I think we all want to believe that because building interesting things is so much more fun than marketing them. Unfortunately, I think we're all wrong. White hat SEO and analytics seem like the most amenable techniques to the coder's way of thinking. Avoiding them because a large component of that world (marketing) is out of line with our view of how things should work will only hurt us.
I'm not picking on you here: I think one of the reasons this issue gives us fits is that we think about it in terms of "gaining and sustaining traffic", which is somewhere in the general vicinity of an actual business goal, but not quite there. "Traffic" is not valuable per se and often people optimize for it in ways which are counterproductive for the business. For example, if you get featured on TechCrunch, you will certainly get a lot of "traffic", but what does that get you, really? A few dozen comments saying the design of your front page would look better in blue, your VPS chugs under load, and "didn't Taxi.derm.me already do this last year"? And then tomorrow they'll have forgotten about you.
But people spend insane amounts of time and effort trying to get on TechCrunch, when they could spend that time writing pitches to blogs whose owners will melt to know anyone who is reading them, or writing systems to make pillar content which will be attract qualified, interested prospects for the next decade.
And even for those few (arguably) lucky folks who make it onto Techcrunch, they might assume that if getting a big splash "gains" traffic then subsequent big splashes are needed to "sustain" it, throwing an increasing amount of resources to further pessimize their business to fit the preferences of a narrow, fickle audience who does not spend money on the stuff they have to sell.
1) they send less traffic than they did back in, say, 2006. It's only 3,000-4,000 uniques.
2) one backlink from them doesn't appear to be worth crap any more in Google. Google (as far as I can tell) looks much more at the heterogeneity of your linkbase than it does any individual link from a high PR site (maybe Wikipedia's excluded here, but I don't know).
Patrick and Peldi's advice to target 40 smaller blogs with personalized pitches (there's that interesting content thing again - interesting to an audience of one each time) is probably better than TechCrunch traffic in every single way you slice it.
Traffic coming a TechCrunch link still converts way higher than most traffic. Why? Because the article is read by ~50,000 people and only those actually interested follow the link.
