Why the current live version of Alldone.app is embarrassing
Every founder or product manager will tell you that you should release your product as early as possible to have real customers use your product so you can learn and iterate based on the reality on the ground and find your product-market-fit. Every testing you can do before you release your product into the wild can also be very helpful — but nothing beats the reality check of actual users using your product every day. This is why you typically define and build a product version which you consider an “MVP” — so the Minimum Product version required to be “Viable” in the market. The Cambridge dictionary defines “viable” as “able to succeed”.
Typically MVPs are build for the target group which the business case is based on — so basically the first 100 or so customers you have in the business case in the first few months. The problem with those users is that they typically don’t know you and rightfully expect a very polished experience. So when they use your MVP — and it does not work as intended — it is really hard to differentiate between the 2 sources of failure:
- Was the approach not good enough …
- … or didn’t it work because the idea wasn’t done in good enough way?
In order to avoid the second reason you need to polish every feature a lot which slows down your learning & iteration cycle considerable in your quest to find product market fit. This is especially a problem when you are building your product as a “side-business” in a bootstrapped way without considerable funding and resources at hand.
Here is where I am trying to take a different approach with Alldone.app: I have been building the first “MVP” version of Alldone not for the usual target group of the first 100 random users but for “friends & family” — people I know, i work with and i can interact directly with and do “handholding” :) The hypothesis is that those people will have a much higher tolerance in terms of “look & feel” of the product and therefore I can develop & test out features much faster than for the usual target group. Let’s call this version “MUP” (=Minimum Usable Product) since it is clearly not viable yet in terms of being able to succeed in the market.
This is why the current Alldone.app version looks embarrassing — it is only an MUP made for myself and people i have been personally onboarding. However it works and I have been using it everyday as a daily driver for over a year by now. I even had a team using it which worked better than expected given the current state and lead to the realisation that Alldone will probably have business users as the initial target group and not product people as originally intended. Lots of learnings and even a first pivot right there ;)
After using the MUP to iterate and work towards Product-Market-Fit I now feel confident enough that I want to build the real MVP based on the learnings of the MUP. The MVP will be built by the Cubans as described here. However I am planning to keep the MUP alive while building the MVP to further test out features and approaches. The plan is the following:
- Step 1: MUP Frontend + MUP Backend
- Step 2: MVP Frontend + MUP Backend
- Step 3: MVP Frontend + MVP Backend
So only after Step 3 is completed I will probably retire the MUP version of the product and do normal Iterations & A/B testing in the MVP version.
This will take a couple more months — let’s see if it works :)
You can find all stories about Alldone.app here
