Key Takeaways
- Tenex pays software engineers strictly by completed story points instead of billable hours, linking compensation directly to delivered code.
- Pay across engineers at Tenex shows a 4x spread between the highest and lowest earners, with top performers clearing over $1,000,000 per year.
- Traditional agency models punish developers who speed up using AI, since finishing work faster reduces the total hours billed to clients.
- Tenex pairs developers with AI strategists tied to client retention to stop point inflation, sandbagging, and bad pull requests.
- The model faces difficulty with non-coding duties like client discovery and system architecture, where story points fail to track value.
The Hourly Rate Punishes Fast Engineers
CJ Hess hit a wall as a freelance developer. As he adopted modern AI coding tools, his work accelerated. He finished projects in a fraction of the usual time, solved client problems, and immediately watched his billings collapse. When you charge sixty or one hundred dollars an hour, doubling your speed cuts your paycheck in half.
“I'm a software engineer by trade and I just felt my output getting better and faster and ultimately getting paid less,” Hess explained. “I was charging hourly to my clients.”
Most engineering consultancies operate on this backwards incentive. If an engineer spends forty hours manually writing boilerplate that an AI-assisted developer finishes in three hours, the agency and the slower developer pocket more cash.
Tenex, founded by Alex Lieberman, threw out the hourly model. The firm pays engineers based on story points delivered rather than time at a desk. If you clear forty points in a week because your workflow is dialed in, you get paid for forty points.
Lieberman observed that this approach exposes raw skill differences: “There's a very wide dispersion in earning capacity between the highest paid engineer at 10x and the lowest paid engineer at 10x. It's probably something around 4x difference.” Top performers make over a million dollars a year by shipping at volumes traditional agencies never thought possible.
The Battle Over Points and Architecture
Tying compensation to story points introduces obvious operational traps: sandbagging estimates and dumping low-quality code. When cash attaches directly to completed tickets, an engineer has an incentive to claim an easy bug fix is worth eight points. They can also prompt an AI to generate sloppy code and file rapid pull requests just to rack up points.
Tenex counteracts this by pairing developers with AI strategists. The strategist manages the client account and gets evaluated on client retention. If a developer bloats points or ships broken code, the client leaves, and the strategist's pay drops. The strategist acts as a check against ticket stuffing.
The bigger structural headache is non-coding work. High-level developers spend hours thinking through data structures, talking to stakeholders, and setting architectural boundaries. That effort rarely maps cleanly onto a story point.
Lieberman noted that this remains their toughest puzzle: “Actually the biggest challenge has been how do you compensate engineers for all the time that's not spent coding whether it's architecting meeting with clients.”
Hess pointed out that raw coding represents a shrinking slice of the job anyway: “I was coding a lot coding a lot coding a little less and now we're kind of getting to the point where the models are more or less given the right context and can get you to 90% of the solution.”
When AI produces 90% of the code, a system based purely on code-delivery tickets risks overvaluing the final 10% of syntax while underpaying the thinking that scoped the problem.
What to Do With This
Audit your agency contracts or contractor agreements this week. Pick one upcoming milestone and price it as a fixed-ticket bounty based on difficulty rather than an estimated hourly budget. If you pay contractors hourly, calculate what your fastest engineer would make under a bounty model to see if your current setup penalizes your best talent.