Why You Can Almost Guarantee Google is Using Your Analytics Data

    September 2, 2011

While many webmasters and publishers use Google analytics without a second thought, smart publishers, marketers, and SEOs are left to speculate… Is Google using this data? What might they use it for? And am I doing myself more harm than good using it?

Truth be told, unless Google ever does a full disclosure about analytics data (which to date hasn’t happened), we are left to conjecture and guess. However, if we look at Google’s business decisions, especially since Larry Page has assumed control, the only conclusion you can logically come to is that yes, they are using it. But let’s not get ahead of ourselves.

Conference attendees have been using Q&A sessions to try and pin down Google engineers about whether or not they are using bounce rate, exit rate, time on site, or other specific factors. But, by asking these narrow questions, they allow the engineers to sidestep the real question with carefully worded answers. What we need to do is stop trying to figure out exactly what factors they are using so we can try and exploit them–instead, we need to get answers to the larger question: “Is data from Google analytics currently used in search engine ranking pages?” Why is focusing on Google’s use of one factor bad? Much like that kid in class who asked “Is this going to be on the test?” you’ve lost the plot. That kid has stopped focusing on learning and simply wants a good grade by regurgitating specific facts back to the teacher. It’s unlikely that Google is looking at any single factor; instead, they are looking at more than one aspect to determine overall quality. Engineers from Google constantly tell us to stop focusing on these narrow factors (like pagerank) and focus on the big picture and what it says about our website.

Ever since Larry Page has assumed control, one of his big focuses has been shuttering some of Google’s non profitable projects, like Google Labs. It’s not that Google is no longer innovating; it’s that they are taking a more pragmatic approach. Projects need to be cost effective, and they’re not giving them the same length of time as they did in the past to get to that point.

Let’s take a look at analytics. Maintaining uptime for the massive number of sites that Google analytics runs on requires a huge investment in hardware, software, and skilled engineers and technically skilled labor. Next we need to add in the programmers, Q&A, and resources associated with maintaining and updating the user interface/reporting side of gGoogle analytics. Simply put, it’s a huge investment of time, resources, and money. With Google’s recent shift in direction about projects needing to be profitable to stay alive, there simply isn’t any logical conclusion you can reach except that they are using the data. There’s a true saying about this situation: “if you aren’t paying for a product or service, you are the product being sold“.

There are multiple business uses for this data, including forcing up prices on adwords keywords, determining adwords quality scores, understanding consumer usage, and validating organic search engine ranking factors–to name just a few. Yes, Google adwords may have started as a supply/demand bid driven market system, but once adwords quality score got factored in, it became a black box model and prices could be raised artificially as needed. Don’t believe me? Try and explain why quality score forces me to bid $5 to display ads for my own name in adwords, but I can show ads for Matt Cutts for $0.30. Unless, of course, you want to defend the bizarro logic that I am more relevant for someone else’s name than I am for my own.

The real takeaway here is that Google is the data Borg. Without feeding the Borg signals that people are visiting, using, and returning to your site, you have little chance of ranking organically. These are the types of signals that real businesses and brands will send the search engines. These are the signals that become difficult and expensive to fake without large distributed botnets, malware, or hacking. The question you need to ask yourself is at what point does it become easier to spend more time building these signals the right way and less time on faking it via black hat spam techniques …

Check out Graywolf’s SEO Blog for more articles by Michael Gray