Watching the LA tech community grow over the years has been one of my favorite parts of calling this place home.
This week, I was thrilled to be in a few rooms with builders and dreamers at LA TechWeek. I’ve been more protective of my time this week, so I used a not-so-secret metric for each event: "time to wow." And luckily, my TTW was pretty short!
Here are 5 themes that I took away from events focused on building in AI, product development, and working as a fractional employee in the AI era.
Oh and if you’re an AI-inclined solo builder or if you have a small team working on an AI project, keep reading for things you can try!
Every builder should be tracking "time to wow"
Leandrew Robinson spoke about delighting customers in the Founders Scaling in AI panel. "Every founder should be tracking how fast you can get customers to say 'wow'." How do you do it? Go the extra mile by loading publicly available customer data into your demo environments to show relevant use cases.
AI products, like many others, need to capture attention in the first 60 seconds of your demo. ChatGPT set the bar at 3 seconds. If your product doesn't make someone say "this works" almost instantly, you've lost them.
Try this: Record your next client call or demo. Watch it back and time how long until your client's tone shifts from polite to engaged. That shift is your TTW. Now build your next demo to hit that moment in a fraction of the time.
Get buy in from the top
Josh Christensen from Mercato Partners synthesized the AI Native Product Dev panel with one truth. "To build an AI native company you have to start with buy in from the top. Without it, you can't really start." This pattern showed up in many events. The successful founders weren't fighting uphill battles for resources. They were working directly with the decision makers.
When you have buy in from the top, your job becomes less about convincing and more about making the customer successful. If you're implementing AI without executive commitment, you're being set up to fail.
Try this: Before accepting your next project, have a 15 minute call with the actual decision maker. Ask them directly: "What happens if this works?" If they cannot paint a clear picture of success in 30 seconds, decline the project. Your time is too valuable to waste on uncommitted customers.
If you're trying to figure out where AI fits in your product strategy, let's chat! I'll help you identify the 1-2 AI features that will deliver the bulk of your value. Plus, you'll get all the support you'll need to deliver.
Cross functional "tiger teams" beat specialized AI teams
With any special project, AI included, the most effective teams are cross functional teams with members from different backgrounds, where the company knowledge sits. Roaring like a tiger may also make your team more effective (if you try this, let me know). 🐯
What I see is that domain expertise beats AI expertise for most use cases. The winning combo is people who understand the problem deeply, can implement quickly, and keep their focus on the value you bring over the specific tech.
Try this: As a solo builder or small team, you're often all you've got. Map out the three roles you need for your next AI project: problem owner, technical implementer, and customer voice. If you can't fill all three yourself, reply to this email (happy to help!) or find specific people in your network who can fill each gap. Propose a short collab or knowledge exchange.
Go "forward deployed" to understand your customer
The forward deployed engineering model involves deeply embedding yourself with customers to understand their real problems. Whether you have a big team or you're working solo, this model is a no-brainer. The Fractionals breakfast I attended reinforced this, too. Warm leads and deep engagement beat marketing funnels every time.
You can't understand the problem from your home or office. Generic solutions fail because context is everything. Greg Schoeninger (Oxen.ai) runs weekly virtual sessions that are open to virtually anyone. He brings practitioners together to study and implement together. That's what I call forward deployed community building.
Try this: Pick one online community where your ideal clients spend time. Reddit, Discord, LinkedIn groups, or Slack communities all work. Commit to showing up three times per week for the next month. Answer questions, share insights, and help people solve problems. If you can help it, try not to mention what you're selling right away.
This is an approach I teach in my 3 day email course, I show you how to use AI powered research to deeply understand your clients. Then, use what you learn to write authentic, human messages that get responses.
Keeping humans in the loop is table stakes
The consensus on building safe AI was clear. "Human in the loop is key. Checkpoints and deeply understanding your workflow are critical." Josh Christensen added, "You have to have targets or evals to ensure what you are building works." This is not safety theater. This is competitive advantage.
This approach builds trust, catches errors before customers see them, and creates continuous improvement cycles that make your AI product better over time. This is the difference between a cool demo and a system you can bet your business on.
Try this: For your current AI project or client engagement, create a simple one page document that lists every critical decision your AI system makes. For each decision, write down who reviews it and how often. If you can’t identify a human checkpoint for the top three riskiest decisions, add those reviews into your workflow. Share this document with your client to show them you are thinking about reliability.
I'd love to know what resonated most with you from this recap! Hit reply and let me know. I read and respond to every email.
And if you found this valuable, forward it to a founder or product leader who is building in AI.
Thanks for reading,
Igor