Technical Debt Calculator

Measure the interest, not the mess. Count the hours the team loses each week to workarounds, slow builds and fear of touching the code, and price them at fully loaded cost.

Compare the annual interest against the cost of paying it down. A payback period under a year is a straightforward case; a payback measured in years usually means doing it alongside related work instead.

Because it recurs whether or not you act, and it grows if the debt does. The metaphor only becomes useful when both sides are actual numbers rather than adjectives.

Track it for a fortnight rather than guessing. Ask people to note time spent on workarounds, waiting for builds, and re-running flaky tests — the total is usually higher than anyone estimated.

No. Debt in code nobody touches costs nothing to carry, which is why the interest matters more than how bad the code looks. A payback of several years is a legitimate reason to leave it alone.

Usually not. Debt paid down as part of feature work in the same area is easier to justify and easier to schedule; a standalone rewrite is the highest-risk way to spend the principal.

Anything where a past shortcut makes present work slower — missing tests, duplicated logic, a framework nobody understands, a build that takes twenty minutes. Code that is merely old or unfashionable is not debt.

Frequently. Shipping something imperfect to learn whether it matters is a sound trade, and it is only debt if the shortcut is recorded and the interest is watched. Unacknowledged shortcuts are the problem.

With this arithmetic rather than adjectives. An annual interest figure and a payback period are the language a budget conversation runs in; 'the code is bad' is not.

Usually not. Incremental payoff alongside feature work in the same area is lower risk and easier to schedule, and a standalone rewrite is the most dangerous way to spend the principal.

It compounds, because workarounds built on workarounds get harder to remove. The growth rate is an input here because it varies enormously — some debt is stable, some accelerates sharply.

Measure it. Ask everyone to track hours lost for two weeks and the disagreement usually resolves itself, because the areas actually costing time are rarely the ones people complain about most.