I am posting this as a separate item because I didn't want to take attention away from the twitvid discussions, and I feel this particular topic deserves its own thread.
I know the obvious, generalized answers: patents, copyrights, trademarks, creating a strong brand, creating a following, blog, PR, SEO, making the product complicated, trade secrets, etc..
What I'd like to discuss instead, are unique ideas for barriers to entry. Also, real-life case scenarios and stories from YC founders and HN users, who have successfully spun-out their start-ups with competition high on their tails, about what they've done to leverage and to keep their first-mover advantage.
Google - PageRank. Google was not the first with the product anyway, there were many other search engines in play already at the time - AltaVista, Excite, AskJeeves... Google's infrastructure is not easy to replicate, by the time others caught on to PageRank, Google was already growing exponentially, past the "start-up" phase. By the time others replicated the functionality, Google was a house-hold name.
Twitter - SMS, simple interface, API, brand. Would you rather tweet, or send jaiku's?
YouTube - progressive Flash download, reply features, similar videos actually worked compared to other sites.
Facebook - clean, organized interface, college accounts validation.
One point to mention though, none of these sites had direct knock-offs, icon for icon, like the twitvids. Google, Facebook, Twitter, etc.. had all grown in popularity before they were knocked off. Or at least that seems to be the case, unless there were knock-offs that simply were not nearly the same caliber and just never made it past grandpa's blog as far as spreading use. What I am talking about are early-early stages of a start-up, like the twitvids. And looks like they have similar caliber potential judging by their features currently offered.
P.S. If you want to see a nice Facebook knock-off, take a look at vkontakte.ru. THE most popular site among the Russian internet users. Probably the only thing it's missing is the API, and the fact that it found it's niche with the Russian language.
There is something to be said about knock-offs, and that is why there is reason to be concerned about it. When listening to presentations of enterpreneurs from successful start-ups like Facebook and LinkedIn, they will often talk about "what keeps them up at night."
None of the features of the services you mention are "difficult" to do. PageRank is very easy. The scale is difficult, but lots of companies can scale too. How many YouTubes are there? How many social networks are there?
The issue here is that YouTube continues to improve. They never stop. They just recently "knocked off" the "turn down the lights" feature at Hulu. They could have said, "aw forget it, no one likes turn down the lights" and refused to implement it.
YouTube continues to run hard and fast. That's why they continue to be the winner. If they stop and they don't give their users a constantly improving product that matches the rest and betters the rest, then youtube will cease to be the leader.
Remember, you aren't building a startup, you are building a company. If you're only building a startup, then just write a few lines of code and push it out.
What do you really want? Do you always want to be a startup? Or do you want to be a business? Businesses aren't startups. Businesses are built to last decades. Are you looking for a quick exit?
Is that what you are really asking? How do you prevent another startup from exiting in your place?
Why are you even thinking about another startup? Envy? Greed? Do you want more users? Is it a competition?
What do you really want? It is impossible to prevent your startup from being knocked off. No number of patents or trademarks or copyrights are going to prevent someone from knocking off your startup.
Do you think Nike asks, "How do we prevent adidas from making tennis shoes?"
You should be asking, "How does one make a successful company? How does one last through the startup phase?" Neither of those questions have anything to do with competition.
What I mean by start-up, is its literal definition - a business starting out, with very limited resources, small market share, no brand recognition. Nike and Adidas are already established brands / companies. I am assuming though, that they had similar woes when they first started out - its natural to have competition and to differentiate yourself and your brand from it.
What I would like to see, is a discussion along the lines of "Nike used special stitching in their shoes, and Adidas didn't, because they didn't feel it was important. The stitching caused the shoes to gain popularity." Preferrably for internet-related start-ups, not shoes. :)
I have nothing remotely resembling any real insight into this, but for something like a pure play internet startup where the barriers to entry are pretty small, it seems like app quality, luck and bloody-minded perseverance are probably the most important factors and you only have any real control over two of those.
Web focused analysis follows
Patents - This is obviously great if you can get it, but I'd say the space of possible startup ideas that are coverable by patents that enable you to exploit monopoly power is a lot smaller than the ones that don't.
Copyrights - Well that'll stop someone stealing your name, but what else have you got? Maybe "look and feel" but even that pushes it. It also requires lawyers, which are well, expensive.
Creating a Strong Brand - Requires you to have enough users who like you to use your product and get some buzz. Reduces to building a good product.
Creating a Following _^_
Blog - That'll really help with your page rank in the beginning if you update regularly and is a way to build community but it doesn't really have anything to do with preventing knockoffs. If your stuff is good you'll get to the stage where the vast majority of your users aren't even aware you have a blog.
PR & SEO - Worth doing only if you have something people will stick around for after they get to your site. You must have something there first.
Making the Product Complicated - You mean having solutions to non trivial problems baked into your product? Now that raises the barriers to cloning substantially. And unless you're solving a Hard Problem that has no value proposition (which would be a kick in the teeth if you discovered that was what you'd done afterwards) you've made your product better. As has been mentioned by pg, if you're considering two ways of doing something, one of which is harder, the harder one is almost certainly teh one to go for, because you're only considering it because you know it's better.
Trade Secrets - I was going to diss this one but it's actually among the best on the list, and is I think more or less the same as above.
