Lies, Damn Lies, and Statistics!

Roy McClean
Jun 04, 2013
Lies, Damn Lies, and Statistics!

I recently came across someone who was comparing his website traffic to a competitor who shared some Google Analytics data with him.

"Why aren't we doing as well this past month as our competitor?" he asked. "His visits went way up, especially mid-month, and his Facebook Likes really shot up too." So I had a look at his competitors' data, website and social media campaigns. Something looked a little fishy.

I took several statistics courses back in the day. The first thing we learned in these courses…. "Good data in, good data out. Crappy data in, crappy data out." To quote another phrase, there are "lies, damn lies, and statistics".

In order to be sure data you are looking at is truthful and accurate carry out due diligence.

Is Your Web Data Full of Fake Visits or Followers?

Did you know that there are many ways to 'artificially' increase your website traffic, your Facebook Page Likes, your Twitter followers, and more?

So what do I mean by 'artificial'? Let's switch gears and think about how I define a 'useful' visit to your website. I define this as a visit from a real, live human being (with an actual pulse) that is visiting your website with some sort of intent. This doesn't include translation bots, people who are paid, tricked or cajoled into visiting your website, companies that "highjack" Google Analytics code to artificially inflate someone else's website traffic numbers, or even people (with actual pulses) from your business.

If you can get at the 'real' data you have a much better chance of understanding if that pay-per-click or social media campaign you are running is really working.

If you are a REALTOR® you would like to base your marketing decisions based on web data generated by real people coming to your website with some sort of intention of buying or selling real estate, either now or in the future.

If you are managing a resort hotel you are looking for visitor data from people who have real interest in vacationing and will potentially end up as a "head in your hotel bed". Intent is the key concept here.

Beware of Imposter Data!

An Imposter Can Be Hard to Spot
Sometimes it takes a little work to spot the imposter! But it's worth the effort.

To repeat… you can significantly increase your website visits tomorrow VERY easily and create "imposter data". Create a couponing campaign, a pay-per-click campaign, or pay people to click on your site or Like your Facebook Page with a service like Fiverr and voila…. more visits, more Likes, more followers!

I've got nothing against a couponing site like Living Social but if I have seen websites dramatically increase their visits by using such a campaign and that's all it did…. increase traffic. This bump in website visits had no real impact on the business. Sure, lots of people got to the site but no one bought anything nor signed up for anything. And when the promotion stopped so did the traffic.

I personally have seen a company that served a limited geographic area fluff up their stats. This business had thousands of Facebook Likes. But several hundreds of these Likes have foreign names and generic, stock photos for profile pictures. These profiles just didn't make sense. Impressive numbers when you look at the Likes total but this raised red flags for me.

The vast majority of people, including online marketing consultants, don't artificially inflate their numbers. But when comparing your numbers against another site or social media channels beware. When looking at Facebook Page check out names and icons of people that Liked the page. You can check out Twitter followers too. You can also use tools like FollowerWonk or Qwitter to sort through your Twitter followers and identify fake profiles or profiles that haven't been active for quite awhile.

Phony profiles exist! Just sayin'.

Filter Out the Crap, Use the Good Stuff

If you are truly interested in your customer traffic you should try filtering out your 'local' traffic. Get the IP addresses from your office and use Google Analytics filters to remove this traffic from your data. (See my blog post about setting up Google Analytics profiles and filters the right way for more details.)

If you are really interested in seeing visits from people that have 'intention' visiting your website create a segment in your analytics tool. Make this segment for those visitors who stayed at least 30 or 60 seconds on your website, or those who visited at least two pages of your site. The assumption here is that people who don't have intent aren't going to stay on your website too long. It's usually a pretty safe assumption. Consider making your marketing decisions based upon this data and not your entire data set.

If you are comparing this to other sites find out if they have done any filtering or segmentation with their data. Ask if they have taken part in any campaigns or promotions lately.

Bottom line? Buyer beware! There are those who will resort to some shady tactics to make their numbers look better.

So when you are comparing your data remember to compare apples to apples! After all if you aren't aware, there truly are lies, damn lies, and statistics.

Have You Come Across an Imposter?

Feel free to leave your comments below if you have ever had to deal with imposter data. How did you find out that the data was fake?

Made With In Whistler