I recently joined a product development team and was told by the lead that the product was 80-90% complete. However, after diving deeper, I discovered that there were significant issues and, alarmingly, no unit tests in place. In my opinion, this situation spells disaster for the product's functionality in production. I'm worried that if users encounter problems, they might cancel their subscriptions or even demand refunds. I feel like this isn't a complete product at all; while it might work for developers during testing, it is prone to breaking easily, which I've already experienced myself.
I brought this concern up with the lead and wondered if it was overboard to say "you don't really have a product because it lacks tests, which will lead to it breaking in production and turning users away."
5 Answers
I feel you, but at the end of the day, a product is defined by its ability to generate revenue. Unit tests help in maintaining quality for sure, but many companies have thrived without them. Telling someone they don’t have a product just because they lack tests could come across as harsh. Focus on suggesting how they could incorporate testing down the line, that's the angle to take.
It's definitely an exaggeration. Many successful companies I’ve worked for didn’t have tests until way after they’d started earning revenue. It’s not ideal at all, but sometimes the market’s demands outweigh ideal coding practices. I think it’s more important to recognize the distinction between good coding practices and what actually constitutes a product in the market.
Honestly, it seems a bit extreme to say it’s not a product just because it lacks unit tests. I’ve seen businesses thrive for years with systems that have almost zero tests. While I personally wouldn’t work without them, the reality is that many operations sell and succeed without following best practices. So, calling it not a product is a stretch—I mean, sold products can exist without tests.
I get where you're coming from, but remember, unit tests are just one tool for developers. You can have a fully tested codebase, and that doesn’t guarantee a working product. It's also about automated testing; end-to-end tests are crucial to ensuring everything runs smoothly after changes. It might just be a case of prioritizing the wrong types of tests.
You know, I helped a startup go from scratch to making millions without any tests. Sure, we didn't run into major issues at the time, but I wouldn't want to rely on that approach. Saying they don’t have a product just because of this might not be helpful. That kind of language could hurt your relationship with the team. It's more beneficial to focus on how testing can improve their future work instead of criticizing their current lack of it.
Right, not every product needs extensive testing upfront. The reality of shipping is complex and sometimes you roll the dice. It's about getting stuff out there and improving as you go.