So we take the report on Natural Born Clickers and the results of our Lab Store AdSense Optimization and what do we have?
I’m thinking about a basic model for understanding the potential effectiveness of online advertising based on Engagement. Basically:
The more Engaged a person is with the task at hand, the less Attention they have for out-of-context advertising.
The gross amount of Attention available on the web is finite. That means if you pay Attention to one thing, you have to ignore something else. This creates Attention winners and Attention losers. In general, for any space available for advertising, in-context wins and out-of-context loses. That’s Relevance, right? Therefore, out-of-context ads should be much less effective than in-context ads.
So, for example, if the task is Research, and a person is using a Search Engine, the PPC ads focused on the Research topic are highly relevant and Attention gets paid to them. Also, in the same Research mode, if a person is searching or participating in a Chat Board focused on the topic, display ads focused on the Research topic are viewed as highly relevant and Attention gets paid to them.
However, if the task is (for example) interacting with a social media account, then very little Attention is available for advertising – PPC or otherwise – because all other advertising would be out of context with the task, except ads directly related to the task, such as for widgets or tools. This effect would generally explain the concept of Banner Blindess, since most display advertising is completely out-of-context. People just learn to ignore it.
Not breakthrough thinking in Consumer Behavior or Psychology but for Online Advertising it might be, considering the number of business models nowadays that plan for “advertising” to be the revenue stream. In fact, it’s quite possible that the more Engaging they make these social apps, the less effective the Advertising will be.
It’s about the limited amount of Attention any one person can have.
When ads are in context, you get an effect much more like that of Fashion or Hot-Rod Magazines, where the ads are part of the content, they are part of the Engagement and so get Attention. Out of context, much less Attention, if any. Not part of the content, screened out.
For the same ad, PPC or Display. In other words, it’s not the delivery method that matters, it’s the context and available Attention. PPC ads by their very nature just happen to have the context problem solved.
For example, a TV ad running in the middle of a favorite TV show is much more effective on an individual than the same TV ad that plays in the background while someone is Engaged with a project on the computer. Same ad, different context.
Now, here’s the thing. This idea makes a lot of sense. Can we expect anyone with scale to test it, prove it empirically? I dunno, because an awful lot of business models will get completely hammered if it is true.
The test would be pretty simple:
1. Define Engagement – really not too hard for this, it’s how many “actions” take place per unit of time. Seems to me this would capture the whole Attention thing; if you are busy taking actions, that’s where your Attention is.
2. Run both in-context and out-of-context ads during the measurement period. Display or PPC.
3. a. Measure clicks and conversion, if that is your game
b. Measure Awareness and Intent, if that is your game
4. Compare results
Does anybody think that out-of-context advertising would win, or at least match in effectiveness?
If there is a difference, what does it mean for biz models relying on out-of-context impressions? What can they do to correct this problem?
The next post in this series is here.
Great! This is why I like your stuff Jim – you like to juxtapose things, i.e. think in terms of dynamics.
Can’t wait for you to start answering your last two questions.
Thanks, Chris. ‘Cmon, I bet you have some answers of your own…
I have to admit that I have only a shallow answer – OOC users just have to judge success by a different ruler, and they have to take more on faith. We actually have this issue going on right now and that’s how we’re trying to deal with it. But I really want to hear more from you.
I am a big believer in banner blindness and thus tend to not be helpful to teams that are trying to work with OOC types of online marketing. A web page is such a short quick experience, after all. Print ads are different, especially magazines. TV might even be different. Online rich media ads happening at points during a streaming experience like a TV show are different too. But banners — well, they’re both out of context and part of a chaotic context.
HI Jim,
Hmm, I’m wondering whether all that advertising you described, be it OOC or not, is just not interruption marketing, in the S. Godin’s sense. As soon as you are on a site, the “what’s-in-it-for-me” / “how-the-heck-is-this-relevant-to-what-I-want” mechanisms start to fire. If you came through a PPC on a search engine, you happened to be better qualified, at least in principle, so you respond better.
