500 word or 1000 word articles?

You say in one breath that our SERP data isn't relevant at all and then say "proof is in the real world organic SERPs"? I am surprised I have to sit here and explain to you how you've twisted yourself around, but well, you're supporting my argument with yours at this point.

I know I have, what, 20 posts here, but I've been doing SEO for nearly 5 years now. My argument is laying on the table right now with a few million data points from real SERPs from real SEOs who are ranking for real keywords that actually matter (or else why would they be researching them) and you're really countering with "common sense" and opinion is all you need? I dare you to ask any SEO at all in the world whose side they'd choose in this situation.

SEO is a game of data, opinions don't matter. I can literally prove you wrong with a graph right now. I'm not arguing that you can't rank a site with a small amount of content, of course you can. But not all keywords are created equal. Try ranking for anything actually worth ranking for with no content and it simply won't happen.

This has nothing to do with keyword stuffing or density or any of that. Remember that little Panda thing? It was entirely about content. And you mean to tell me I have 1.5 million data points that are just...not relevant compared to your common sense? Come on man, stop spreading disinformation, it's what ruins the SEO community.

Ok. I would love to see an example of your theory. Just 1 Serp example of a site that has 3 separate pages ranking for a main keyword (not a long tail), that also has got some great long tail rankings as well, all with about 300-400 words (1000/3)

Your theory sounds awesome. Although your theory relies on having lots of content - over 3 optimized pages instead of one, but a lot of content, none the less.

Please show an example, though. It doesn't have to be one of your sites, (as I am sure you have dozens of sites that would show examples but don't want to "out" any of them). And I know you can't "out" your clients, right? But nonetheless, I'm sure there is at least one example somewhere you can share to back up YOUR theory?

The elephant in the room that your post ignores is that for any terms worth significant money (I'm talking stuff like buy weight loss pills or buy satellite tv, not the higher volume short tails that don't feature implied buying intent, just so you don't counter with that straw man position, even those short tails pay great too if you build converting sites), you're going to need a significant link building investment for each page you hope to rank top ten.

You're also going to need an authority site with significant age and overall back link profile for google to treat multiple inner pages as the most relevant results deserving of being frontpaged (not even the 1,2,3 in your example) in these types of niches.

For easy stuff like "overhyped worthless pdf file of the week review", I'd agree with your position. For hard keywords worth thousands of dollars per day, day in and day out, I'll take one longer post covering all aspects of the target keyword (+ plural) and a slew of LSI longtails as well, then consolidate my multi variate linkbuilding efforts into that url. Just what works best from my experiences, YMMV.

I don't think there's a wrong or a right answer, every site is different, every SERP is different, but ignoring the volume of data we accumulated at serpIQ because it doesn't fit with your current world view is rather short sighted, IMO.

Also, I know the site windjc sold because I did some SEO consulting on it with him and I'm rather jealous of how quickly he turned around the ROI on that project. Makes me feel like an amateur.