If there are millions of prospects and you cant filter by anything the point is to figure out that they don't need or want your product. Do it fast and politely.
Or not even that, the actual goal is to put in the calls without the negative psychological effect of mass rejection.
Arguably, you can start calling before writing the business plan when you only have a half finished idea. If you find just one prospect who says your product would be a wonderful thing have them be more specific. Like that it is much easier to stay motivated. Rejection is much harder if you are deeply invested, wrote the plan, wrote the code, found investors, hired employees.
While cold calling is heavily associated with shit products that doesn't mean your product is shit the moment you pick up the phone. Or maybe it is and you need to be told what is wrong with it repeatedly. You need to be talking with people who've made widgets for decades, they know their stuff.
Who knows, maybe you don't even need an idea. If you just call 1000 people in the funeral sector they can tell you what software they need. Then try weddings, laundromats, plumbers etc
Ask the dumb questions, what would be the right time to call someone in the $sectorName sector? What is their software budget? What are the repetitive administrative tasks? Is the sector patient and polite or do they tell you to fuck off and hang up?
There is nothing super wrong with this, I mean, sales is necessary evil of doing business. necessary, because the whole point is to sell the thing, and evil, because you are coercing someone to do something they otherwise would not have done. But I think there is a healthy ratio here, and I try to make a point(often failing for the reason in the previous paragraph) to avoid products that lean too heavily on the unhealthy side of that ratio. those companies that believe sales is more important than the product.