Big companies love to talk about AI. Strategy decks. Executive memos. Innovation summits. But behind the buzzwords, most enterprise AI efforts have stalled. Why? Because they’re trying to solve a culture problem with a tech roadmap.
And the data backs that up. Seventy-eight percent of enterprises are still stuck in early adoption, largely because they don’t trust their own data—especially when it comes to revenue. More than 80% of AI projects never make it to production, more than double the failure rate of other IT initiatives. And this year, 42% of companies scrapped most of their AI initiatives entirely—up from just 17% last year—with nearly half of pilots never making it out of the lab.
So what’s happening? In so many organizations, AI is being handed off like a hot potato—assigned to one team to prototype, another to assess risks, another to write policy. Meanwhile, frontline employees are left wondering what tools they can use, what’s expected of them, and how any of it applies to their actual work because mandates are replacing support.
The enterprise AI problem
Earlier this year, Shopify CEO Tobi Lütke made it clear: prove the work can’t be done by AI before hiring a human. AI fluency became a job requirement, and performance reviews started factoring in tool usage. On paper, it looks decisive. In practice, it leaves teams scrambling without the training or structure to keep up let alone take the first step.
Microsoft has followed suit a few months later by starting to bake AI usage—yes or no—into employee performance metrics. Copilot became mandatory—but support systems didn’t scale with the demand. Training remained ad hoc. Experimentation, optional. The pressure to perform with AI outpaced the permission to actually learn it.
These kinds of mandates are meant to create urgency but they also cause uncertainty which leads to fear. Urgency without direction and support doesn’t drive innovation—it drives burnout, resentment, and panic. This type of leadership confuses usage with understanding. We don’t have to have a Harvard MBA to know this is a really stupid idea.
What Seer did differently
Now contrast that with Seer Interactive. Seer is a search and SEO agency, not a tech conglomerate. But when it became clear that AI would change how people search, the leadership treated it like the existential shift it is.
Founder Wil Reynolds didn’t bury the challenge in a cross-functional team or hand it off to IT. He made it a top priority. Last year he promoted an employee to VP of Generative AI & SEO. Their job wasn’t to run isolated experiments or drop sudden requirements. It was to help the whole company evolve.
The new VP built an internal task force with members from every division. Each one got dedicated time—eight hours a week—to test tools, run experiments, and report back their results to the group. They meet weekly to align on a roadmap, share results, and refine their approach.
The tools weren’t locked behind gatekeepers. Seer provided company-wide licenses to vetted AI platforms and encouraged open exploration. They built a Slack channel to share discoveries. They created an intake process for GPT ideas and have set up several experiments—many of them collaborative. Anyone can pitch. Anyone can build. This isn’t innovation by shock and awe. It’s innovation by human-centered leadership.
What most companies are missing
Most enterprise leaders are trying to make AI safe before they make it useful. That usually means locking it down, narrowing access, and waiting for a clear ROI before investing in culture. The result? A lot of shelfware. And a lot of teams who feel like they’re waiting for permission to try.
Seer flipped the model. They bet that people learn best by doing. That if you want your team to get good at AI, you have to build systems that reward curiosity, not just usage. Encouraging and rewarding curiosity is and will be a key ingredient to thriving with this new technology.
More importantly, Seer didn’t just equip people with tools—they gave them protected time.
The leadership move that matters
Here’s what Seer’s approach shows us: no one has AI figured out. But the smartest teams aren’t waiting for clarity. They’re creating the conditions to figure it out together. At Seer, exploration isn’t a side hustle. It’s part of the job. Ownership is shared—not siloed. There’s structure, but not rigidity. Ideas don’t disappear into the ether—they turn into systems. Experiments become habits.
It’s not perfect. But it’s better than 99% of the bullshit coming from the Enterprise. It’s not a plug-and-play strategy. It’s a leadership posture.
So if you’re a leader wondering how to “do AI right,” don’t start with a memo. Start with your team. Give them time on the clock to experiment. Vet tools and make them accessible. Normalize learning in public—especially when it’s messy. And build rituals for sharing what’s working (and what isn’t).
The companies that win with AI won’t be the ones with the biggest budgets or the business journal headlines. They’ll be the ones that treated AI like a culture shift—not just a production upgrade. You can’t automate your way to real innovation.
T L : D R — Most big companies are getting AI wrong—issuing mandates without offering support, structure, or clarity. Seer Interactive, a small SEO agency, shows a better path: they treat AI as a cultural shift, not a tech upgrade. By giving their team time, tools, and a shared mission, they’re building capability the right way—one experiment at a time.
