on the chessboard.”
They’re not as far apart as it sounds. Both agree that using legacy systems to do things they were never designed for is a trap. The difference is in the timeline and the risk tolerance. Frias is pushing for executive courage — make the hard call, don’t delay it. Dean is acknowledging operational reality — sometimes you have to move around the constraint, not through it. Where you land probably depends on how much runway your business has and how much technical debt you’ve already accumulated.
Sequencing the transformation is an art, not just a science.
There’s no universal playbook for the order in which to tackle unification. The panelists were honest about that. But there was a consistent recommendation: start with strategy and customer understanding. Not with technology selection.
Selim’s framework: identify the problems you’re actually trying to solve. Trace them to root causes. Then evaluate partners based on their ability to handle the parts that don’t make you unique. “We choose to partner with those that have teams innovating for us so that when we wake up every morning, there’s something new we’re able to leverage,” he said. “Then we build the parts that make us uniquely EVEREVE.”
For brands carrying legacy systems — which is most brands — Dean offered a reframe. Ripping and replacing isn’t always the answer. Sometimes the smarter move is pushing legacy systems further down the stack, using AI and integration layers to build around them, and deploying new capabilities without waiting for a full modernization. “Don’t try to make them into something they’re not. Peel back the functionality, use it for something core, and then augment.”
Gibson added one more dimension: the best technology partners bring context from outside your four walls. What’s happening across thousands of brands on a platform is intelligence a single brand can’t generate on its own. That external perspective — benchmarks, patterns, lagging indicators — is part of what makes a partnership worth having.
Agentic commerce isn’t coming. It’s here — and the numbers prove it.
The conversation closed on where retail is heading. The panel’s collective view: the shift is happening faster than most brands realize, and the window to get ready is closing.
Frias cited Shopify’s own data, shared during their most recent earnings call. Traffic and orders arriving through AI channels are growing at double-digit rates year over year — and the customers arriving through those channels convert at higher rates and spend more than those acquired through traditional channels. “Roughly two-thirds of people are using LLMs with some kind of commerce intent,” he said. “It’s not ‘oh, it’s coming.’ It’s here.”
That’s not a projection. It’s already showing up in platform data. Brands that have unified their commerce and fulfillment operations are positioned to capture that traffic. Brands that haven’t are sending high-intent, high-value customers to competitors who can.
Selim put the operational implication plainly. The shopping journey is being restructured. Customers aren’t arriving through the homepage anymore. They’re discovering brands through prompts, images, social, AI agents, and physical stores — simultaneously — and they expect a consistent experience across all of it. Brands that can’t show up coherently across every surface will lose ground to the ones that can.
The through line: every investment in data quality, organizational alignment, and platform discipline is also an investment in being positioned for the agentic commerce reality that’s already here.
Buy what doesn’t differentiate you. Invest your time and energy on the things that truly make you unique.
Watch the full conversation or talk to an expert.
The roundtable is available on demand. Watch the replay or connect with Deposco and Shopify to explore how unified commerce and fulfillment can work for your operation.