Me: So, what is your product?
Startup Guy: Well, it's a X that does this and that (goes on for 5-10 minutes).
Me: How are you going to make money out of this?
Startup Guy: Once we have enough users, we'll be able to attract ad revenue. In addition we're going to use the freemium model.
If I was an investor, I would probably excuse myself at this point. But given I'm not, I try provide an alternative perspective for the Startup Guy. The approach might work with enough luck but you really don't want to rely on luck in business. There isn't enough of it for us all.
The problem with "enough users" is that attaining that magic amount is not entirely trivial. There is a lot of competition out there and building a "community" is not easy nor cheap in case you want to do that fast. This applies particularly to web apps as Paras Chopra points out.
I have my qualms with the freemium model as well. Even if you managed to attain enough users, how do make them convert? What value do you provide? As Ash Maurya points out in his book "Running Lean", it can be a difficult position, especially in the beginning. Rather provide concrete value and aim for real sales rather than something hypothetical. There are probably some exceptions to this but I'm pretty wary about the freemium model. It's simply used too much these days.
As Paras does in his article, I believe as well that it's better to start with the market part of the equation. It is entirely possible that you'll run by a great idea that already has an existing demand and market for it. I do think it's somewhat rare, especially these days. If it's truly a great idea, it's somewhat likely there are already a few teams working on it. If it's a feasible market, a couple of those or maybe just one will come out as the winner.
Market First. Why?
Why to start with the market? Value, and consequently, money is made out of inefficiencies. Let's say you ride by bike to your work place each day. You use the same route given that's what you've gotten used to. What if you discovered a shortcut one day? You could get to work perhaps ten minutes earlier and less sweaty. In addition there could be less traffic lights to bother you.That's what business is all about. It's finding some inefficiency and providing an alternative. I divide these further into necessities and novelties. If you are in a necessity market, you'll likely have a bigger potential market. Novelties can be tougher to sell but in turn have less competition around. You do have to be able to communicate the value you provide particularly well in this case, though.
The given example is bit of a both. It's necessary to go to the work in order to get money to get bread to the table (unless you have a good social security in your country). A faster route is a novelty. You might actually prefer the longer route if you are into fitness or if it's more aesthetic. So the divide isn't always that black and white. It's possible that one's necessity is another person's novelty and vice versa.
In this case the idea of a shorter route is based upon your own observations. Essentially you are the market. I believe this is the reason why many successful startups scratch some itch of their own. They can skip on the market research a bit, even though it's a risky proposition, and then build something for themselves and offer it to similar people they happen to know. If the stars are aligned, the product goes viral and you are in business!
Unfortunately it's not that easy for the rest of us. If we are heading into some other market rather than ourselves, it is our duty to truly know it. This is where Steve Blank's concept of Customer Development kicks in. If you really want to think like your customer, set up an apprenticeship. Follow the potential customer around for a week or two, take notes. I can guarantee you will notice inefficiencies, or in Steve Blank's words, pains.
Your background allows you to interpret these pains and concoct possible solutions for them. It is entirely possible solving some, or perhaps one, of those is your business. You can validate this assumption very fast simply by asking! Would it be worth your money to get this problem solved?