Web Analytics: The Myth of Inaccuracy
I am so, so, so tired of ‘experts’ telling me that things like podcasts, social media or other, harder-to-quantify online media are “hard to measure”.
Bull. The most difficult-to-measure online asset is far, far easier to measure than the best offline one.
More Measurable Than Any Traditional Media
Corporations have blown ridiculous amounts of cash on television, print advertising and radio. Traditional agencies then tell them proudly of impressions and readers.
But what do they really know? Nothing.
I may have walked out of the room when the commercial played. Or flipped past the print ad. Or simply ignored the radio spot. And you have no way of knowing.
Let’s compare that to a podcast; podcasts and other downloaded media are the most difficult to measure.
With a podcast, you at least know someone downloaded it. And you know if they continue to download episodes. So you can measure (not perfectly, of course) how many subscribers you add, and how long they stick around.
Not great, but far better than wondering if I headed for the fridge when your Superbowl spot came on.
Other Web Analytics
I also hear that web analytics aren’t accurate enough to justify spending money online.
Web analytics aren’t perfect, it’s true. But you can still track far more, and more accurately, than in any other medium.
Where else can I compare the conversion rates of one ad to another? Or the cost to acquire a customer using a single keyword?
The Real Truth
I think that the whole inaccuracy myth was created by folks who can’t figure out how to use an iPod.
Web analytics are far superior to anything else. Should you stop advertising in other media? Of course not. But internet marketing is less expensive, more measurable, and therefore more effective than anything else you’re going to find.
Start looking for another excuse, OK?
Joseph Jaffe talked about this in Podcast 105, which was a CEO roundtable. I was driving down to Portland and nearly drove off the road ranting and raving after hearing one of his guests state that podcasts could be a great marketing medium if the ‘problem of measurement’ is resolved. BTW, his podcast rocks. You must subscribe. Right now!!!!
Portent's Founder & CEO
Ian Lurie is founder and CEO of Portent Inc., an internet marketing agency that has provided internet marketing, including PPC, SEO, social and analytics services, since 1995. more >
Subscribe to the RSS Feed
Amen! I worked previously in TV, where fewer than 400 “meters” were supposed to tell us how many people were watching our show(s), representing (allegedly scientifically) millions of local viewers. Right now, I can watch one of my analytics services tell me who EXACTLY has just clicked onto my site and onto what part of it; later, I can see something awfully close to a precise count of how many homes and businesses came by at some point today, and exactly where they went. The only flaw is undercounting — the fact that large employers may have just one IP representing who knows how many users; but if we don’t have an exact user count, at least we have pageviews. :) Internet measurement rocks so much harder than TV measurement, and don’t even get me started on newspapers — a stack sitting in the lobby of my kid’s school, ignored till it’s recycled, counts as “circulation”? Yipes!
“I think that the whole inaccuracy myth was created by folks who can’t figure out how to use an iPod.”
Man, that was just classic!
Ian:
Do you think part of the problem could be web analytics people themselves? Sometimes they start dismissing certain analytics tools in front of the “can’t work an iPod” crowd. This makes someone who doesn’t grasp the concept anyway even more leery than they already are. You know they’ll pop off with “Well who knows WHAT Google Analytics’ bounce rates mean? (insert smug chuckling)
I agree web is the most measurable,testable,cost-effective medium we have; I think the “click counters” need to stop sabotaging themselves and save the nit-picking for when they are out of earshot of the ‘great unwashed’.
My only skill: Biting one-liners.
Yes, but I think the providers themselves are even more to blame.
When a company like Omniture pushes out an expensive product that they can’t support, and that’s hopelessly buggy, it hurts everyone. When they crow about accuracy with claims they can’t support, either, that’s even worse.
Any expectation, though, that an analytics tool will be 99% accurate is ridiculous. It’s reasonable to expect 85% accuracy, which we get, and it’s reasonable to tell everyone that that’s a lot better than any other medium.
I’m rambling a bit here but hopefully you get the idea.