For the last two years, we've all been strapped to the front of the AI rocket ship. The feeling has shifted from "This is cool," to a frantic, "How is everybody moving so fast?" If you feel like you're on an AI treadmill, struggling just to keep up, you’re not alone.
While many companies are still trying to catch their breath, a select few have turned the chaos into a competitive edge. They’ve cracked the code on AI transformation, and their playbook isn’t what you might expect. It’s less about having the most advanced technology and more about a fundamental shift in how they approach problems. This article distills the five most surprising and impactful lessons from their success.
Transformation is People-Powered, Not Process-Driven
Most leaders think AI transformation begins with a top-down mandate and a new tech stack. The trailblazers prove this is backward. Real transformation doesn't start in the boardroom with governance policies; it starts with your people.
The core argument is simple: you can’t build a culture of innovation from a memo. You have to give people the tools and the freedom to experiment. Wade Foster, CEO of Zapier, outlines a clear three-step playbook that flips the traditional model on its head:
- Get the team the right tools. Provide access to a range of AI applications, from intelligence layers like ChatGPT to no-code platforms for non-technical staff. Everyone needs something they can actually use every day.
- Get everyone building and showing off their results. Encourage experimentation through initiatives like internal hackathons. When people have to present their work, they try harder, and leadership gets a direct line of sight into the best ideas bubbling up from across the organization.
- Make AI a part of evaluation and hiring. This step comes last, intentionally. Why? Because you want your team to feel like they've "had a chance to stick their hands on the goo and get familiar with building with AI first," before making it a performance metric.
AI transformation, it happens through people, not necessarily process. You need to give people the tools, you set the culture, you set the standards, That's your foundation.
Once you've committed to empowering your people, the next question is obvious: where do they start? The Portland Trailblazers found the answer not in a complex framework, but in a brilliantly simple question.
Find Your Starting Point by Asking "It Sucks That..."
So how do you uncover AI's highest-impact use cases? Forget the consultants. The Portland Trailblazers found their answer in a single, disarmingly simple prompt.
As part of a program they called "Lunch and Launches," they brought every department in for an informal session with a single, non-technical instruction: finish the sentence, "It sucks that...".
This approach is powerful because it democratizes innovation. It reframes AI from a technical solution looking for a problem into a tool for solving deeply felt human frustrations. Instead of asking people to imagine AI solutions, they asked them to identify real problems. This generated dozens of concrete inefficiencies to solve, like a customer feedback process that consumed a staggering 50 hours a week across three departments. By framing the challenge around shared pain points, they unlocked innovation from everyone, not just a specialized AI team.
The Most Powerful Results Can Be Unexpected Cultural Wins
After identifying their 50-hour-a-week feedback problem, the Trailblazers built an AI-powered workflow that slashed the work down to one person, three hours a week. But the most powerful outcome was something they never planned for.
They realized they could "flip the system." The same workflow designed to efficiently handle negative customer feedback could also be used to capture and share positive feedback. They created a dedicated Slack channel called "Brand Hugs," which automatically populated with glowing messages from fans.
The impact was profound. The "Brand Hugs" channel became a massive source of motivation, boosting staff morale by sharing real, positive stories from the community. It helped employees connect more deeply with their customers and the impact of their work. This discovery proved that the most valuable AI outcomes aren't always about efficiency; sometimes, they're about humanity.
I just love that we're using AI to create human-to-human connection at scale.
The Smartest AI Systems Still Need a Human in the Loop
As companies scale AI, a common fear is losing control. The most effective systems solve this by keeping a human in a position of authority, especially for critical tasks that involve customer communication or key decisions.
Faced with a deluge of customer feedback, Anna Marie Clifton, Director of AI and Agents at Zapier, didn't hire more staff. She built an AI system to act as her team's frontline analyst. Her system triages incoming feedback, classifies it, updates the product roadmap, and even drafts a personalized email response. But the final, most important step is manual.
The agent's intelligence goes way beyond keyword matching. It analyzes a customer's specific use case and compares it against the product spec on the roadmap to confirm a feature fit. Only then does it draft a response. Before any AI-drafted email is sent, the system sends a notification to her in Slack for review and approval. She can approve it as is, edit it, or reject it entirely. This "human-in-the-loop" approach embodies the principle of "AI suggests, humans decide." It delivers the speed and scale of automation without sacrificing the trust, control, and personal touch that only a human can provide.
You Don't Need to Be a Coder to Be an AI Innovator
Perhaps the most significant lesson is that the barrier to building with AI is lower than ever. You no longer need a technical background to be a powerful innovator.
Andrew Harding, VP of Marketing at Slate magazine, is a perfect example. Tasked with building a new subscription product, he faced a massive lead generation problem that would have taken his team months. With "no technical background," he used Zapier to build an AI agent using "nothing but plain English."
The agent's task was to go online, identify specific faculty prospects for the new subscription product, and drop their contact information into a shared spreadsheet. This single agent saved him "dozens of hours per week for months." More importantly, it delivered incredible results, generating prospecting emails that achieved 45% open rates and 30% reply rates. Andrew's story shows that the people closest to a business problem—the subject matter experts—are now empowered to build their own solutions.
Conclusion: From AI-Curious to AI-First
The message from the front lines is clear. Transformation isn't about the tech itself, but about orchestrating it. It begins by trusting your people over process, focusing them on real pain points by asking "It sucks that...", and embracing the unexpected cultural wins that follow. It scales by keeping a human in the loop for critical decisions and empowering the non-technical experts closest to the problem to become the innovators themselves.
The winners won't be the organizations with the most tools, but rather "the ones who orchestrate those tools into systems that actually work." As we move forward, the next 12 months will "separate the AI-first companies from the AI-curious ones."
So, as you leave here today, ask yourself the same question the Trailblazers did: What is the biggest problem at your company that makes you say, "It sucks that..."—and how could you start solving it today?
