· 2 min read
What I removed from Optiva, and why
Tax return, invoices, payroll, timesheets: three weeks of work removed from the product over one summer. Not because it didn't work. Because nobody needed it.
- Method
- Product
Mid-June, I add a full Australian module to Optiva: tax return, payroll, leave, timesheets. Tables, screens, calculations. I am pleased, the product "does more".
On 4 July, I delete the tax return and invoice creation. A bit more than 2,000 lines gone, a migration that drops tables. On 13 August, payroll and timesheets follow.
What triggered it
A remark from my business partner. Our clients already have Xero, MYOB or QuickBooks, and Optiva already sends them the data. So why rebuild on our side a tax screen their accounting software does better, with a real lodgement to the tax office that we didn't have? We were doing the work twice, and the second version was the worse one.
Optiva is not accounting software. It is the automation layer that works with it. I knew that. And I still built the opposite for three weeks.
Why I fell into it
When I build, I tend to push to the maximum. One more detail, one more feature, one more case covered. And when my head is too deep in it, I lose my way: I build what would be nice to have instead of what clients actually need.
It is not a lack of skill, it is a lack of attention. The code was clean, tested, working. It was just useless. And useless code in a product has a price: screens to maintain, possible bugs, a longer menu for the user, a blurrier message for the prospect.
What I changed in the way I work
Since then, the decision is written in black and white in the project's roadmap: Optiva automates on top of the ledger, never replaces it. When an idea comes in, it goes through that sentence before it goes through the code.
And I kept a more general rule. What is not alive has to go. A screen nobody opens is not neutral, it costs. Deleting is a feature like any other.
For you, if you are having a tool built
The question to ask is not "could the tool do this?". The answer is almost always yes. The question is "who would do it, how often, and instead of what?". If the answer is vague, it does not go into the first version.
That is also why I start every project with a written scope, with what is in and what is out. That list of what we are not doing cost me three weeks to learn. It is now part of every scope I write.