Competitive Insight, Girl Scouts and Hells Angels

If one of those “panel based” web statistic providers (think Alexa, comScore, Nielsen) tries to sell you their products, or worse, someone insists on making a business decision based on statistics from one of those “We really don’t have access to real statistics, but we have access to web servers so TRUST US” companies (think, run for the hills…


If you hadn’t noticed, Google was down to upwards of 14% of it’s users at one point this week.

amtrak_crashYes – down – as in “very slow” for some, even “not working at all” for others.  Only for an hour mind you, yet down nonetheless.  They even blogged about it.  Given what we know about Google’s market share, that’s a LOT of people who lost their ability to search the only really good search engine on Earth.

And that prompted me to think once again (yes, I have a very bizarre way of connecting dots) about web analytics – the tracking and analyzing of site visitor information that is so important to my small and mid-size business clients…



Personally, I learned about the down-time thanks to an article by Jordan McCollum, who confirmed that the End Times weren’t actually upon us (yet).  See, I wasn’t directly affected.  Either because I was surfing elsewhere at the time, or I actually was at Google (probably looking to see where my mesothelioma law firm clients were in the SERPs for one of their key word phrases – yes I really do have a client in that arena) and in that moment, wasn’t one of the fourteen percent.  Heaven forbid.  I would have been extremely upset.

Well, not really – I get paid whether Google is down for an hour or not.  I am, after all, the owner of my own company.  And I ALWAYS pay myself for my time, regardless of circumstances outside my control.  🙂



Depending on who you believe, Google owns anywhere from  65%,  to more than 70% of all search on the web.  Andrew Lipsman over at comScore tried to explain the variations in these statistics earlier this year.  Of course, his spin attempted to put comScore in a good light as compared to the others.  Because it’s in his financial best interest to convince you that comScore is reliable, trustworthy and necessary.  Except the truth is they use old school statistic methods that come out of the Television marketing world.  Archaic.  Barely not really plausible even in the 20th century.



Anyone who knows me, knows that I’m a statistics agnostic.  I used to be a crime statistician, and if it suits me, I can show you in five minutes how to turn soup into nuts, water into wine, and five visits into five thousand.  On digital paper at least.  So when it comes to statistics, I inherently know that you can’t take them at their face value.

Now, I don’t go around just mocking stats and web analytics companies.  I always put them in context, and fact check.  And use other information to help form business decisions.

However I firmly believe that when it comes to Competitive Insight, statistics are inherently a bear trap waiting for the unsuspecting camper to meander right into.  Especially when it comes to Internet analytics.



For example – if a web site is in a truly niche market – let’s say they cater to Girl Scouts, hard core bikers, Gen Y, or hundreds or thousands of other niches, then it is most likely that solutions from comScore, Nielsen, HitWise, Alexa, Compete, and all the rest, are going to be completely misleading at best, and patently false at worst.

Why?  Because none of the players have any real grasp of those  user bases, or, in the case of “comparisons to your competition”, access to the real traffic statistics on those web sites.

Sure, Alexa may have slipped their toolbar onto the computers of a small number of unsuspecting Girl Scouts.

But a “few” is far from any kind of large enough share to be a valid basis for extrapolation.  And its’ more likely that the only reason those Girl Scouts have computers that got that Alexa ToolBar on them is because their older brother was probably surfing for illegal music downloads at some point. Or porn.  And the statistics from that household are then tainted, to say the least.

The point though, is that if the market focus of a web site is Girl Scouts, Panel based stat companies can’t possibly have enough Girl Scouts on their panel to give you real data to work with.


Calling All Volunteers

Exactly how many hard core, shiv carrying, barroom brawling bikers do you think voluntarily allow their computer to be loaded with a 3rd party user activity tracking solution from comScore or Nielsen?  Okay, so YOU may not cater to that crowd, but again, it’s illustrative of the diversity of markets served through the web.  And a perfect example for this article.

Hells Angels Disclaimer: If you know of any Hells Angels who are participants on the ComScore or Nielsen Net Ratings panels, please let me know and I will peacefully and most respectfully revise this article.  Or I’ll just call one of my friends from the Diablos and let them work it out with the Angels.  😉


Commercial Break

If you want an affordable  domain and hosting solution, I recommend

[adrotate group=”10″ banner=”14″]




In my example above, with say, Girl Scouts, there are probably hundreds, if not thousands of web sites – perfectly legitimate, professionally designed web sites, that cater to people in the Girl Scout demographic.  How well do sites like HitWise or Alexa provide insights into market reach for sites like those?  Not well at all.  Terribly in fact.  Beyond pitiful. And it’s even worse for specialty sites with that market focus.

