Quite unambiguously, I rate the Google Adwords Search Query report a number out of 10. That’s right, a stunning #/10.
The Google Adwords search query report is designed to help you see the entire phrase your visitors searched for, when that search matches a ‘broad match’ or ‘phrase match’ – in other words, when what the user searched is not exactly what you were bidding on.
Awesome I exclaim with as much glee as PPC can make me muster – which is exactly not much… but really, this should be the talk of the town – this is the first made-for-the-public way to view all of these long-tail keywords. This represents the first simple end-user glimpse at the real truth of how people are reaching your site via those mystery-meat Broad and Phrase match terms.
I don’t fear telling you, I’m in love with the long-tail. It’s a torrid affair that dates back years, to the halcyon days of my youth when search engines were mere tools, not omnipresent overseers of the Internet. And in my heart of hearts, I truly believe that this romance would become infectious to the general web-site-running populace at large, if made just a little easier to understand, just a little easier to see.
This simplicity is what the promise of a no-frills Google report would seem to provide. Stuck doing the occasional PPC work on semi-virgin accounts (from an optimization standpoint), I would love any help in the gathering of insightful offerings for friends and clients.
But Google misses the mark by more than a little here. Remarkably they have chosen to exclude phrases with a low impression rate (what I affectionately refer to as… the long tail) in favour of saving money on processing power and server space, and perhaps bandwidth (I’m not going to source that, as it’s made up, (edit: oh hey maybe it’s not!) but I can’t think of any other reason – the data is obviously there).
Okay so perhaps my experience here is atypical. It is a rather niche site where, as you can see, the impressions over the course of a year are often in the hundreds if not less. For a larger advertiser there may be some more useful information to gather on the mid-tail – but I still expect the long of the long tail to cumulate into a good collective portion of conversions, and so a tool which would appear to be designed for peering AT THE LONG TAIL would be a little more useful if didn’t filter out low impression keywords. Just my take.
Okay rant over, here are some ways you can pull the same data while Google upgrades to 486’s:
> Look at your server logs. Better yet, tag your links in adwords (with Google Analytics tags, for example) and write a script that parses your server logs and couples the tagging with the referrer phrase – this offers a little more information at least. If you’re unfamiliar with server logs, you might first want to check with your hosting provider to see if they give you access to your ‘raw server logs’ – then have a read through this U. of Michigan primer on the topic.
> Hack Google Analytics: Our friendly neighbourhood ROI Revolution handed along this fancy-dancy little chunk of code a few months back. If you use Google Analytics, this is the easiest, cheapest and most useful solution (not to mention reveling in the irony of solving the problem with google’s own tools via some custom code is supremely satisfying).
Anybody got a third?