Basic SEO is very simple. Choose good titles first and foremost. Make sure your homepage links to all of your most important content. Make sure google can spider all of your content, and create a sitemap.xml to help with this.
Secondly, setup google analytics, and keep an eye on how people are landing on your site from search engines. A high bounce rate is sometimes inevitable, but try to build you pages in a way to entice the user to keep clicking to other pages.
Make signup super simple (I'm sure 90% of people on Hacker News disagree, but OpenID is going to confuse 99% of internet users).
To really get your SEO kicking in, have links to your site from other sites. A good widget strategy can help with this. If someone puts your widget (and you build the links properly) on their blog, and their blog has 2,000 posts, you instantly have 2,000 links to your site. If this blog has high pagerank, great! Some of that pagerank will trickle down to you.
SEO is important to a lot of businesses (especially ecommerce and UGC plays), but not worth early focus for a lot of others.
"but not worth early focus for a lot of others" Total crap.
"Here are a few examples of companies that didn't follow this path (at first)". Note the "at first". Ebay is certainly focused on SEO right now.
The point is that you can get great without SEO-- and companies often do. And the (unspoken) point was that for some companies, search isn't how their products are found.
Low cost distribution is key for startups, but SEO isn't the only way to get that, and often it's a really inefficient way.
If you were starting salesforce.com today, would you hire an SEO specialist among your first 5 hires? Your first 20? How about 37 signals? What percentage of their traffic do you think comes from generalized search (beyond search for their brand name(s)). How about Slide and RockYou? iLike? Do you think SEO should be an early priority for a widget company?
FWIW, I ran a web consultancy for 7 years with a lot of focus on SEO. I'm a big believer in the value of SEO for many types of businesses... Just not all.
Also FWIW, I'm currently running a business that is growing 7-9% per week for the last 16 weeks (that's 3000%+ annual) with no marketing and negligible SEO (virtually all organic search is for a brand name search). We've got some interesting SEO plans, but I maintain that it's a bad thing to focus on for us right now.
I insist, even if you have a viral web app/widget, to not focus on SEO from day 1 is missing the whole point of the net and will cost you bucket loads in the long run at best.
FWIW I would be very interested to hear of the app / widget that you are enjoying the 8% weekly growth. Mine is myplaylist which is also enjoying high viral growth.
Again, the point I was making (with 10 years of professional SEO experience under my belt) was that for some companies, search isn't how their products are found... If that's the case, an early focus on SEO (the time consuming part-- not the "low hanging fruit") is distracting from more effective distribution channels.
The degree to which you focus on SEO should be proportional to how much your customers are searching for what you offer (you should also roll into the equation the KEI/competitiveness of the target keywords).
There are plenty of businesses whose target customers do NOT USE GOOGLE as a tool find the product. 37Signals is a great mom-and-pop example. Building the next great search engine is another example (people wouldn't search for "search engine"). There are countless others. In addition, any business that's diving into mature verticals where Google is claiming the "2nd click" (weather, images, etc) ought to seriously consider eschewing SEO as an early strategy in favor of other distribution efforts.
0) Have an old domain, New domains are penalized thus take 6 months or more to get top for 2 word search terms. 1 word search terms can take 2 years.
1) Correct SEO, your site will be indexed properly and optimized for your targets. Miss this and miss your target.
2) Link building, this is getting harder all the time. Google see link building as the enemy and are doing all they can to discount unnatural links.
3) Remove all barriers to entry, no sign ups etc, immediate gratification.
4) Innovation and content. Killer apps that people find useful and content that is compelling.
For the record, I have about 10 popular domains, have been recently featured on Techcrunch twice and all of the suggested blogs, which are all good, but at the end of the day, search traffic is what counts in the long run.
This is precisely why Facebook and other social platforms perform so badly for advertisers, people using them are simply not in search mode. That and the adverts online are totally crap!