Just to test and either validate or debunk my own theory, I ran a check on three web sites that fit this arena.

This site, run by Girl Scouts of Eastern Massachusetts, sells uniforms, books, insignia, camping supplies and more.  They’re a legitimate web site selling authentic Girl Scout products.  A small site, for sure, yet if I want to market to Girl Scouts, and I want competitive insight into other web sites that cater to them and come up in the top five at google, they go on my list.

This site is not focused specifically on Girl Scouts as their market.  In fact, according to their web site, they are “An online information center for youth leadership education and development.”  Their market is Teachers, Parents, Kids and Teens.  A site that would make a perfectly fine destination for any up and coming Girl Scout looking to have a positive impact on the world.  Better still, with such a large scope, the site must have lots of visitor traffic data.

According to their web site, Learn and Serve America supports and encourages service-learning throughout the United States, and enables over one million students to make meaningful contributions to their community while building their academic and civic skills.  Again, this is quite an appropriate web site suited for Girl Scouts.  And since it’s a U.S. Government site, surely there’s plenty of traffic.

Now, I picked these three not because any one of them might be just like your web site from a products or services perspective, but to illustrate a point.



While all three are appropriate for the Girl Scout market, checking on them statistically at Alexa is worthless.  Alexa shows NO statistics on any one of these sites.  As though they don’t even exist.  Or aren’t “worthy” of tracking. Because if they’re not in what Alexa deems the top 100,000 web sites (at least the top 100,000 that Alexa has statistics on), then you don’t get any info at all.


Well heck.  In order to be Alexa’s top 100,000 sites you need to be pretty special.  How Special?  Let’s take a look.

One of our clients, Carlos Santana (maybe you’ve heard of him?) has a web site called – a site that Alexa says is ranked number 219,860.  And that 0.00049% of all web users have visited the site in the past month.  And another client’s site, (have you heard of Sammy Hagar?), doesn’t register on the charts at Alexa either.  (Sammy only got a paltry 0.00019% of the web’s users last month according to Alexa). SO even if you’re name is Carlos Santana or Sammy Hagar, you can’t even get comparative data of any significance from Alexa.  What a waste of time.


Like I said though – the basis for their statistics is massively flawed so even when someone comes to me and says “We’re ranked #1 at Alexa in our market”,  I say “SO WHAT.”  What I care about is – do you come up higher than your competitors in Google organics?  What is your conversion rate for your PPC ads?  What’s your Return on Investment in your total Internet initiative? And anyone who wants to advertise on your site or invest in your company should be asking the same questions.  If they’re not, they’re participating in what I like to call the Statistics bubble.  (Think Mortgage Bubble, Internet 1999 bubble)



Compete claims that their data is the only real and valid data because they have arrangements with the major service providers.  They’re supposedly given direct access to Internet Service Provider traffic information.

So okay – let’s find out.

I ran all three sites through the system.  Amazingly, what I found were apparently  numbers. supposedly had 549 unique visitors in April, had over 15,000, and was reported to have had 12,973.  Well the problem with these numbers is that I’ve done a number of tests over the past couple years with sites like this.  While I don’t have access to the actual server logs or analytics reports for any of these examples, I do have such access for over five hundred web sites – clients I have served.


But wait.

How trustworthy are the numbers?



Well for, Compete reports that in April, they had 38,387 total visitors and 18,413 of those were unique.  Well, according to Google Analytics, they had 68,946 total visits, and of those, 37,194 were unique.  And from my previous article, while we know Google Analytics is flawed, if anything, they UNDER-report visits.


So the next time someone tells you to check out, tell them I said to take a bite out of that wax apple and let me know how it tastes.

And Suggest they install Google Analytics.  For Free.

For all it’s flaws, you get much more reliable information on your own site than any panel based extrapolation.  And if you think you can learn about your competition from any of those companies or, fuggedaboutit.  (HEY – I lived in Brooklyn before moving to California)

And if one of those “panel based” statistic providers tries to sell you their products, or worse, someone insists on making a business decision based on statistics from one of those “We really don’t have access to real statistics, but we have access to the big ISP servers so TRUST US” companies (like, please don’t believe them.  PLEASE.


I beg you.

And for the record, Hells Angel Photo copyright 2008 Alan Bleiweiss. (Do you really want to know the story behind that?)


What do you think? Am I off my rocker? Or do you have similar experiences?

About Alan Bleiweiss

Just another guy. Who happens to have a lot of experience living, breathing and sleeping organic SEO. So that's my primary focus - high end SEO audits and consulting for sites ranging from thousands to tens of millions of pages. In my spare time I blog, rant, write eBooks, and speak at industry conferences.

Read more from