Perspectives
Executive briefs on commerce, growth, and the decisions behind them.
Short, operator-grade notes for CEOs, boards, and investors. Written from inside the rooms where these decisions get made.
Transformation · June 2026
Why Most Digital Transformations Fail Before Technology Is Ever the Problem
Most digital transformations are framed as technology programs. A new platform, a new stack, a new tool that will finally make the organization modern. Two years and several million dollars later, the platform is live and almost nothing has changed. The roadmap slipped, adoption is thin, and the business case quietly disappears from the next board deck.
The technology is rarely the reason. In nearly every stalled transformation I've seen or been brought in to fix, the platform worked roughly as advertised. What failed was the operating model it was bolted onto, the way teams were structured, how decisions got made, what people were measured on, and who actually owned the outcome. You can install the best commerce platform in the category on top of an organization that still runs in disconnected silos, and you will get an expensive version of exactly what you had before.
Consider what a transformation actually asks of an organization. Traffic acquisition, on-site conversion, and post-sale retention have to operate as one performance loop. But in most companies those three live in three different departments, with three different leaders, three different budgets, and three different definitions of success. The platform doesn't fix that. It exposes it. The moment the new system goes live, every seam in the operating model becomes visible, and the teams default back to optimizing their own metrics because that is still what they're rewarded for.
This is why the sequence matters so much. The organizations that get transformation right almost always do the unglamorous work first. They align the teams around a shared outcome before they buy the tool. They redraw accountability so one leader owns the end-to-end result. They change what gets measured so the incentives finally point in the same direction. Only then do they layer technology on top, and when they do, it compounds, because the operating model is ready to absorb it.
The companies that do it backwards, technology first, operating model never, end up with the same plateau they started with, plus a depreciation schedule. The tool becomes the scapegoat. Leadership concludes the vendor underdelivered, or that the category is harder than it looked, and the real problem, an operating model built for a different era, survives untouched into the next initiative.
There is a useful diagnostic question for any leader weighing a transformation. Before approving the platform, ask: if this system were already perfectly installed tomorrow, what in how we operate would still stop us from winning? If the honest answer is a list, that list is the transformation. The technology is just the part that's easy to buy.
None of this means technology doesn't matter. It means technology is the last 20 percent of the work and the first 80 percent of the budget, and that imbalance is exactly why so many transformations underdeliver. The leaders who reverse it, who treat the operating model as the product and the platform as the enabler, are the ones who get the compounding return everyone else was promised.
Jordan Ste. Marie
AI & Discovery · May 2026
AI, Search, and the Next Shift in Customer Discovery
For twenty years, getting found meant getting ranked. A customer typed a query, scanned a page of blue links, and clicked. The entire discipline of digital discovery, SEO, paid search, the content economy, was built on that single behavior. That behavior is now changing faster than most brands' operating models can keep up with.
Increasingly, the first surface a customer encounters isn't a list of links. It's an answer. ChatGPT, Perplexity, and Google's AI Overviews are synthesizing a response and, often, recommending specific products and brands before the customer ever reaches a results page. The question is no longer only whether you rank. It's whether the systems generating these answers know you exist, understand what you do, and consider you credible enough to name.
This is a meaningful shift in the economics of discovery. In the old model, you could buy your way to visibility, bidding on the right keywords put you in front of demand. In the emerging model, a growing share of discovery happens inside a generated answer where there is no ad slot to buy and no link to bid on. Visibility is earned through how comprehensively and credibly your brand is represented across the sources these models read, not purchased at auction. Brands that are invisible or poorly represented to those systems risk being quietly left out of the consideration set, before a human ever evaluates them.
I'm deliberately careful here, because this is exactly the kind of shift that invites overreaction. AI discovery is not replacing search overnight, and it is not the only thing that matters. Traditional search, retail media, and direct channels still drive the overwhelming majority of commerce today. The right posture is not to abandon what works and chase the new thing. It's to understand a structural change early and position for it while it's still cheap to do so.
What that looks like in practice is less exotic than the headlines suggest. The brands that show up well in AI-generated answers tend to be the ones with clear, well-structured, authoritative content about who they are and what they sell, the same fundamentals that have always underpinned good discovery, now read by a different kind of reader. It also means treating your presence across the broader web, reviews, third-party coverage, structured data, as part of how the machine forms its view of you, not as an afterthought.
I led one of the first national-scale efforts to position a major brand for this shift, building a generative engine optimization strategy alongside traditional SEO rather than instead of it. The lesson wasn't that AI changes everything. It was that the organizations that treat it as one more discipline inside a coherent discovery strategy, rather than a bolt-on experiment, are the ones that will compound an advantage while competitors are still debating whether it's real.
The honest summary for any leader: this is early, the data is still forming, and anyone claiming certainty is selling something. But the direction is clear enough to act on. Understand your exposure, get the fundamentals right, and treat AI discovery as a capability to build deliberately, not a trend to chase.
Jordan Ste. Marie
More to Come
Operating ModelBuilding a Digital P&L That CompoundsForthcoming
Growth StrategyWhy Founder-Led Brands Stall Between $5M and $20MForthcoming
Private EquityDigital Diligence: What Operating Partners Miss in Consumer DealsForthcoming
Retention & LTVThe Retention Math Most DTC Brands Get WrongForthcoming