How minimal should my MVP be?

Aim for exactly 5 people in your first round of Interviews. Schedule a follow up for each one right on the spot. Ask for other similar people they know.

The first MVP should be wireframes / mockups only.

To test the assumption "I understand what they need and the solution I have in mind will solve their Problem"

In the 2nd Interview you ask them to "use" the paper prototype or click trough the Powerpoint mockups... Don't show them - observe their behavoir and assumotions if they see your mockups.

Let them imagine they could only have one of the features. Ask them to name the one feature which they can not live without. Build that as MVP v0.2

First comment, hope I can help.

How deep should the market research be?

I like to validate ideas through talking to people online and creating landing pages. In the past few months I've done that and I always felt that if you building a landing page that isn't promising something that you cannot deliver, this is a pretty good indication of if there is a market for your product or not.

Given, I am a UI Designer / Developer as well so building those minimal good looking landing pages is never an issue for me, but there are plenty of services (or even templates if you want to customize them yourself) to streamline this process even more.

Last but not least, if the project is a "weekend project" or I strongly believe that it will succeed, I will just build it in a minimal way without a landing page.

How minimum should the MVP I'll build be?

This question is harder to answer. I'd say your minimum product can be significantly smaller than you think. An example from my recent experience: I tried to raise $1M for a start-up but failed, so I iterated over the idea to remove more and more features while looking at the final idea again and again, asking myself if it was still useful. In the end I ended up with something, for my test, would be a little too small, so I re-added some features and am now in the process of launching it this week.

I think that iterating over an idea and removing as many features as possible while still making it viable is the key here. This can mean that instead of having a fully fletched SaaS, you may just have a one-page website with Stripe.js, a webhook that auto subscribes people to a private email list, and maybe a 1 page setting page that you can log into just with your email address (though a "magic link email" or similar).

Be creative in the ways and don't let your emotions keep you pivoting a bit if necessary.

Hope this helps!

Congrats on deciding to start something!

1. How deep should the market research be?

In my experience, you will never get definitive results from your market research (to illustrate this; I did pre-orders for one of my products last year and even after over $30,000 in pre-orders I still wasn't convinced the market was good enough). I just a matter of testing the waters and using your gut/instinct to decide wether to proceed. Additionally, it's probably also a good idea to think a little bit about what your long term plans are with this project (creating a side-project generating $2,000 in extra monthly cash vs looking to build a large scale business with employees, etc). One final thought on market research; a common mistake founders make is that market research is a one-time affair which will clearly indicate wether your idea will work; I have found this not to be the case. It's pretty much an ongoing activity which provide with useful data on what the next steps might be.

1. How minimum should the MVP I'll build be?

What your first MVP looks like should be dictated by what you're testing. Keep in mind that the MVP approach is a methodology used to test hypothesis. So you start with an assumption, and then build a MVP to test that assumption. The "minimum" refers to the minimal product needed to test your hypothesis. So how your minimal your MVP should be depends on what you're testing and how you will be testing it. Another common misassumption about the MVP approach is that is has to be an actual product; which is not necessarily the case. The "product" part refers to the result of effort, not an actual product. Although it can be an actual product, it does not have to be. It can also be a survey, an interview or any other means of testing a hypothesis.

It all comes down to formulating a clear assumption/hypothesis first and start from there. What can you create/do to test that assumption?

Hope that helps!