How technical debt actually gets paid down

Sample article. This piece exists to demonstrate formatting, headings, and typography for the writing section. Replace it with real writing, or delete it — it is not one of Dave's actual published views.

Every engineering organization has technical debt, and nearly every engineering organization has a version of the same argument about it: the team wants time to fix it, and the business wants to know why shipping should slow down for a problem no customer has ever mentioned.

Both sides are usually right about their own part of the argument, which is why it rarely resolves through debate alone.

Debt is a financing decision, not a moral one

The word "debt" is doing real work in that phrase. Debt isn't inherently bad — it's a financing choice, made under time pressure, that trades future flexibility for present speed. The mistake isn't taking on technical debt. It's losing track of which debts were taken on deliberately, and which accumulated by accident.

Treating it as financing changes the conversation. Instead of "we need to fix this because it's bad," the conversation becomes "this loan is now costing more in interest than it's worth, and here's the payoff plan."

Make the interest payments visible

Most technical debt doesn't get prioritized because its cost is invisible to everyone except the engineers absorbing it. A slow test suite, a fragile deploy process, a data model that requires tribal knowledge to touch safely — none of it shows up on a roadmap, but all of it shows up in velocity.

A useful exercise: for each significant piece of debt, name what it's actually costing today — in calendar time, in on-call load, in onboarding time for new engineers. That's the number to bring to a planning conversation, not "code quality."

Pay it down inside the work, not around it

The version of debt paydown that survives contact with a real roadmap is rarely a dedicated quarter of cleanup. It's a standing discipline: every project that touches a fragile area leaves it a little better than it found it, and every planning cycle explicitly reserves some capacity for paying down the debt that's now costing the most in interest.

That only works if leadership treats it as non-negotiable capacity, not a nice-to-have that gets cut the moment a deadline tightens.