«ИИ за вечер сделает»: три метода оценки показали, что 300 часов — ниже плинтуса

The author of a Habr article estimated the refinement of a corporate accounting system at 300 hours — the client replied that this was inflated, hinting: "AI will do it for me in an evening." In response, the developer recalculated the same scope using three independent methods and got a result the client did not expect.

Three Methods — One Conclusion

The developer applied three standard approaches to the same task scope, without relying on any single one of them.

• Bottom-up by user actions — a detailed breakdown of every end-user operation in the system
• Function points with ISBSG benchmarks — the scope is converted into function points and compared against the ISBSG (International Software Benchmarking Standards Group) database of thousands of real projects
• COCOMO II — an algorithmic model from the University of Southern California that accounts for scale, logic complexity, and team characteristics

All three methods produced consistent results. According to Habr, each approach showed that 300 hours is not a ceiling but a lower bound: the figure turned out to be below average industry values.

Why 300 Hours Is Too Few

The ISBSG database accumulates production data from thousands of real corporate software projects worldwide. The average team productivity when developing accounting systems is significantly lower than it appears from the outside: function points account not only for writing code, but also for requirements analysis, integration with existing modules, and regression testing.

"300 is below the industry floor," the article's author concludes

after calculating with three independent methods.

COCOMO II adds complexity multipliers for business logic and technical debt. For corporate accounting systems, where every change affects adjacent modules, these multipliers increase the final estimate — not decrease it.

Where the "AI in an Evening" Argument Goes Wrong

Modern AI tools do genuinely speed up writing boilerplate code, but they do not eliminate the key stages: auditing the existing architecture, requirements analysis, integration and regression testing. These stages account for the majority of the actual scope when enhancing corporate systems.

The client's mistake is typical: AI is visible in a quick demo ("a prototype in an evening"), but invisible in what the prototype lacks — reliability, compatibility, and data security. The ISBSG and COCOMO II methods explicitly account for these hidden stages.

What This Means

Three independent methods gave a unanimous answer: 300 hours to refine a corporate accounting system is a realistic lower bound, not an inflated figure. The case shows that a well-grounded estimate backed by industry methodology is a working response to the argument "AI will do it faster."

Frequently Asked Questions

What Is ISBSG?

ISBSG (International Software Benchmarking Standards Group) is a non-profit organization that collects production data from real development projects around the world. Its database allows you to compare a function-point estimate against the industry median and justify it to the client based on statistics rather than intuition.

How Does COCOMO II Differ from Bottom-Up Estimation?

COCOMO II is an algorithmic model that produces a probabilistic estimate taking into account project scale, team characteristics, and technical risks. "Bottom-up" is a detailed decomposition by specific tasks. When both methods produce similar results, this removes disputes about "expert subjectivity" and increases confidence in the estimate.