They say it was easier prior to Panda. Not sure about it. 7 years ago when I started my first website that was to sell Rock N Roll t-shirts and related merchandise, I have realized eventually that one should treat Googlebot as a human, a very smart human. There was absolutely no way to trick it back then, just like there is no way now, with Panda. It is no surprise that all this emphasis on keyword density and other mechanically determined, and, under circumstances, mechanically generated aspects has been depreciated among the SEO community. Keyword density, for example, should be pursued no more and no less than in any college essay – there are things like thesis, supporting terms and supporting paragraphs, transition words, supporting material and citations… just like that. If you can write a good college essay, such that your pedantic teacher would give you an “A” for, then you can design an optimal web-page, or the entire web-site for that matter. It was true before, it is true now - with Panda's “ravaging” legacy or without it.
It is not that simple (if we have an audacity to call it simple, of course) for commercial websites, however. Here the competition is too intense, and your ability of creating a unique content extremely limited.
I was an amateur SEO developer back than, which I believe I am today too. However I was lucky to be able to figure out that in the highly competitive segment of my keywords: t-shirts, global monthly search 16,600,000 – competitiveness: High – according to Google Keyword Tool (while this is the current data, it always has been just as congested out there) Keyword Tool there was virtually no way for a start-up to gain traction, and that there perhaps was another way to promote the website into the same segment but without using the keyword “t-shirts”. I realized that there should have been no small percentage of people who were searching just for “shirts”, not “t-shirts”: “Beatles shirts”, for instance. On the other hand, there were literally no t-shirt websites that would optimize their pages by “shirts”. The common belief has been, I think, that the “shirts” search numbers are miniscule compared to that of “t-shirts”, and therefore it was worth it. It was worth it, however, and this is why: even if it was just 5% of those 16 million searches, would not it be more than enough for a 2 person start-up? Within 3 month we were able to account for such a share of this supposed 5% that we had over 3.000 relevant visitors to the web-store every day. I understand that all these contradicts to the widely adopted opinion that the highest competitive keywords should be targeted by the SEOs, and there certainly is logic behind it. But promoting on the sidelines might yield astonishing results, as long as one can find a niche, yet unclaimed one. As I mentioned before, I personally consider Googlebot as a human, a smart adversary, and an ally too under the right circumstances. Therefore, I am pretty sure that it knew at once as to what segment of searches we were targeting with our website. But what could be its argument against showing us in the searches that included “shirts”? Nothing. We grabbed the niche and Googlebot acknowledged it, and started showing us above the well-established sites even though it knew precisely well that we did not carry nearly as good back-link rank, or any other rank that it considers it the SERPs. How do I know that it knew we belonged to the same industry segment? Well, easy: because it showed us in the same company. It had to – there was only us with “Beatles shirts”, and there was a bunch of others with “Beatles t-shirts”. Also, later on it made us a full-fledged member of the segment, showing the website at the same level but within the “t-shirs” SERPs now. Still, it has to do with a niche, whether you can find one or not. Therefore, my thesis here I guess is: Go for the niche first, and that just a proper essay may not be enough.
I have done this a couple more times since then, with relative but yet success. Google's Keyword Tool provides an invaluable support in looking for the niches.
I can be contacted at the student Web Designer blog here - temporarily.