You could build this. Here is what it would cost.

Genuinely — it is not complicated work, just a lot of it, forever. These are the numbers three customers gave us after trying.

Build it yourself

Two engineers, first year, fully loaded

Initial build · 14 weeks
$118,000
On-call rotation
$34,000
Infrastructure
$21,600
SOC 2 evidence work
$26,000
Maintenance · 20% of a role
$38,000
Features not shipped meanwhile
unpriced

First-year total

$237,600

Use ours

Scale plan, 25 seats, same first year

Subscription · 12 months
$14,280
Implementation
included
Infrastructure
included
Compliance evidence
included
Integration work · 2 days
$3,400
On-call
ours

First-year total

$17,680

Not included: your own data modelling, and the meeting where everyone argues about metric definitions. That one is on you.

Build
$237,600
Buy
$17.7k

Assumes $140k fully loaded engineer cost, EU rates, and no rewrite in year two — which there usually is. See the full workings.

">

The prompt

Design a buy-versus-build comparison with Tailwind CSS. Use a max-w-4xl section with a heading, a paragraph acknowledging that building it is genuinely possible, and two columns of equal width. The left column, headed "Build it yourself", is a bordered card listing six cost lines as rows with a label and a figure — engineering time, on-call, infrastructure, compliance work, ongoing maintenance and opportunity cost — with a hairline before a bold first-year total. The right column, headed "Use ours", is a neutral-900 card with the same row structure but far shorter, ending in its own total in white, plus a small line noting what is not included. Below, a horizontal bar comparison showing the two totals to scale with figures at the ends, and a closing line in muted text stating the assumptions behind the estimate with a link to the workings.

Paste it into Claude, Cursor, v0 or any AI coding tool. Prefer the finished markup? .