Over lunch recently, an online publisher shared with me an RFP that he had received. The advertiser wanted to know how many impressions they could get for a million dollars.
publisher looked at me and said, "See, they still aren't getting it, are they?" When I asked what he meant by that, he said, "Broadband isn't about building reach. It's about building
relationships. Broadcast is all about how many saw
the advertising. Broadband is all about how much time they spend
with the advertising.
should be asking is not how many impressions they can get for $1 million, but how much involvement they can get for that amount."
"And you can tell them how involved viewers were in their
advertising?" I asked.
"Absolutely," he said.
He went on to explain how he had recently run a commercial for an advertiser on his site for six days. The commercial was two
minutes and 20 seconds long. Rather than pre- or mid-roll, it was user-initiated, requiring the viewer to opt in.
Over the course of six days, he had 1,542 unique streams to this
commercial. If an advertiser were buying reach or impressions, this sort of number would receive no more than a grunt at best.
But when you factor in that the average time spent with
the two-minute and 20-second spot was one-minute and 50 seconds, well, it certainly impacts the "worth" of those 1,542 streams, doesn't it?
After all, it means the advertiser received
169,620 seconds of time spent with his brand's message. Or, 47 hours worth.
The cost to the advertiser? Three thousand dollars.
The publisher then asked me what I
found to be a most interesting question. How much would it normally cost an advertiser to get 47 hours worth of time spent with his brand?
My answer was that there is no way of
knowing, as advertisers aren't asking for a measurement of involvement in their commercials. All they want are the number of impressions, regardless of whether the viewer is involved in the
message or not.
Besides, in the intrusive methods deployed in broadcast, and now being adopted on broadband, there is no way of telling whether viewers are actually involved or not.
Yes, they are in the room, or in front of the screen, but involved? Who's to say?
By allowing viewers to opt in to the advertising, this publisher was able to monitor the
viewer's intent rather than the advertiser's intent. When viewers opted out, on average, after watching 78% of the message, he had an accurate measurement of time spent with the brand.
Now what's interesting is that this publisher can still go and sell his site on a CPM or CPC basis. And, to media agencies, this will continue to be his strategy. After all, media
agencies are paid to deliver eyeballs to, not involvement in, the message.
But he's also starting to go directly to advertisers to talk about building relationships, rather than
reach, on his site. He's also letting advertisers know that he'll help them optimize the different touchpoints
across his site in an attempt to increase the amount of time that people
spend with individual brands.
His logic, although quite simplistic, seemed difficult to refute. "There are only 24 hours in a day," he said. So the more time a person spends
with one advertiser's brand, the less time that person has to spend with the competitor's brand.
In this way, he said, "offering a good 'return on involvement' starts to offer the
advertiser a pretty darn good return on investment."
As for answering the question with which we started: How much involvement will $1 million buy?
If $3,000 was able to achieve
47 hours of time spent, then $1million should deliver around 15,666 hours of time spent with the brand.
Or, 652 days.
Which is why we feel that this just might be a question
that more advertisers will soon be asking their media agencies.
And if they don't know the answer -- well, I happen to know a publisher who does