Of course I agree with the great insights by Future Now guys Robert Gorelli and Bryan Eisenberg on many questions of user engagement (though by necessity in my practice I use dumbed-down versions of what they do)… but there’s something left out in the most recent round of responses to Nielsen finally putting the page view to bed as a meaningful audience engagement metric for the purposes of selling advertising space online.
Specifically, we still need (see above), a meaningful audience metric for the purposes of selling advertising online. (Probably several.) Let’s assume for a moment I’m a large content publisher. Heck, assume I’m MSN. Maybe I’ll sell ads in a variety of formats, based on advertiser demand. I’ll need to price the ads in some way, be it on a formal rate card, through negotiation, or through an auction.
Or if you prefer, assume I’m a publisher of deep content and the operator of a pivotal resource online with a related light social networking component — let’s say it’s called ILoveMyBackYard.com. I’m not a powerhouse, but all the outdoor patio vendors and such happen to be clamoring for ad space on the site. It’s a quality site, so I can attract higher rates than “just AdSense.” A good position to be in. How to fairly price the “inventory”?
I’ve got to make money. I don’t want to sell subscriptions, because that limits openness and growth. So, in the mix of a variety of sponsorship and listing options, *based on heavy advertiser demand from those who simply want to include online display advertising as part of a broader online-offline mix*, I’ll sell some display advertising.
First of all, (a) what’s wrong with that? (b) if there’s something problematic with the pageview due to AJAX, then we need to replace it with something. (c) if tabbed browsing and other perverse incentives mean it’s not useful to measure time spent, then we need to replace it with something. (d) If certain events achieved by users are hard to prioritize fairly, then we need to replace it with something. And with all due respect, Robert, that something is not a sales pitch for persuasion scenarios! Remember, we were talking about my revenue needs as a publisher and community resource site, and the needs of media buyers 90% of whom may never want to figure out how so-called persuasion scenarios fit in with their preconceived notion of allocating a large media budget.
So meanwhile, in the offline world, the 30-second spot is not quite dead yet. Full-page newspaper advertisements and magazine ads are not dead yet. Far from it. Their online counterparts are doing pretty well, too.
As the interruption marketing world (not my favorite personally) moves towards an increasingly diverse ad mix, let’s face it, ad folks are still filling spaces with ads and coming up with a fair formula to charge for that space. Insofar as it’s a marketplace, it’s their right to do this if there’s a demand. The main scandal would be if the main metrics are so laughably primitive as to be “gameable” by the least scrupulous players. And often such players are the biggest-name publishers. So I’m in agreement with that part of Gorelli’s argument. Give me something I can reasonably fall back on like visitor counts and audience size. Beyond that it gets harder to measure user engagement, to be sure, so I proposed something like a “multi-point scorecard” as a potential helper metric, but at the end of the day, this is impractical.
I think the problem is, we’re talking about too many things at once. There are quite a lot of rational scenarios for advertisers big and small to shell out for different kinds of listings and units online in places generic and specific. Advertisers will need to get more sophisticated about how well the ads “perform,” but some really deep-pocketed ones are willing to just go for saturation tactics in key verticals, without over-measuring. So be it!
And meanwhile, for the average observer not in the process of transacting an ad buy, it just muddies the waters as the agencies like Netratings and comScore publish these broad-based “rankings” of web properties that we’re supposed to look at. These have always been vanity metrics, easy to manipulate and ostensibly useful for impressing people at cocktail parties. I recently overheard someone from a long-defunct search engine company (still, sadly, in operation in shell form) mention that his company was a Top Ten Web Property in Canada! Did this mean he attracted a large crowd of groupies hanging on his every word? Absolutely not! People aren’t stupid. They know a bunch of legacy page views from old personal homepages, and a gaggle of motley content acquisitions, can easily add up to a Top Ten Web Property (or more like “OK, OK, sixteen now, but we were ten for a month in 2006”), but that’s irrelevant to most of us. Soon, I’m sure word will get out that some of the Netratings numbers factor in AIM usage where they shouldn’t, (not that the word needs to get out, but I’m sure it will if it needs to).
I mean, when you think about it, the new “time spent” metric adoption by Netratings doesn’t literally “hurt” Google, now does it? Nor will AOL get a “boost” in any tangible sense.
If we can agree on anything, the online ad world should be critical of easily gameable metrics, and should be wary of laughably unsophisticated measures relied on by ratings agencies that have this unwarranted godlike status. But my sense is that most sensible players already are.
But in a time of unprecedented supply and demand for online advertising, it might not be the best idea to split too many hairs as individual players try to practically execute on these vital transactions. Gullible advertisers who get completely ripped off have themselves to blame – but this has been happening on a grand scale for a half century, and online is more measurable than ever before. If Party A, Publisher wants to sell inflated page views to willing Party B, Advertiser, — even after page views have been declared dead — I won’t lose a whole lot of sleep over it. I can’t afford to let some online publishers’ cynicism get to me — although I came very close to going insane watching what unfolded in 1998-2000. 🙂
And if a few journalists wish to wildly misinterpret the time-spent metric, most will understand its limitations, so again, I’ll sleep as soundly as ever.
I’d love to think that we have a mechanism in this place for taking Occam’s Razor to the whole process (no, we won’t be able to apply the razor to this rambling post, there’s no antidote for that), and we do: it’s called an auction. As distorted as some marketplaces might be in practice, in theory a good auction platform builder can facilitate an environment whereby advertisers are in total control of what they’re bidding on. When advertisers control and monitor their own measures of success (be these time spent, events, conversions, etc.), they decide what happens to be significant to their business. Publishers that satisfy criteria sought after by more advertisers will thrive, and get paid more for their inventory. Today’s auctions aren’t perfect; Google, Yahoo, ContextWeb, Quigo, and others, are working to perfect the craft.
Not much has changed, then: the higher-level data proffered by NetRatings and comScore are all too easy to misinterpret.
