Two seemingly unrelated announcements in the past few weeks are actually pretty significant when combined with a few other ideas.  I’m talking about the release of the Google Analytics Data Export API and the recent announcement that Google is jazzing up the information available from the referrer strings.  This post addresses how both can be used to create a profoundly optimised site structure that pretty much tends to itself.

But first, a warning:

This is not a post of tips, tricks or tweaks. It will not benefit those who haven't already been working hard to build a) a compelling web product, and b) external authority in the eyes of the search engines.

For those amongst you that have been working hard, hang on to your hats! 

This post sets out valuable, strategic techniques to evolve your site into a living ecosystem of link strength with a sprinkle of automation to help you get on with growing your business and not forgetting some nifty report outputs to keep your editorial workflow efficient and focused.

Phew, right, that's the niceties out the way - let's get started:

The requisite mindset

If we go back about a year, you'll recall that a popular piece of advanced SEO was second-page keyword targeting.  For those who haven't read about it, it’s all about how to use Google Analytics to identify those phrases for which you receive traffic from non-first page SERPs, the logic being that you clearly have some authority on those terms to be anywhere at all in the SERPs, you just haven't given them the push they need to poke their head up over position 10.

Cool, but not earth-shattering stuff.  Yet.

Now that Google's telling us which terms rank and where they rank through a fuller compliment of referrer string parameters, we can address this method easier by using rank rather than simply capturing the SERP page number.  James Morell has an excellent article explaining this here so I don’t need to.

Prioritised keyword focus

By following those few posts and indeed several others, we can quickly arrive at a bulk list of priority themes to target by taking that list and overlaying a combination of:

• The likely traffic gains using any number of keyword volume tools (Google, Wordze, Wordtracker)
• Rank vs calculated CTR (total available traffic/my visits on that term)%
• Home-brew relative competition metric (SEO for Firefox export or any number of others)
• Associated acquisition yield (conversion value, etc.)

This bit’s not all that important for the aims of this post, and there truly are hundreds of ways to settle on your business-critical themes.  What this step enables you to do is identify which valuable terms can be relatively easily bumped up to a considerable increase in traffic.

Understanding the demandscape

As we know from clickthrough analysis old and new, ranking gains bring exponential traffic growth.  Sadly, this CTR data is long out-of-date and doesn’t take into account the different competitive landscapes for different verticals of demand, or indeed the quality of the resulting SERP.

So to help us along, I’d like to bring forward a complimentary measure to enhance our understanding of rank.  For now, let’s call it POAT or proportion of available traffic.  Catchy, eh?

With our new-found transparency on rank vs. traffic share then, we can understand SERP CTR for every single one of our own terms, removing that old and dusty layer of AOL CTR assumption.

A quick example.  I’ve kept it broad but you can of course delve further into each semantic trunk in individual cases:

I'm getting 200 visits per month on terms containing 'blue widgets'.  It’s a high-performing set of terms for me and I want more.

My traffic projection tool of choice tells me that 200,000 clicks are available from that trunk.  (It doesn’t really matter which tool you’re using here, just that you stay consistent and update the base data regularly.)  vlookup() and sumif() in excel are your dear friends.

This means that I'm getting 0.1% of the available traffic or demand for that trunk.  Digging into it by term, Google’s even telling me which ranks on which terms that proportion of available traffic comes from.  

Even better, I have real data as I climb or drop through the rankings and can even test increasing this POAT with better calls to action in meta descriptions and titles rather than attributing all the glory to rank.

Scientific, huh?

Now, set up correctly, these are all spreadsheets that can self-update with a single data dump or web-based external data source, like Google Analytics API and your favourite keyword tool.  That way, we're always appraised of our shares/positions/opportunity.  Run PPC at the same time and you can practically map the CTR of the whole SERP for any term.

Shoot, time for some work

Unfortunately, you don’t get off completely scot-free.  So now that you have your target themes and an incredibly detailed understanding of the opportunity available, you’d likely output these terms in a nice daily or weekly report and drop them into your SEO ‘ranking’ workflow queue.  It probably looks something like this but is markedly different by industry and, again, not the centrepiece of this post:

• 301 redirect your current highest ranking page for the keyword to a proper landing page
• Sprinkle in some UGC to loosen up that lingo
• Point some internal links to the page using your target language
• Garner some external links if you need them
• Rinse and repeat

How to deal with the internal linking

Landing pages don’t need covering here, but here a few notes on the internal linking method.  You can link to your destination pages using any of the following site-wide measures:

• What's hot/new/popular boxes with a list of links
• Contextual phrase catching link-maker; think server side intellitext but with flat links
• Footer/Breadcrumb; well used but potentially useless

Worst case (read; most labour-intensive) - you are editing this all by hand.  Good luck with that.

It’s probably fair to assume that many of you are automating the population of these outlets by updating a master list of target terms.

Think of it as a shallow tray of link juice.  By changing the terms in these linking outlets, you're slightly tipping the tray toward those themes and landing pages that drive the most value to your business.  That’s a pretty sophisticated eco-system you’ve got going there, well done.

COMING UP: Closing the loop

So just to retrace, as this is pretty heavy going.  At this point you have:

• Prioritised those terms that hold the most value for you
• Set them off in your 'ranking' workflow
• Tilted your tray to give a bit more love to those terms that you get the most benefit from

Next we’re going to:

• Let the content breathe with a term variation catcher to hone your most successful language usage away from the Wordtracker herd
• Self-level the tray of link juice by automating some of the key functions of this process, freeing you up to get on with the ‘ranking workflow’

And that's it for now - pat yourself on the back, you are a pretty advanced SEO.  Much more importantly though, you have a robust and scaleable site infrastructure and you are ready to take on Weapons-Grade SEO PART 2: Automating the Legwork.