# Derek Bosch, Founder/MD at Stoneframe Digital — read of Margin Protection AI, May 11 2026

> "15 years running a custom dev shop, currently 14 people, two of whom exist because I thought we had better margins than we did."

## How I got here

Googled "scope creep tracking software agencies" on my lunch break after closing a project last week that came in 23% over hours and I had to eat most of it. Third result was a G2 listicle, scrolled to a comment thread, someone mentioned this. Not a paid ad. Just a comment. That felt less gross so I clicked it.

## What I clicked first

The line "Your team is busy delivering work. Scope creep accumulates in small increments." That's real. That's my actual week described in two sentences. So I kept reading. The hero also had "margin gambling" which felt like someone had been in my backroom conversations. It pulled me in fast.

## Where I paused

The stats block. "2.3% Average portfolio margin lift. 8 days Earlier scope drift detection. 18 Change orders captured per month per team. $22k Average annual margin recovery per client." I stopped and stared at these for a while. They're specific enough to seem real but there's no source. No "n of 47 agencies over 6 months." No customer name. Just the numbers floating there. I'm not saying they're made up. I'm saying I've been in enough pitch meetings to know that four clean stats with no denominator is a slide deck move, not a data move.

## What I distrusted

Scrolled down expecting to see integrations. What tools does it connect to? Harvest? Teamwork? Jira? Linear? Clockify? QuickBooks? There's zero logo section, zero "works with" list, nothing. Just "Connect your project management, time tracking, and accounting systems." That's the vaguest possible promise for a product that lives entirely on data integration. If it doesn't connect to my stack natively, the integration burden lands on me and my developer and now I'm paying for setup time I was told I wouldn't need.

Then I hit the honest disclosure section and everything shifted. "We don't have live customers on this idea yet. We shipped the strategy package; you ship the customer conversations." Wait. This is not a product. This is a business idea for sale. I had been reading a product page for three full minutes and only found out it was a blueprint shop at the bottom. That's a bait-and-switch even if it's technically disclosed. I don't want to buy a strategy doc for $5 or a code starter for $99. I came here to buy software that protects my margins. Those are completely different problems.

## What would convince me

If this were a real product in market, I'd want one case study from a dev shop or agency with 10 to 30 employees. Not "professional services." Specifically: what PM tool were they using, what changed in month one, and what did the founder say when they saw the first alert fire. That's the memo I'd forward to my ops person.

For the actual thing being sold here (a build blueprint), I'd want to see a screenshot of the working code starter before I spend $99 on it. Not a demo video. An actual screenshot of the repo structure with a README that tells me what's wired and what isn't.

## What I'd ask in an email reply

1. What specific integrations are live in the code starter? Not "planned." Shipped.
2. Has anyone actually built and launched this from your package, and if so can I talk to them for 15 minutes?
3. The Fermi math shows negative $20k year-one take-home. That's honest, but it means this is a long-haul product play. What's the assumption that swings it positive and how confident are you in that variable?

## Verdict: on-the-fence

The problem framing is the best I've seen on a page like this. But I walked in looking for software and found a strategy blueprint shop, and that gap is big enough that I'd need a conversation before I'd move. The honesty about having no live customers is genuinely refreshing and almost earns a reply on its own.

---
*Memo by skeptic persona, generated 2026-05-11. Studio breaks own self-grading loop.*