# Marcus Tillman, Senior Engineer / Sole Proprietor at Tillman Dev LLC — read of custom-app-development-agency, June 19 2026

> Nine years writing code for other people's companies. Been freelancing solo for two years. Thinking seriously about going agency-shaped. Two kids, Saturday soccer coach, drive 35 minutes each way to a coworking space in Bellevue because I can't work from home when school's out.

## How I got here

Googled "how to productize a web dev agency" on my lunch break Wednesday. One of the results was some kind of idea marketplace I'd never heard of. Clicked through skeptically. The domain was wishdeal.com and I had never heard of them. The specific URL landed me on what looked, for the first eight seconds, like an actual agency's homepage.

## What I clicked first

The hero. "Bespoke Web & Mobile Apps / Custom software designed and built for businesses that demand precision, performance, and a partner who understands your vision." I thought I was reading a real agency. Portfolio section, three case studies, process section. Looked fine. Standard. Then I kept scrolling and hit the score panel and did a full stop. This is not a real agency. This is a template for one, being sold to me.

That realization scrambled the entire first half of the page retroactively.

## Where I paused

The Fermi numbers. Hard stop. "$-1,585 Year-1 take-home." That is a negative number. They put a negative number on the page in a box labeled "Year-1 take-home" and are still selling this to me. And then right below it: "1 in 6 Meaningful-success odds." 

I respect the honesty, I genuinely do. But I am sitting here staring at a product pitch that opens with "you will probably lose money and have a 17% shot at something meaningful" and somehow expects me to pull out a credit card. The Fermi estimates deserve a much longer look because either they're a gimmick dressed up as rigor, or this idea is genuinely not worth pursuing and they're just selling the dossier anyway.

## What I distrusted

Two things. First, the portfolio. "SaaS Analytics Dashboard, Real-time reporting platform for fintech startup. React, Node, PostgreSQL." No company name. No link. No logo. No "the client asked us to keep them anonymous." Just three one-liners that any person could type in twenty seconds. These case studies are mockups for a template, not proof of work. They are props inside a fake storefront.

Second, the scoring section itself. "buyer clarity: 10/10" and "speed to mvp: 9/10" are the two strongest axes, which is interesting because those are the easiest things to score well on a generic agency idea. Of course anyone can understand what a dev agency does. That tells me nothing about whether I specifically can build one profitably. The two worst axes are "financial upside: 2/10" and "pain intensity: 4/10." Those are the only two scores that actually matter for a service business and they're both underwater.

The phrase "We shipped the strategy package; you ship the customer conversations" jumped out as well-written but also a clean disclaimer. It means: we cannot be held responsible for what happens next.

## What would convince me

Show me one person who bought the $99 pack six months ago and is billing clients today. Not a testimonial quote. A Twitter thread, a Substack post, a YouTube video, a Reddit comment, anything with a timestamp and a real handle attached. Show me someone mid-journey, not a finished success story. Rough is fine. Real is the only thing that moves me.

Also: explain the Fermi model. What assumptions go into "-$1,585 year one"? Hours worked, average contract value, how many clients, what's the cost structure? If I can see the spreadsheet I can stress-test it against my actual situation. Right now the number feels like a stat designed to signal honesty without actually being useful.

## What I'd ask in an email reply

1. The portfolio on the page, are those real projects you or a past operator ran, or are they example cases you wrote to illustrate what the business could look like?

2. The $5 dossier versus the $99 adopt tier: what specifically is in the "working code starter" and how production-ready is it? Is this a Next.js boilerplate with a landing page or something more substantial?

3. You scored pain intensity at 4/10. Custom dev is a crowded, commoditized market. What is your actual thesis for why someone buying this in 2026 carves out margin that a random Upwork dev cannot?

## Verdict: on-the-fence

The honesty of the scoring model is genuinely unusual and I respect it more than I expected to. But the negative year-one take-home and a 2/10 financial upside score on a product you're actively selling is a combination I cannot ignore. Five dollars to read the dossier is low enough that I might do it just to understand how the Fermi math works, but not because I'm sold on the idea.

---
*Memo by skeptic persona, generated 2026-06-19. Studio breaks own self-grading loop.*