Hey y'all, around the start of the year I signed a letter of intent to sell Megaphone for $1,128,000 (before add-backs).
We got inbound interest, and I thought maybe I wanted a bigger buffer of cash in the bank. AI was a big reason why.
Megaphone helps founders and investors build their authority with the audiences that matter most to them. That’s potential investors, hires and partners, and customers through content on X, LinkedIn, and newsletters.
So watching AI get better at writing made it pretty easy to get swept up in imagining a future where that business was worth a lot less.
I also worried about something bigger: what if AI made it harder to build wealth at all? Taking a meaningful amount of cash off the table “while I could” started to make sense.
So I took some steps to move forward with the acquisition.
But then I changed my mind and turned it down. In fact, I went the opposite direction and doubled down on service businesses, launching JOLT! right around the time I turned the offer down.
And now I have no desire to sell either even as they’ve continued to grow.
I want to walk you through my thinking, because I think a lot of founders are trying to reconcile the same two ideas I was: AI makes building a business easier, and AI might destroy the business you're building.
This week’s skill is called Exit Math. It helps your AI figure out what price would actually make selling your business worth it, and whether an offer on the table is a good deal.
It works out your walk-away price: the cash you'd keep after taxes, fees and escrow, compared with what the business would pay you if you kept it
It shows how much more a buyer could make by running your business with AI the day after closing, so you don't hand them that upside for free
It checks the offer itself (earnouts, seller notes, holdbacks) and tells you how much of the headline number is actually cash
Hit reply and tell me what would’ve made you sell.
The future I was selling into
Selling would have been buying into a pessimistic vision of the future:
AI keeps getting better. More of the work clients pay us for becomes cheap. Competitors show up, or more gets done in-house. Prices fall. The business gets harder to defend.
Selling would turn that uncertainty into cash I could keep while the AI-disruption-dust settles.
That concern wasn't irrational.
If what you’re selling is writing, AI really does put pressure on it. I still think plenty of agencies are going to struggle.
But I was sleepwalking into a much bigger assumption: that the work becoming cheaper meant there would be less opportunity to build valuable businesses around it.
Three things changed my mind.
1. Content is still human
If you see a post in your feed, you’re probably able to tell if it was written by AI or not.
Personally, I don’t think most people care if the information they consume is AI-written or not, but they definitely care if that same AI-writing is presented to them from someone’s personal brand.
It just feels icky. Like the person is trying to pull one over on you, even though they know that you know it’s AI.
Opus 5.5 came out in September and models will continue to get better, but I don’t see this “gap” (that’s a joke) closing anytime soon.
2. Going AI-native ourselves
Aside from content, AI does get a lot right.
And while we were under LOI, we were building too. I set up our company brain and built the data feeds and execution layer on top of it.
It helped us get more efficient while also expanding capabilities. Basically, we’re now doing more with less.
We automated existing processes and set up best-in-class research and trendspotting capabilities (seriously, I’ll put our research arm up against anyone).
It helped our cost structure but also genuinely allows us to do better work, and that combo makes the business even more interesting to me than it was before.
3. Expertise still wins
Just because “intelligence” is getting cheaper doesn’t mean expertise is.
AI can write up a content strategy for you, but you won’t know whether it’s actually good or not if you don’t have a content expert by your side.
My bet is that as more companies go AI-native they’ll actually look towards on-demand, hyper-specialized outside experts more than before.
They’ll keep their “team” small and favor generalists to use AI for straightforward tasks, while bringing in outside specialists for complex, niche, and rapidly changing work.
Like, for example, what’s actually worth saying in their niche right now based on the “current thing” and how should it be said to uniquely position them in an interesting way.
The TLDR
A founder doesn't wake up wanting 30 drafts. They want the right investors, candidates, or customers to know who they are. They also want to spend less of their week making that happen.
Those are much harder problems than generating text.
As I made Megaphone AI-native, my services-over-SaaS thesis became more concrete: sell the work the customer wants done, use software to do more of it with less, and keep an expert responsible for the quality.
Talk to me in a few years about whether keeping Megaphone rather than selling it was the right personal financial decision.
What I can say is that the business I've turned Megaphone into makes more sense to me than the version I was preparing to exit.
Get the next one.
One essay every Sunday on how to build leverage in your life and work.

