Thread by @cyberomin

A strong product manager should be a person with a balanced view of business (requirements) and engineering, else there's bound to be problem and this could lead to disintegration in the team. S(?he) must know how to manage stakeholders expectations in relation to timeline.

In the quest to move fast and break things, product managers unconsciously encourage engineering teams to throw caution to the wind and ignore the fundamental tenets of software development. This practice only leads to frustration and unmanageable technical debts.

Product managers do this in a variety of ways:
1. Change features direction mid-sprint without any change in delivery timeline.

2. Demand half baked features to be released into the wild.

3. Get people to build not well thought-out ideas, only to cancel it before it ships.

There's a reason Agile is the most successful software development methodology today. Sprints, scrums, retrospectives aren't waste of time. They add to the bottomline. They help improve the overall health of the product.

There's a reason why engineers spend those awful hours writing tests and adding monitoring to the infrastructure. It's to allow everyone sleep well at night. I will paint a scenario why these tests are important. Grab a bucket of popcorn.

Payment gateways typically collect payments in the smallest unit of a country's currency; Cents for the US and Kobo for Nigeria. When payment is made say NGN50, the developer is essentially charging the user's card N50 X 100Kobo = 5000K.

If it so happens that there's no unit test at least to check the currency conversion, bad things can definitely happen. Imagine for a second that the developer missed the conversion by a zero(0).....

...What this means is that, for every supposed N50 paid, he/she charging the customer 50 X 10 = 500. Converting this to Naira gives us N5. So, for every N50 item, the business s losing N45...

...if this goes unchecked for about 1000 transactions, the company would have lost N45,000. All of this happened because someone missed a zero. This isn't fairy tale, I have seen it happen before and it isn't always pretty.

So, dear PM, while you're getting these awesome features that needs to go love yesterday, take the time to explain to the stakeholders why if not built correctly, bad things can happen. The best way to make them see reason is to show them the impact in Naira & Kobo.

Let the stakeholders understand that writing tests and adding things like monitoring will, at the end of the day, five everyone peace of mind, and most importantly, someone wouldn't have to explain avoidable losses to the board.

The end.