The Trojan Horse Playbook for Selling AI in the Real World
Welcome back to Founder Mode
As we build Pretty Good AI, I see that most enterprise deals fail. It’s not because the technology isn’t good enough. They fail because we ask for too much trust, too early.
Founders love pitching the big vision. The platform. The transformation. The future state. But most buyers are buried in today’s problems. They don’t have time to imagine a new world. They just want one annoying thing to go away.
This week, I have some lessons to share. They are about joining real groups. They also cover how to build trust fast and sell AI well. None of this comes from theory. It comes from hard work, missing deals, fixing errors, and seeing what works.
The Trojan Horse Utility Tool
Selling a big platform change in the first email almost never works. It rarely succeeds. It’s too much change, too much risk, and too much thinking.
At Pretty Good AI, we flipped the approach. We looked for a small, shared headache. Our customers really disliked it. One small issue everyone noticed but learned to accept as normal. For us, it was collecting documents at check-in.
We took that fix out of our larger product. Now, it’s a simple, free tool. No long pitch. No contract. Just immediate value.
That tool earned trust. More importantly, it earned integration rights. Once we were inside their workflow, the conversation changed. The flagship AI product was no longer abstract. It was the natural next step.
Solve one small problem perfectly, and you earn the right to talk about the big one.
Demographics Dictate Tech Needs
We learned quickly that sorting by industry was not enough. Two clinics may seem the same on paper, but they can cause very different pain.
The difference often comes down to patient behavior.
Specialties that serve older populations get crushed by phone calls. Their patients do not use apps. They do not log into portals. They call. Over and over.
Other specialties do not feel that pain at all. Their patients self-serve. They schedule online. They upload forms without help.
We focused on friction. This made our targeting better. We stopped selling to everyone. Instead, we focused on the group that created the problem we solved.
If the users refuse to change, the business has to. That’s where AI becomes urgent instead of optional.
If It’s in the Database, It’s Already Burned
Early on, we relied too much on standard B2B databases. Titles. Emails. Firmographics (bro! I learned a new word).
The problem is that everyone else is using the same data. Those inboxes are already burned and abused.
The real edge came from doing the unglamorous work. Mapping ownership structures. Finding the private equity groups behind dozens of local brands. Linking parent companies to operators on the ground.
At Pretty Good AI, this manual detective work made a big difference. We decided to reach out to people others hadn’t contacted yet. This way, we avoided the same old list.
If your data looks clean and easy, it’s probably not differentiated.
The Check Writer vs. The Pain Holder
One of the biggest mistakes in enterprise sales is using one message for everyone.
We learned to split our outreach into two tracks.
The first is the innovation executive. This person cares about positioning. They feel competitive pressure and don’t want to be left behind. They respond to strategy and future impact.
The second is the operations manager. This person is drowning. Short-staffed. Overworked. They do not care about vision. They care about making today survivable.
Pitch strategy to the person in pain, and you lose them. Pitch tactics to the person writing the check, and you lose them, too.
At Pretty Good AI, matching the message to the role helped get more replies. It worked better than any copy change.
Why This Works
All of these lessons point to the same idea. Reduce the cognitive load.
Make it easy to say yes. Make the value obvious. Make the next step small.
AI adoption doesn’t fail because people don’t believe in the future. It fails because they’re exhausted by the present.
When you meet them where they are, trust builds naturally.
5 Key Takeaways
- Lead with a utility, not a platform. Small wins open big doors.
- Follow user behavior, not industry labels. Demographics create urgency.
- Do your own data work. Differentiated lists create a real advantage.
- Split your messaging by role. Strategy and pain are not the same.
- Lower friction everywhere. Less thinking leads to faster adoption.
Final Thoughts
Building Pretty Good AI keeps reinforcing the same lesson for me. Selling AI is not about convincing people it’s powerful. Everyone already knows that.
It’s about making it feel safe, useful, and immediately helpful.
Cut out friction. Fix real problems. Get how people think and act. Then, AI feels less risky and more obvious.
That’s when adoption happens. Not because of hype, but because it finally makes life easier.
See you next week!
-kevin
Recent Social Posts
Recent Podcasts
2810 N Church St #87205, Wilmington, DE 19802
Unsubscribe · Preferences
