My AI Journey: Not One Lightbulb Moment, but 100 Sparks that Gave Me Deep Conviction (& why you're not behind on AI)
+ My top 11 lessons from my AI journey so far
If you were forwarded this newsletter, join 4,179 weekly readers—some of the smartest GTM founders and operators—by subscribing here:
Hey y’all! 👋
Jason Lemkin’s “tough love” message in his opening talk at Saastr has stuck with me: adapt quickly, be relentless, and work harder to navigate the evolving landscape and ensure continued success. At the core of this moment is adopting AI.
Every week, my conviction around AI grows stronger (at least in the GTM space, which is the corner of the world I know). Watch his full talk here: SaaS + AI: It’s Time to Move Much Faster.
As Jason put it:
Your team should be both scared and like a kid in a candy store.
That describes my feelings about AI to a tee.
Everyone’s bragging about instant AI wins online.
But my reality looked different. It took me 9+ months—hundreds of wrong turns and dozens of micro-breakthroughs—to get real leverage from AI.
Today, I wanted to share the roadmap I wish I had before starting down this winding path.
My hope for you (and to myself!) is that this encourages you to keep learning, because we’re still early. Yes, early. You’re not behind. (The voice inside my head just reminded me: you’re not behind either!)
Let’s get into it.
My AI Journey: Not One Lightbulb Moment, but 100 Sparks That Gave Me Deep Conviction (+ why you're not behind on AI)
I've been in B2B SaaS GTM for over a decade. I've started a couple of venture-backed SaaS companies, and I've been at a couple of big companies (Zoom Video and Apollo.io).
But it has taken me 9+ months to really understand the power and leverage that AI can give me in GTM. And, maybe more interesting, I believe I'm early in my journey in discovering what's possible. In fact, I think everyone is early in discovering what's possible. I don't think we yet know where the best places to apply AI. But there are a few areas that are emerging as great places to deploy AI today.
My journey
The best way I can describe my journey is that I did not have a single (big) light bulb moment. Instead, there were a hundred (small) sparks that added up to give me the proof I have today that AI is going to change ~everything in GTM. I think it's going to touch everything (like electricity).
Testing
Shipping experiments is by far the most important thing that I've done to learn.
Building a workflow, throwing that workflow away.
Writing a prompt, throwing that prompt away.
Vibe coding an app, throwing that code away.
Using a new tool, throwing that tool away.
9 out of 10 things I try with AI do not work. I have the luxury to have a decent amount of time on my hands right now to tinker with this stuff. And yet, I still I don't feel like I have nearly enough hours in my day to keep up with how fast the space is moving.
Examples in GTM
I want to touch on some of the use cases that I'm starting to see in GTM (I’m working on a dedicated post on tactical examples that will go live later this month).
I'm starting to really believe that having AI do things like research are incredibly powerful. Think about your best rep and what they do when they research an account and the people at that account, how they're spending the 20 minutes doing all that research — that can now be done with 95% accuracy with AI workflows and agents. This is a huge unlock.
Another way to think about this is over the years you've probably outsourced work to somebody overseas that is doing manual work and then uploading a CSV into Salesforce. AI replaces this worker. And the quality of work is better. It's faster and it's cheaper. This reshapes what is possible in GTM. New workflows emerge that weren’t possible before (because of constraints in time or money).
The other things I’ve talked about a lot are things like: signal-based selling, evergreen plays, micro-campaigns and using ai & automation to scale, instead of headcount alone.
Lean into learning
Again, I want to call out the fact you may not have had a light bulb moment about AI. And there's a good chance you won’t have one single (big) magical aha moment.
But, I am confident that if you give it enough time and play with these new tools, you will have a hundred small moments that will add up to the realization that AI is the biggest thing in our lifetime, and it will certainly impact GTM in ways we can't even imagine today.
My top 11 lessons from my AI journey so far
(Also shared on LinkedIn last week)
1. AI gets you to 80% in minutes—the last 15% takes 95% of the work.
You have to push to get over that initial hump. And, don't expect 100% with AI.
2. You can't learn by LinkedIn posts, tweet threads, YouTube or Substack.
You have to build things with AI first-hand. Find a small problem to solve, and figure out how to build it. Either with LLMs directly, or other AI-native tooling.
3. Things change SO fast (eg: weekly!).
Don't get married to a model or even an LLM. If there is a problem today (eg: hallucinations), keep an open-mind that in the not-so-distant future, that may be solved ~entirely.
4. You won't have one big lightbulb moment.
Instead, it will be 100 little flickers that all add up. Just keep tinkering.
5. Fall in love with learning.
Be curious. If you're not interested (or incentivized) to learn this stuff (literally hours a day), it's going to be hard to stay up to date. You have to find the best sources and go direct to those places to learn. Then, test yourself.
6. 9 out of 10 experiments I've tried with AI fail (probably more tbqh).
Assume the vast majority of things *won't* work—today. But, remember to revisit experiments as models improve and new things are created (like MCP servers).
7. Build fun, silly things. Stuff for personal life, not just professional life.
Building things you're really excited to exist (even if it's just for you) is what will help you do that second shift after you put the kids down for bed.
8. Think of AI like an entry-level co-worker that you manage.
It's almost always easier to do a task yourself ("the old way"). But, if you take the time to pass off your work to AI, you'll gain wild leverage.
9. Don't be overwhelmed by feeling behind.
I still feel super behind. I honestly think I *am* super behind. That's okay if we're still early with AI. And, I think we are. "Best time to plant a tree is 20 years ago. Second best time is today."
10. Lose the ego.
I feel like a poser for posting this. I am, probably (lol). But, this is a 'public diary entry.' A note to myself. And, hopefully it helps 1 or 2 people to start hacking on that agent or finally vibe code that website.
11. My (current) favorite tools I'm using:
- Lovable, ChatGPT o3, Superwhisper, any of the Deep Research models, Delphi, Clay, Sumble, Sonnet 4.0.
- I'm currently dabbling with n8n and Lindy.
- And I need to learn more about MCP servers and get better at vibe coding.
Thanks for continuing to be on this journey with me (and I appreciate you letting me have this cathartic release of some ideas that have been swirling in my head over the last couple of weeks).
As always, thank you for your attention and trust.
Keep exploring,
Brendan 🫡