I would suggest that it is particularly, if not only, while on a search engine, that people are the most responsive to ads. Once they’re on a site, they’ll start ignoring ads on that site, while trying to figure out if they can pursue the goal that first brought them there.
If I am correct, well then, advertising based sites should get back to promoting impressions as a metric; at least we can’t “prove” people were not paying attention at the time.
Jacques, that’s an interesting thing to bring into the discussion. I’d leave them in two categories — OOC and interruption marketing. I think OOC just gets ignored or not noticed (generally) while interruption marketing has an emotional negative component, plus an additional cognitive load. Because of those two differences, I’d suspect that the outcomes would be different. Different outcomes would be the real reason to keep them separate, conceptually.
But I’m just conjecturing, although I’m basing it on my psych background (such as it was). I’d love to see behavioral studies or, better yet, some biological tracking such as GSR or brain imaging!
Hi Chris,
I meant “interruption marketing” in the general definition Seth Godin gives it in Permission Marketing, which is basically advertising that wants to get your attention while you’re doing something else (watching, reading, etc.). So, I don’t know if there’s is “out-of-context” and “in-context” advertising really. My point is that it’s all interruption marketing, whereas people coming in search mode might be more receptive.
However, I like Jim’s categorisation of purpose, which definitely would have an impact on ad responsiveness. Anyway, I don’t know; just thinking out loud.
Oh, and just another quick comment: “Engaged”, which we are trying to measure (cf. Eric Peterson’s effort) would then NOT be good news to many sites ;-).
Good points both.
Jacques, I’m not sure we’ll ever see “permission” in the display ad world (online or offline) though you could argue the nature of PPC creates an implicit “invitation” to advertise if not permission.
I’m with Chris on the notion these two ideas are disctinct yet overlapping. Intrusive is different from Relevant, the first being an “invasion of personal space” concept that has more emotional baggage with it, as opposed to simply “tolerated but useless to me”.
I’d guess both ideas contribute to lack of effectivness in display advertising; Intrusive gets an outwardly hostile reaction (Negative Awareness?) whereas Relevance says the ad is simply ignored based on experience (Blindness). This is the difficult environment display operates in.
On top of that, if you are actively Engaged, then Intrusive just enrages you even more, even interrupting your work flow, while Relevance says if the ad is in context, it might break through the Blindness.
PPC does not generally face any of these challenges as long as it is a relatively intelligent effort from a marketing perspective.
What is interesting to me, and kind of shoves this Engagement layer out to the light, is BOTH our AdWords (PPC) and AdSense (Display) campaigns on MySpace suck horribly. That is unexpected and suggests to me there is a new factor in play with social networking.
By the way, the creative was always the same across all usage. In other words, the exact same ad that generates great results in Google Search and pretty good results in Google Syndication generates horrible results in MySpace, both in the AdWords and AdSense formats.
And then this same ad, running on topical chat boards and other high-context locations through AdSense, generates really great results.
That leads me to ask, what’s different about MySpace?
Hmm, you make a persuasive demonstration here, especially the last two paragraphs. Yes, what would explain the difference with MySpace? Most probably what you have been explaining. I think you have here enough good stuff for a great white paper, or maybe an upcoming DD newsletter?
Jim, can you define “suck horribly” a little more? CTR or visit quality? I’m wondering if MySpace is particularly bad at matching on relevance. I just did a search for “skin care” and one of the resulting ads was for a Dog Food Comparison Tool. In what context (ahem!) are your ads appearing, do you know? And if you’re doing broad matching, what actual search terms are being matched to the ones you are paying for? Are people seeing your ads after they do a search? what kind of search (for people? all of myspace? videos? etc).
I’m curious because we have gotten really good results from Myspace (in terms of visit quality) but the site in question was geared to the 12-25 group.
Any metric you want to look at – clicks, # pages viewed, time on site, and so on down the chain.
I addressed the potential for poor match problem earlier in comments a couple of times. Certainly possible but I can’t control that. We rarely use broad match except for on teenie tiny search volume concepts.
What kind of site was yours, was it related to social networking or similar (ringtones, etc.)? What was the success metric for visitors?