(1). New technology that you have developed secretly and roadmapped into the future (so that as your previous release is copied, you are ready to release the next one). Of course, those developments must also be needed by customers (eg. enhance an attribute that they care about; meaning that the initial release was lacking in this aspect, yet it still had some core value that made it worth buying). It's a bit like having your own private giant upon whose shoulders to stand. Because you've done the experiments, you've seen more. You do more experiments, and see more again. This is keeping ahead of the engineering curve, and so it's a supply-side advantage.
(2). Customers that have developed a habit of using your product, especially if they have changed their behaviour in order to adopt it (they might have done so because you were the only one to offer the technology, and then being the one with the best technology - due to point 1). This is related to switching costs. Demand side.
(3). Related to point 2 is when you not just have customers, but you are recognized as the leader, because you have dominant market share (here's where it's an advantage to focus on a niche: it's easier to dominate a small pond that a big one). Psychologically, people like winners. It also means that you win all sorts of races, where people have to make a choice: third-parties choose to support you; big companies choose to standardize on you; users choose to train for your product. Demand side.
(4). Network effects! Again with the customers, but here, because of interaction between customers, the value of the product is proportional to the square of the number of customers. This makes it harder to stop you, but you first need to be ahead - so it's more of a multiplier than a CA in itself. Demand side.
(5). Feedback. Once you have customers, you get feedback from them, and you can improve. If you can get this feedback before your competitors (eg. by launching before them), it can help you keep ahead. It's a kind of secret information, as you learn about what the market really wants; and it's also specialized to your product. In these senses, it is proprietary (owned by you). Another private giant. Supply side.
(6). Revenue. Once you have money coming in, it's a kind of magic, because you can reinvest and grow. Smaller competitors can't match this (assuming you can spend the money in ways that make a difference - like advertising - but it can be surprisingly difficult to find such ways...) And even large companies don't want to bother competing with you, once you really get going (they might buy you though). Supply side.
(7). Although particular skills can help, there's always someone smarter, more experienced etc. However, I like to think that a particular combination of skills can approach a hard-to-replicate advantage (though of course it has to make a difference to the start up. It's no good being an art-expert + TCL programmer unless that's somehow needed).
(8). There are also milder ones like good PR (a kind of advertising), being attractive to talented workers (which in turn increase your success), being inspirational (more advertising).
I agree about patents. They're a good piece of paper as a bargaining chip in acquisitions.
Now, the difficult thing for me is to apply these abstract ideas to my own actual startup (it's easier to be wise when you're not involved). They all seem to point to agility and getting your product out there now. Scary.
This is nice but almost all of these depend on an already existing product with a significant userbase. My presumption is that most people applying for YC or similar seed-stage groups don't have an existing product. So that makes addicted customers, revenue, etc. irrelevant in this context.
I am looking forward to seeing what the author considers "competitive advantages". The only one I can think of is the team behind it. That's all you have when you're starting out anyway.
reality can be inconvenient. Why not accept the challenge of guessing what the author will say (eg. by researching on wikipedia or Porter's competitive advantage)? It's one of the most effective learning strategies - get it right or wrong, you won't forget.
One approach is to not expect to have customer-based competitive advantage at the start, but have a plan of how to get there. Eg. a product that has network effects + a plan for how to get there.
I think the idea of enduring or sustainable competitive advantage is a bit misleading, because things always change, so competitive advantage is dynamic. It can be more helpful to in terms of the duration of a competitive advantage. For example, when you start off with an easy-to-copy product that is in demand, you do have a competitive advantage - it's just that it doesn't last long. But you can use that brief period to build the next competitive advantage. (in rock climbing, a dynamic move is when you are not statically supported, but rely on your momentum to reach a stable point).