Writing
Notes from 20+ years building B2B software companies.
Mostly about growth, product, GTM, hiring, AI, bootstrapping and mistakes I've made along the way.
I had to end my $5M ARR business
18 months trying to make a 12 year old help desk agentic, and the architectural reason it was never going to work.
The four layers of an AI-native product, and the only one that's a moat
Data, context, model, outcomes. Everyone is competing on the layer they rent.
12 months to rebuild 12 years
What AI-native costs when you take it literally, and which parts of a 12 year old product turned out to be worth nothing.
All 10 of my engineers code with AI. This is not vibe coding.
The difference between AI-assisted engineering and the thing that phrase has come to mean.
Overnight migration is a feature now
The switching cost that protected incumbents for 15 years just collapsed, and most of them are still pricing like it didn't.
We deleted per-seat pricing
$0 per seat, $0.50 per resolution, and everything that broke internally the week we shipped it.
Deciding what not to charge for
The billing exclusions were harder to design than the prices, and they're the reason the pricing page survives a hostile read.
50 calls before the requirements doc
I owned a help desk for 12 years and still got the first 15 calls wrong.
I spent $100K of my own money to demo a product that didn't exist
SaaStr, a gold sponsorship, and 125 demos of software that wasn't ready.
What a $100K booth actually buys
The full accounting from SaaStr, including the line items most conference recaps leave out.
Copying Benioff's No Software campaign 26 years later
Pandas, sign holders at Gate 15, and what a 1999 playbook still gets right.
Reach was never the point
456 likes, 33 ICP-fit engagers, 24 net new conversations. The post does the qualifying.
We burned our sending domain in a week
Post-conference outreach volume took helply.com to Bad in Google Postmaster. The diagnosis and the ramp back.
Three years flat at $5M ARR
What a plateau feels like from inside the company, as opposed to how it reads on a chart.
Running a declining business on purpose
Groove funds Helply. 85% of R&D, 100% of marketing. Managed decline as a capital strategy.
Writing my way to the first customers, twice
The Journey to $100K MRR series built a 50K newsletter and nearly all of Groove's early revenue. Doing it again in 2026 looks nothing like it did then.
Hiring for a product that doesn't exist yet
When the honest pitch is: we're throwing away the thing that pays your salary.
Buying out a co-founding partner before launch
Significant equity, pre-revenue, no clean playbook, and what I'd structure differently at the start.
Owning 88.02%
What majority ownership actually changes about how you make decisions, and why $10M and $100M aren't the same target.
The exit that didn't finish anything
A team of 10, a $15M acquisition, and the thing nobody tells you about the year after.