Bridging the gap on short term debt
14 Oct 2021 - Estimated reading time: 3 mins
On this day 12 years ago, Martin Fowler published a seminal blog on tech debt. In the world of technology blogging, this may feel ancient, but its lessons still ring true today. In his article he describes tech debt as anything that slows down the pace of development, which is one of the many definitions people have thrown around. The concept of tech debt has stuck with us for so long because so many features of debts and loans can be applied to tech debt.
The key concept Martin Fowler introduced was the distinction between reckless and prudent tech debt. Tech debt can be reckless due to total size, risk or high interest, which means that it places a large drag on the team compared to the cost of cleaning it up. As much as possible, we want to avoid any reckless tech debt.
However, the real world does not always allow us to follow theory to the letter. An example of this was modelling changes due to the government’s consultation on RPI. Due to this regulatory change, we had a time-sensitive opportunity to create more value by delivering quickly.
Our team realised that we couldn’t deliver the full minimum viable project with a prudent level of tech debt to commercial deadlines. These delays would have dropped our take-up of the new feature set from 75% to 25% for March 2021 valuations. Over a third of our private sector schemes have this valuation date. We didn’t throw in the towel. Instead we took a step back and worked out how we could phase delivery to capture the most value and realised that we could only make the delivery date by incurring a high level of tech debt.
After pragmatic discussions, we agreed to go ahead with the release on the earlier date. However, we made this contingent on getting agreement from the project sponsor and the business owner that we would commit to cleaning up the tech debt immediately after release. It’s reckless to build up high interest tech debt without any consideration. Incurring it with a clear plan to clean it up is prudent and deliberate.
For our debt analogy, we can look to the property market. If you’re lucky enough to be a homeowner and are moving to a new house, you’ll try to complete the sale and purchase on the same day. If you buy first, you won’t have enough money for the purchase and if you sell first you may end up temporarily homeless. But that’s not the full story. Let’s say you’re buying your dream house and are under pressure to complete before you can sell your current house. Rather than lose your dream, you could take out a bridging loan. This is a short-term loan to tide you over until you can sell your house. It sounds fantastic. However, the downside is that it has a high interest rate and you may take a big hit if you can’t sell your house for a year.
This is exactly what we did with our code. We took out a short-term loan on tech debt to end up on the right side of a valuable deadline with the intention of paying back the tech debt immediately, which is prudent and deliberate in the long term. The challenge is the next Urgent Project. There will always be pressure to push back the tidy up and move onto something more visible to the business owner.
By using analogies like the bridging loan, we build wider understanding of the cost of delaying or cancelling the clean-up. This allows us to have honest and pragmatic discussions about the trade-offs between running a high interest tech debt long term versus the value of a new project. After all, when would you decide not to pay back a bridging loan?