Who reads my e-commerce blog? Really surprising answers from Alexa.

So I was looking at Alexa for another site of mine and decided to look at this website while I was there and this is what they told me:

Based on internet averages, thewhineseller.com is visited more frequently by males who are over 65 years old, have children, have no college education and browse this site from school.

My reaction was…really?

I mean, OK, I can buy that I have more male readers than female, business is still mostly male dominated so that fits with my expectations. Over 65 year old men who have children? That one surprised me but, still, within the realm of believability. If you are running your own successful business, you are probably a bit past middle age so I am can buy that as well.

But over 65 year old men with children that didn’t graduate college and browse this site from school? What school are these over 65 year old men in? This one makes no sense at all to me. Or does this mean that most of my readers just work in a school?

Now their page doesn’t say how they measure these stats but you are welcome to review them yourself here.

I have to believe they aren’t accurate but, what do I know?

So let me ask you this, just for fun. If you had to guess who reads this website, what would be your guess?

And, hey, if you are an over 65 year old man with children, no college education and are browsing this from a school, say hi! 🙂

Discover more from The Whine Seller

Subscribe to get the latest posts sent to your email.

5 Comments

EventHorizon1984 on March 18, 2010 at 1:32 pm

> I have to believe they aren’t accurate

The demographics do have the caveat “Confidence: low”, making your assessment 'Confidence: high'.

http://blog.alexa.com/2009/04/alexa-and-love-yo…
“We're careful to calculate estimates of confidence whenever we produce a demographics score for a site. It turns out that the accuracy of a demographics score depends not only on the site's Alexa ranking — which in turn can be thought of as a proxy for the total amount of traffic the site gets — but also on how the site's traffic is distributed over the demographic categories. For example, depending on user behavior, we might be very confident about a site's score for the 18-24 year-olds, but less confident about the 65+ year-olds, and this effect is somewhat independent of the site's total traffic.

So it's a bit tricky. But rest assured, before we serve any demographics data to our users, we perform our error analysis and judge the accuracy of each and every demographics score. If we find the accuracy of a particular score is unacceptably low, we simply don't serve up that data point. And for every data point we do serve, we boil the nitty-gritty of the error analysis into three simple categories: confidence low, confidence medium, and confidence high.”

And no, no comment, no, no, no, but waving anyway.

“I don't get facts wrong! It's everything else I screw up.”
Flynn Carsen, The Librarian: Quest for the Spear (2004)