What do you think is unique and valuable? Test THAT!

Q I've an startup idea. It is 40-60% same as of others, but a company started 2 years back which is 95% same as mine. It has >10M users but not even 0.1M active. I've little better idea & we could implement in better way. But our MVP won't be enough to distinguish b/w us. So what should be approach?



A What do you mean your MVP will not be enough? What is it about your solution is the unique part? Isolate that as an independent variable and test for that. If you can tell me the solution, and the unique part of your solution (even if that unique part is your team, or a single feature, or your acquisition strategy) I will design an experiment for you that proves that specific thing.

Elaboration for Bhuvaneesh Kumar J J

One of the biggest problems with the term MVP is the P. The PRODUCT. When you have a product in mind you automatically think you have to BUILD something and an experiment is conflated with a prototype. In your case, you already know that your solution is not 100% unique, for which others on this thread have rightly given you kudos. But you DO have an implicit assumption that something about your solution is superior to the competition. For something to be better, by definition it must be different, even if only slightly (think about how humans and apes share 99% of their DNA). 

As for the experiment, my opinion, and the purpose of experimental entrepreneurship, is to try and run an experiment that proves your hypothesis with as few resources (time, money, energy, relationships, etc) as possible, while still preserving the same outcome (leanings) of "building the whole thing". 

For example, if your hypothesis for Quora was that it was better than Yahoo! answers because of the credits system, then you would set up an experiment that controlled for the credits system and you would test for that. 

There are tons of nuances of the science, controlling for the right things, getting the right data sets, making the right inferences, knowing that you are trying to disprove rather than prove your hypothesis etc. 

But at the end of the day, it is relatively simple to isolate what you think is the UNIQUE value prop of your team, or your solution, and testing that thing independently to see if the "market agrees". 

You will need to be careful about false positives, false negatives, and a whole host of other beasts, but you will reduce the risk of building something that no one wants or over betting the pot on an assumption that you could have disproved with a 1 day experiment. 

Now I ask you this, what is your UVP. What do you want to control for?

Post a Comment

Popular Posts