When to MVP?
One thing that’s perennially top of mind is feature / product scope, what is an MVP, when to ship fast, and when to wait to ship, etc. There are so, so many product best practices; A/B testing, JTBD, solution validation, KPIs, impact/reach prioritization frameworks, proof of concept, etc. If we used all of them it’d take up a huge amount of time - so how do we choose?
In general anything we build has a pretty long shelf life. We’ll be supporting and iterating on things we ship for a while. This is different from say a consumer internet company where they A/B test heavily and throw away products / work when things don’t get traction (Twitter Fleets!). There are two mental models that I think are really useful in this context. Both are related to thinking about value and risk. First, is from Casey Winters CPO @ EventBrite, Product Risk & MVP Mindset) (great article, highly recommend reading it), and talks about ambiguity in the problem and knowing how it’ll help our users.
Shipping things in phases helps with two areas key types of product risk. Do people want this , and can we build it fast / well enough? When we build a small thing first, it allows us to de-risk and understand the complexities of what we’re building. This is hard without production usage. It also helps with compounding user learning, as real merchant usage is the highest bar. Key here is small incremental milestones that are shippable to merchants (Beta’s count!).
Second mental model I like is one about value and thinking about a portfolio of features / work. People at Square call this “Big Rocks and the Pebbles”. I like to think about it like Brandon Chu (GM Shopify Platform) does in PM Mental Models. There’s an inherent Time Value of Shipping. If you only look at the benefits of a feature, then you usually pick the biggest ones. However you do have to invest in the big things as they compound more over time and are more valuable as the time horizon increases. But we can’t always build those things as they take a long time!
Related to the value of shipping are different types of product development and the need for various types of tradeoffs (speed vs quality). Experiments in this context are usually related to high traffic funnels like onboarding, consumer experiences, etc. In general I don’t think that A/B testing a big feature is worth it, unless there are big business risk bars and we have to be sure that we don’t break something. Platform investment needs to focus on quality due to the leveraged impact that it has on the product portfolio, developers, etc.
Hope this helps!