In order to do well in SEO I'd say you should take advantage of your programming abilities. Look up this company called Narrative Science. Using algorithms to create automated stories. Now I'm not saying make spam, but leverage something like that to give you a competitive advantage. Plus, I know Demand Media gets tons of hate, but check out "site:livestrong.com" and "site:ehow.com". Millions of pages indexed, and thousands more created day after day. You're competing with people like them.. how the heck are you supposed to succeed?! Create quality content, but also find a way to make it scalable.
SEO is something I consider to be the game of major companies...and hackers like us now as well. However, it isn't enough to make a page, slap ads on and SEO the hell out of it.
If you don't have lasting value in your site, you don't have staying power. No staying power means low future profit and therefore lower sustainable startup opportunity cost.
There is always a risk in affiliate marketing, and at this point I'd consider the major success differentiators to be 1) innovation and 2) risk mitigation. You can make a really innovative site and get lots of quality traffic, but you had better do your damnedest to mitigate your overall risk or you may find your source of revenue gone overnight.
You can mitigate risk by spreading your site out so that it covers many products, brokering custom deals, making your own community (FatWallet.com and Extrabux are great examples), or doing something really innovative in terms of functionality (HipMunk, Mint.com).
As an example of what can happen, we had an affiliate site running for the better part of 3 years on Google AdWords. We owned the top spots in a pretty significant vertical and made quite a lot of money. The client (the agency of record for the product) was super happy, we had exclusive terms and business was booming.
All was well until one day I get a call. The merchant decided to drop the agency and eliminate their entire affiliate program. It didn't matter that we generated them huge sales each month, that was it. Nothing we said could convince them otherwise.
That was it.
In order to do well in paid search or SEO for the long-haul, you've got to provide a lot of clear information and really guide the user. You have to make something people would miss if it didn't exist.
The best approach to this day, imho, is not to make a straight-up site any longer. So many people create them that you are one voice in a massive crowd competing for attention. What gets that attention now is unique functionality, ie: software. Focus on developing a unique solution instead of the same-old-same-old.
Case in point: RetailMeNot was the most successful of the coupon sites. They sold to WhaleShark for about $36m and had about $1m/month in commissions. However, anybody and their mom made a coupon site, which made any search term for "insert-brand-here coupons" get bombarded with results, all pushing the same thing. They succeeded because they have great social features, deal in printable coupons now, work closely with merchants, etc.
As for us, most of my time was at first in copy/video/graphics; that ended up changing to more code related time as we started packing unique features into our sites.
Eventually, I figured out that we are a marketing-heavy software company. Haven't looked back since.