Solving a pain point
A product must create value - but creating value is a subset of this criterion. Good products solve real problems - problems that people must solve one way or the other. OK products are “nice to haves".
You only get paid if you solve a problem. Solving someone’s problem is why people would ever buy your product in the first place.
What about Facebook, the iPhone, Instagram? People got on just fine before these things - what pain point did they solve? Well, great products create new problems that can only be solved by the product itself.
When Facebook became popular and everyone got on it, you had a new problem: your friends were all sharing things on Facebook and you weren’t.
Defensibility
Ok, so you’ve ID’d a pain point. What’s stopping your product from becoming a commodity? Is it the technology - i.e., is no one able to replicate your product because of IP or trade secrets? Is it network effects? Is it economies of scale?
Without defensibility, you may still make money, but you’re going to be competing for scraps with the rest of the market.
If solving a pain point is why people would buy this product at all, defensibility is why people would continue to buy your product.
Scalability
Scalability, in my mind, is the confluence of two factors: a large potential market and a viable growth model. Consider this: you are the sole owner of wedding photos for a recent wedding. Your potential customers really want to see the photos (pain point) and you’re the only provider (defensibility). So - you’ve got the makings of a great product. But, there were 100 people at the wedding, 50 more that didn’t know but care enough to see the photos, each willing to pay $50 for it. Too bad.
You fell over because the market was too small.
Well what about a big potential market? You can still fall over if you don’t have a viable growth model to eat up the market. The growth model essentially means that your cost of user acquisition is lower than your customer lifetime value. I’m sure there are other nuances, but that’s pretty much it. Even if you solve a problem for someone, if you’re paying more to deliver the product to them (with “deliver" meaning everything from the averaged costs of developing the product to creating it to marketing it to supporting it and beyond) than you can receive for it, you don’t have a viable growth model.
So if solving a pain point is why people would by this product, and defensibility is why people would buy your product, scalability is why you’ll be able to create a business around your product.
Concluding thoughts
So solving a pain point, defensibility, and scalability are all necessary. But are they together sufficient? I haven’t thought that through, but my hunch is that they are.