Why we built a /services page

Up until tonight, tarmaclabs.org was a single-arm company surface. The home page sold Trim and AutoBrief. The portfolio showed past work. The journal told the story. The tools page proved engineering credibility. But there was no public route that said here is the other half of what we do.

/servicesis that route. It's the surface for custom-build engagements — companies that need software shipped by someone who already has a 132-test rules engine running in production, not someone learning on their dime.

What changed

The positioning didn't change. The communication did. TarmacLabs has always been "an aviation-based software company first that also leverages AI to provide business solutions to other businesses." That sentence got locked the night someone asked me for a business card and I realized I needed a one-line answer for what the company actually does.

Two arms: aviation products first (Trim is live, AutoBrief is live), AI-leveraged custom builds second. The custom-build arm has always been real revenue — it's just been word-of-mouth. /services puts it in writing so a prospect who lands from a journal post or a tools-page link can see the path: paid discovery → SOW → scoped engagement → delivery.

You don't build a paid-discovery offer when you're desperate for work. You build it when you have enough demand to need a filter.

The three tiers

The page shows three engagement sizes, anchored against the pricing research I did with the CFO advisor a few nights ago.

Foundation is $15–25K — a single bounded build, 2–4 weeks, one decision-maker, one acceptance test. Think: replace a clipboard process, automate a recurring report, ship a real tool that one team uses on Monday. Most engagements fit here.

Transformation is $30–75K — multi-module build, 6–12 weeks, integration into an existing stack, requires real product thinking. This is where AutoBrief-shape work lives if a school wanted a customized variant.

Partnershipis $100K+ — long-running, embedded, joint roadmap, the kind of engagement where I'm the de-facto engineering partner for a non-engineering company. One slot per year, maybe two if the fit is right.

The paid-discovery gate

Every engagement starts with $1.5–3K of paid discovery. Two reasons.

First, free discovery is a tax on the wrong customers. People who won't pay $2K to discover whether a $25K build is the right shape almost always won't pay $25K either. The $2K is a self-selecting filter, and it pays for the legitimate hours that go into scoping a real SOW (interviews, audit of existing systems, written deliverable, fixed price for the full build).

Second, discovery is real work. The deliverable is a written SOW with milestones, acceptance criteria, and a fixed price. That document is worth $2K to the buyer whether they hire me for the build or not — they can take it to another vendor. Most of the time they don't, because if I've done the discovery right, I'm already the cheapest path to execution.

What we'll say no to

The /services page has an explicit no-fit list — not as a flex but because saying no is the entire purpose of a positioning page. We don't do: pure design work, marketing sites, app stores compliance, hourly maintenance contracts, anything that needs a team larger than one. If the work doesn't leverage what we've already built — pure-function rules engines, FAA-data-only pipelines, AI-leveraged business automation — then it's a worse fit than our actual competitors.

That's not modesty. It's economics. The only way TarmacLabs ships great work at the rates on /services is by reusing the patterns that already exist in Trim, AutoBrief, and the eight free tools. A custom build that has nothing in common with any of those is going to cost the client more and ship slower than if they hired someone whose primary business is that other thing.

The cadence

One to two new engagements per quarter. That's the capacity gate. Anything more dilutes the products, and the products are why anyone wants to hire us in the first place. The whole pitch is "the company that shipped this also ships custom builds." If the products stop getting better, the pitch stops working.

So: /services is live. If you're reading this and you have an aviation or AI-leveraged business problem that someone with a CFII and a Vitest test suite could solve in a quarter — say hi.