The Startup Casino
Startups are risky. They are risky like wagering. Like wagering your time and money and energy and resources in the hope of future payouts. In fact becoming an entrepreneur is an awful lot like walking into a casino.
To extend the analogy... various market opportunities are games to be wagered on and for better or worse, the government runs the casino -- setting and enforcing the rules of the game(s) and taking a rake (taxes) off the top of all the wagering that takes place.
Anyone who has ever been to a casino knows that the jig is up the moment you choose a game to play. In fact hitting or staying, raising or folding, dollar or quarter, red or black.... these are all trivial and ultimately meaningless decisions relative to choosing which game(s) to play. Some games have poor odds, others have better odds. Some (in fact most) are absolutely certain to loose you money, and others actually give skilled players a chance at making significant money. Games with good odds are relatively cheap to play, and those with bad odds are very, very expensive. Of course no casino games (or businesses) guarantee success, but there is a kind of spectrum from the truly horrible odds to the only kinda-bad.
House games are certain losers, thus expensive: all slot machines, roulette, keno, craps, etc.
Some games are in-between: e.g. blackjack where it's possible to essentially break-even with lots of skill and even make money if you cheat (it's arguable as to whether counting cards is really cheating or not, but you get the idea).
Pari-mutual games -- notably poker and some sports/track betting -- actually offer opportunities for profit to players with significant skill/strategy and/or informational advantages. These games are still subject to massive amounts of luck and variability (especially over short time spans or small trial sets), and the house always takes a piece, but the skilled players can profit over the long term and at the expense of less-skilled players.
It's the same in the business world. When you are considering starting a company, often the type of game you choose to play essentially determines the outcome.
On the long-odds end of the spectrum are the big broad consumer plays. These are often best-or-bust businesses with terrible odds and enormous network effects. Online you know them as Facebook, Youtube, Pinterest, Twitter, WhatsApp, etc. Offline, entertainment businesses tend to fall in this category (e.g. movies, books, video games). This category is roughly analogous to the house games in the casino. The successful players here are all run by smart and savvy crews, but the reality of the matter is that the roulette ball just happened to settle (at least for a time) on their number.
In the middle-of-the road are a broad swath of services. Businesses offering things people actually need or want. Retailers, luxury goods, apps that save time or make life easier. This is by far the largest part of the potential startup landscape. This segment is most like playing poker. The best player can always loose, but the risk can be managed somewhat with diligence and effort.
On the farthest end of the business spectrum are commodities. Raw materials, energy, food, real estate, etc. If you've got them you can generally make some money selling them to people who need them. These types of businesses don't find simple analogues in the casino world (since no casino game has such favorable odds).
So choose what to play wisely. Most startups can't afford the commodity game. Most of the attention goes to the riskiest business games with their neon-bright press coverage and huge payouts.... Just as casinos promote their winners with huge billboards of winners standing with a huge check in front of the paying slot machine.
The good news is that for all the flash and glamor, you don't have to play business-roulette. You can play poker instead.