HOME
FAQS
CONTACT US
Compete.com – Not So Competitive?
March 26, 2009
Just recently, I’ve had at least three people come to me about a web site that tracks your web site visits, called compete.com. They called because of what is being reported by Compete.com’s analysis of their web site – either in a panic because it looked like they were getting no traffic, or crowing because it said they had a bajillion visitors.
Compete.com – what do they claim?
Compete.com claims to show “competitive metrics on every site on the web, powered by the largest pool of online consumer behavior data in the industry”. Essentially: “show me my competitor’s traffic!”
They even have a funny tagline: “Track your rivals. Then eat their lunch.”
(That’s funny: I thought it was “eat them FOR lunch.” Why would I want to eat their lunch? Gross!)
Anyway, they get their data from numerous sources. Their biggest source is apparently the two million people who have downloaded their search toolbar. They take the data from each person’s history, and use that to guess how much traffic any web site gets. On the surface, it’s a really cool tool. And it does work…kinda.
Should you be concerned?
One of those who seemed concerned initially was WNYT.com – Albany’s local NBC TV affiliate. They said that another local TV outlet was going around town with printouts of the compete.com web site statistics. Of course, those statistics in hand indicated that the “other guy” was getting much more traffic to their web site recently.* They came in saying: “Quick! Please buy from me today while we’re ahead!!”
It sounded funny.. in both ways.
Funny Ha-Ha
Funny because the ‘other guys’ (in this case the local CBS outlet) get spanked regularly in regular TV news ratings by WNYT. I mean, every Neilsen rating, without fail, in almost every age group and demographic. It’s been going on for so long that the other news outlets have a fixation on who can just be number 2 . One such CBS guy was heard to say: “Hey, sure they beat us in the 35-50 demo… but we kill in the 90+ demo between 11-12 a.m. during early bird lunch specials at the diner!”
(OK, not really. They don’t ‘kill’ there either. But trust me, that’s very close to what they say.)
Smell Test Funny
It also seemed funny because the web site traffic comparisons seemed off.
It doesn’t make sense that the web site of the TV market’s “first place leader” would rank behind the web site of the TV market’s “first place loser” . If you watch news on one TV station regularly, it seems that you’d follow that one on the web, too. Think about that: If you buy at Wal-Mart, do you think of Target.com first on the web?
So I asked WNYT to supply their monthly traffic from LiveStats, their highly regarded on-site statistics program.
Our own little Fact Checking
I took a look at what WNYT provided from their actual web site statistics. Wow. For February 09:
Actual Statistics show: 168,717 unique visitors.
Compete.com statistics indicated: 59,569 unique visitors.
That’s pretty bad. But I gave compete.com the benefit of the doubt. Maybe they just didn’t get one month right. So I kept checking, going backwards.
January: Actual 261,683; Compete claimed 300,000
December: Actual 272,602; Compete claimed 250,000
November: Actual 272,602; Compete claimed 90,000
October: Actual 148,335; Compete claimed 100,000
Holy cow. Those are some big differences. Don’t trust compete.com alone with your daughter.
Keep Looking.
And, being one of the area’s largest web design firms with 500+ web sites, I took a look at some of our own actual Google Analytics statistics from a bunch of the web sites we manage. Google Analytics is installed right there on the site – and it tracks completely accurately, and privately.
Here’s a graph of what Compete.com said:
Wild and Crazy Swings!
According to Compete.com – their best month was under 10,000 visitors. their worst month was 1,800.
Will the real statistics please stand up?
That’s right…their worst month ever was just under 10,000. Who is running compete.com..are they working for an Albany politician, too?
Nothing to see here. Move along.
So you see, both fact checks seem to bear witness: Compete.com is cool for water cooler talk. That’s it.
And if you use it for water cooler talk, you need to find a better hobby. Really.
* Oh – one more thing. What were the results of the Pepsi Challenge?
You may not have noticed the asterisk above in the first few paragraphs. Those other guys from CBS, showing the “great” results vs. WNYT? Did they get too much of Liz Bishop’s hairspray in their eyes? Here’s what compete.com said:
WRGB - we're #2!
It shows that WRGB.com has about 1/2 of the traffic that WNYT.com does. But we know compete.com is way off… maybe CBS’ web site does get a lot of traffic. The question is: how off are they?
I invite CBS to send me their data, and let’s settle this the real way – with real statistics from your actual web site statistics. I’ll post your results too and then you can have a news team rumble.
Otherwise, I’ve got to assume that WNYT remains the true leader.
I received an email from Aaron Prebluda, Director of Market Development at compete.com about my post, asking to clear up any misconceptions about compete.com’s metrics. I suggested that if I said anything wrong, that I could “take back” what I said on my post. Here’s his response:
“I don’t think you necessarily need to “take back” anything that you wrote, but I think that it is important to highlight some of the differences between pixel/cookie-tracking and panel-based metrics. At the end of the day, our data is based off the online behavior of about 2 Million American consumers, which represents about 1.5% of all Americans online. The challenge for some sites (specifically those that have a small audience, B2B focus, or regional appeal), is that Compete (or any panel for that matter) is not going to have robust data. If you think about our methodology as it relates to your local news channel, we have 2 Million data points nationwide. Of those 2 Million, there are only a small percentage from the Albany area, and of that group, only a certain percentage of them actually traffic to your site. As such, our estimate for this site (with a very local audience) is likely a bit skewed. We do make every attempt to account for any bias in the data with our normalization process, but in the end it is not perfect.
Here is a link that summarizes how Compete gets data: http://blog.compete.com/where-do-these-numbers-come-from/
Other things to take into consideration:
- Compete’s data is U.S. centric, meaning that we take out any traffic from non-U.S. sources (Canadian, Indian, Chinese traffic for example is not represented in our data)
- Compete counts actual people only – any automated bots, spiders, crawlers, etc that may be picked up with site-side analytics data again is not included with Compete.”
Comment by Dave Borland — March 30, 2009 @ 9:19 am
This is Chris White who manages WRGB’s Web site.
I encourage you to use our actual Web site: CBS6Albany.com. You were comparing others to our re-direct site.
Comment by Chris White — December 18, 2009 @ 9:28 am