Why do we pay for so much code when most of it is not used?

There is a lot of discussions regarding how we should work faster and the impact of AI. Really good ones. But if we look at it, how efficient are we today?

by Fredrik Kangevall

60-80 % of all code is never used

Most software never gets to touch a moving vehicle. In traditional development, between 60% and 80% of all written code is estimated to create zero or even negative value for the customer. If it even reaches production at all.

To me, that’s insane. Stop investing money that doesn’t make a difference.

How much software was written that never ended up in this truck?

Imagine doing the same thing in real life. You decide to renovate your bathroom. Build it. Never use it. Tear it all down. And start over. That isn’t just wasted money in the moment. It is about so much more.

How fast will a project progress when half the effort is spent on unnecessary work? What does that system cost to maintain over time? And how do you keep your team members motivated, pushing hard against tight deadlines, building features no one will ever use?

We want every line we deploy to be running in a real vehicle, out on the road, as soon as possible.

You have just improved you capability, so what do you do?

How? By creating an environment where every change has a purpose, iterations are frequent, and deploys are kept small and controlled in scope — allowing us to see, get feedback, and refine.

How much of the code from your last project is actually running in production today? And how do you ensure you aren’t paying for features no one will ever use?

Key take aways

Benefits

• Potential cost savings in development: 60-80% of code in traditional projects are never used, huge potential for cost savings
• Improve your time to market: Spend your time on things that add value, ignore the rest
• Reduce your life cycle cost: A bigger system costs more to maintain and support

To consider

• Keep team members motivated: how do you keep your team members motivated, pushing hard against tight deadlines, building features no one will ever use?