Tech Debt
How to balance your time & resources for systems
This is an official term that describes how you get stuck with the technology you choose.
For example, if you've built your systems in spreadsheets, at some point they will be too complex to manage, or too heavy to use.
This is true for full blown enterprise software or apps too, where the debt is usually created by complexity of code or lack of clarity on what features are really needed.
It's good to be aware of this from the start, but people usually find out after they've spent a lot of time, money and effort to build something. Then you are stuck. The only choice you have is to either fix it or start from scratch. Both look equally difficult, without spending much more resources on it.
So, how to combat this?
In my experience the following things help immensely (read from bottom to top if you're already deep in tech debt :)
Start with only what's really necessary. This will keep the system and complexity to a minimum. Really choose only the minimum. Even if it's easy to create, remember it has to be used & managed forever!
When you add on more features as required, make sure they're systematically added, not random. This is the biggest reason for tech debt. You may think you need to add something urgently, but this is like adding wheels to a car. It will not actually make it faster or more efficient in the long run.
If you already are deep in tech debt, the best way is to simplify and document. If systems were made randomly, it may also call for a system wide restructure. For e.g., instead of having 10 sheets for accounts, a single system that covers all 10 aspects would be much more effective. The best example I can give for this is Google. They had 20+ products like Gmail, Chrome, Android, each with their own UI. It was difficult for them to manage and confusing for users. That's when, they basically stopped all scattered work and created a single material design system that we use accross all of Google's products today. It became extremely easy for them to manage - even for brand new products and consistent for users. Similarly, IBM, Microsoft, etc.. all have gone through this.
One more aspect to take care of us not to go overboard with optimising too early on. You will waste your time. Do it only when your systems get painful enough, otherwise you will waste your time & resources on systems instead of focusing on your work.
The best way to create systems is:
Make it work
Make it right
Make it fast
Which stage are you on?
Cheers 🔥