You could never isolate it completely with only a single site, but that doesn't mean you conclude that you can't measure anything, or estimate at all. If I made all the changes in a few days, gave it a few weeks for google to notice the changes, and organic search traffic didn't increase in any noticeable way ... that's strong evidence that it didn't have much of an effect.

Now if someone is going to tell someone else that they should do xyz, what are they basing that on? Their gut? Or did they do these things in the past, and measure an improvement, even when they didn't isolate all the variables as you are demanding of me! Why not demand the same from the person making the suggestion?

> Now if someone is going to tell someone else that they should do xyz, what are they basing that on?

Fair question. I would say these things are current, accepted "best practices". I'd liken it to suggesting you use a version control system instead of emailing code back and forth. Your existing system might work, but many other people who were also emailing code back and forth are seeing gains using this new-fangled VCS thing, so it's almost definitely worth doing.

You might try searching for "SEO before and after" or "SEO case study" to see how on-site changes have worked (or haven't worked!) for other sites. At any rate, many of the current "best practices" are trivial, under the hood changes so it doesn't really cost anything to put them into place. Most of them simply help search engines understand how your content is structured and benefit both you and your users.

yes, I agree with you sort of. I"ve done a lot of SEO in the past, nothing you mentioned was new to me. But here is my experience, one of the startups I did had a brand new domain, which had nothing on it before launch, and didn't hyper optimize the type of SEO you mentioned. However, due to the nature of our site we got a lot of links for big name groups like peta, the nra, etc. Follow links on their site, all within a few months of launching.

We ranked really well really fast, whatever sandbox people think google may or may not have didn't seem to affect us. Like I said, within a few months, on a brand new domain, without an SEO expert optimizing things like H1 tags, we ranked really really well for competitive keywords.

That's the basis of my hypothesis, without good backlinks, it doesn't matter a whole lot, and good backlinks matter a lot more than tags on a page, whether you have /posts/title, etc.

But I do really appreciate the time you took to look at my site and offer suggestions. You did catch a missing index :) I didn't notice because no human goes to tag pages, and it didn't show up in 'new relic' because my other pages are fast and get hit a lot more.

btw, i'm curious if you see my site on this search, http://www.google.com/search?sourceid=chrome&ie=UTF-8...

I rank decently for long tail things like that, if I did your suggestions would that page go up? I see it in 7th position.

Yeah, SEO is really just about having good content and having links from other good sites to it. All the tweaks are just to help the search engines determine exactly what your content comprises and where it lives in relation to your other content (see Microformats/rel attribute on hyperlinks). Without good content, none of that matters.

Yup, I do see your site on the first SERP for that query. Here's a shot of the SEO breakdown for that SERP: http://i.imgur.com/URUr2.png Not bad positioning considering there's zero backlinks to that page on your site. If you had 1-10, you'd probably rank higher.

You probably wouldn't see any rise by just doing tweaks, but they're probably still worth doing overall. I would focus on encouraging your users to spread your content. Give them a "share this link with friends/embed this link on your web page" widget. Digg is dead/useless, get rid of that icon. Get a Tumblr and Stumbleupon share icon on there instead. I'd probably space them a bit more away from that big, ugly, unrelated image ad, too. :)