I regularly report radio ratings, so I thought I’d attempt the more dangerous task of ranking local media websites.
Somewhat like radio’s Arbitrons, these ratings measure the share of the local audience looking at a particular site, within the universe of 50 or so local news sites I provided.
The data comes from Experian Hitwise, and is based on visits from people within the Minneapolis-St. Paul “designated market area,” or DMA. In other words, local people visiting local sites.
A few caveats before we get to the rankings, which I plan to update monthly.
Experian Hitwise buys data from Internet Service Providers like Comcast, so the data is directly measured, though it’s not complete. (You might use an ISP Experian Hitwise doesn’t purchase data from, or get your digital information some other way.)
Also, unlike radio ratings, these only measure visits, not time spent on-site. They also don’t measure page views, how well the sites perform for advertisers, or if a site’s content is viewed on another domain (like YouTube).
Finally, the company gave me a Top 20 from my dataset. I might have missed a few; the non-listed are at the bottom of this post. Feel free to chime in with any I missed (I accidentally omitted K102). This is the shakedown cruise, and I’m sure I’ll have to make some tweaks. Disputations welcome in the comments.
OK, the numbers!
Let’s just say the Strib dominates the local ratings in a way radio stations can only dream about. WCCO is the top TV station, KQ wins the web-radio race, and you have to go all the way to #14 to find a shop without legacy-media ties (us).
The four TV stations combined can’t match the Strib’s visits, but two of them beat the Pioneer Press. The biggest radio surprise — for me anyway — is that women-talker MyTalk 107 ranked 12th but KFAN didn’t make the list at all. And the suburban newspaper chain ThisWeekLive made it but rivals Sun Newspapers didn’t.
The Top 20, by the way, accounts for 98 percent of all visits to the Top 50. I have some numbers about local traffic to national sites, but I’ll pass that along later.
Here were the other sites I asked to be reviewed:
Newspapers: Minnesota Daily, Downtown Journal, Southwest Journal, Onion Twin Cities, Minneapolis-St. Paul Business Journal, Minnesota Sun newspapers.
TV: The CW.
Radio: WLTE, JACK, WCCO, KAN, KDWB, KTLK, The Patriot. (Again, I’ll add K102 to the list and possibly update.)
Magazines, Mpls.St.Paul, Minnesota Monthly, Metro.
Websites: BringMeTheNews, Secrets of the City, Momslikeme, The Uptake, Twin Cities Daily Planet, Minnesota Independent.
What are the visitor numbers?
While I believe the Star Tribune is at the top of the heap, I have a hard time believing they’re getting three times the traffic of their next competitor. How accurate this is really depends on how hitwise is analyzing their data. The Strib still has their shady 10-minute page refresh, which almost certainly drives up their traffic figures. If hitwise’s sampling for visits is short, the Strib will wind up which vastly inflated numbers, as we see here.
I could be all wrong, but without a little clarification on their methodology, I’d take the disparity between the Strib and everyone else with a big grain of salt.
It’s an interesting idea, although I can see a couple of problems.
I’m not in love with the accuracy rate for Hitwise, for a lot of technical reasons. Quantcast is more accurate because it directly measures traffic. But only to sites that sign up, so while MinnPost has a direct number, sites such as K95.com don’t. And others (like WCCO) hide their data from the public.
The other problem is more philosophical. How do you define “local web site”? If you define it by geography, then you’re comparing sites that draw traffic primarily from the local DMA (such as Vita.mn) with sites that draw traffic from all over the world (according to Quantcast, about half of StarTribune.com’s traffic comes from outside MN).
Then of course there’s the problem of web sites that are based locally, but primarily oriented nationally. I would put my own modest effort (AllYourTV.com) into that category.
Still, I hope you keep at this. I think with some tweaking you could come up with a great resource.
There are probably some technical reasons why KFAN isn’t showing up; it could appear to the browser that it’s a standalone site when it’s a subdomain of a larger Clear Channel site. It’s actually pretty common in a shared hosting environment; you would see it as kfan.com but on the server side it would be kfan.clearchannel.com or something like that.
This is why server-based tracking tools have their limitations and the more accurate tools are user-based, like Quantcast. Google Analytics isn’t as accurate as you’d think, especially since Google doesn’t maintain historical data on specific pages.
If time spent per visit was included, MinnPost would move up because of its strong content.
TPT’s low rank is kind of startling. If they had some decent local news programming maybe they’d get more visits. And no, Almanac is not decent local news programming, it’s stale broadcast time for local talking heads.
I would love to make sense of this with respect to all local blogs, etc. I know lileks had an insanely large following that might rival some of those on this list, but I think he was the only one. The inclusions Brauer has asked for are also very important. But we still have no solid and comparable count for actual readership of some of the blogs that are vying for influence
Kudos to you David for taking a stab at rankings. While taking qualitative data might be the most scientifically objective, it definitely short-changes certain websites. Case in point: time of visits. Anyone can surf onto KSTP.com off of a top Google News result, but keeping them around is an important judge of a site’s effectiveness and the quality of their content.
Furthermore, using DMA is inaccurate — both in terms of the numbers that are reported and using it as a rubric for success. ALL of the media outlets listed have a reach that extends beyond the MSP DMA. I’d like to see a list doesn’t segment the numbers to DMA just to measure the state-wide reach of the area’s biggest news outlets. Keeping it segmented to one DMA severely handicaps the race for Twin Cities readers and doesn’t give an accurate representation of a website’s overall influence in our state.
Perhaps instead of just incorporating one metric, you can report back a few of them so we see where organizations are succeeding in the Twin Cities, in greater Minnesota, in pages accessed and in the average visit time.
Lileks has a Quantcast rating at 98,810, Erik. That’s the rough estimate, and in my experience Quantcast actually overrates the rough estimates versus the real numbers. If Quantcast is to be trusted, he’s not close to anyone on this list.
What is the denominator? So we can figure out a rough estimate of visits to each site.