Become an engineer who PMs love to work with
serverless paradigm from product managers' point of view
Product managers are mini CEOs of products in a company, who decide what to build and are responsible for the success and failure of a product feature at the end of the day. As engineers we may not understand their point of views, they seem to ask this and that and we do all the hard work; but since they are finally responsible for a product’s success and failure, we may also have ignored the pressure they are under: one decision mistake could cause the loss of million dollars (either the real cost or opportunity cost) — especially for the PMs who are higher up in the product scope.
In this post, we are going to discuss how the PM may think of building products in a serverless paradigm, using several great (and well-respected in their community) product managers who developed (and keep developing) their product management theoretical system.
Nobody really knows whether a product will make a kill or get buried in graveyard before the real release
Here is a Google product graveyard we can get when we do a Google “Google product graveyard”:
Same thing happens to Microsoft, and here is the list created by Alberto Savoia, who coined the product management methodology “The Right It”
The take away is both of the great software companies made mistakes that cost millions dollars because the law of market failure that Alberto referred to.
If product failure is inevitable, how can PMs increase their odds of building a successful one?
As an engineer, I really liked the way that Alberto transformed product management process, which typically was regarded as some art-like procedure, into somewhat science-like, by giving a Bayesian analysis to model the process (if you want to entertain a bit math, please watch this video). From product management perspective, the methodology outlined how to progress the decision making:
I am not going to go into the details of the above Bayesian formula, if you are interested, please watch video above he gave — super interesting and extremely straightforward.
In his thesis and proposed methodology, the key to build a successful product is that we should be able to collect data at each step of our product building to increase the odds that we are building a successful product or feature (or discontinue it if the odds are against us): from concept forming stage to customer feedbacks collecting stage.
How to collect data with an MVP release in a least expensive way: consider the serverless paradigm
For building an SaaS, rolling out an MVP may be an inevitable step to collect the data. As we wrote in a previous post, using the serverless paradigm to develop a product feature may shorten the product development cycle (and therefore the development cost) 10x. If a product manager hears that an MVP can be built in 1 week instead of 1 quarter, then that’s the music in his/her ears. The reason is that it not only reduces his data collection lead time but also sometimes gives him a new tool to collect real data that otherwise would not be available.
Even though not necessarily every product is suitable to be built in a serverless paradigm, the gained velocity of being able to collect the data for product managers to make further data-driven decision (by the way, most SaaS product features are very much suitable to the serverless paradigm). This is why an engineer or an engineering team knowing how to build a product (or at least MVP) in the serverless paradigm, i.e., being able to roll out an MVP in an inexpensive way, will likely make good friends to their product manager peers — you give them the weapon to make smart decision. Even though the data may show that the current feature is not the right one, in which case we can quickly make decision to reduce the sunken cost, we give him/her exactly the weapon that he/she can build the “the right it”.
In the next post, we will talk about what complexity and needed team member supports a serverless paradigm app will reduce from a docker / k8s/ Django (or Spring boot) stack. If you want to follow, please subscribe our newsletter 👇
Summary:
Product managers need to collect data to make smart decisions, and being able to build initial product in an inexpensive way is the key to collect the data for the product manager;
Serverless paradigm gives us an inexpensive way to deliver the MVP, so keeping that under your belt will very likely make you an engineer or engineering manager that product managers are willing to work / collaborate with.
Please leave the comments or share with your friends if this may relate to their day-to-day work.