Taking On ARF, Engagement, Interruptive Advertising… And Whatever Else You’ve Got!
In 2005, The Advertising Research Foundation (ARF), through their MI4 Initiative, decided to embark on the Quixotic quest of defining engagement. The impetus was finding a more appropriate and applicable metric that could stretch across the rapidly expanding number of channels that were exploding through digital delivery. So, the good folks at ARF assembled a […]
In 2005, The Advertising Research Foundation (ARF), through their MI4 Initiative, decided to embark on the Quixotic quest of defining engagement. The impetus was finding a more appropriate and applicable metric that could stretch across the rapidly expanding number of channels that were exploding through digital delivery. So, the good folks at ARF assembled a bunch of agency people, various publishers, and yes, even a few search marketers, to try to thrash out a definition for a standard metric that could apply equally to video, print, digital display, audio, and text. I watched the proceedings from the sideline with skepticism.
While I agreed that current metrics, based on mere exposure, were outdated, I thought it was highly unlikely that there was a “one size fits all” metric lurking out there somewhere. It just wasn’t that easy. And nowhere did this become clearer than when you looked at the nature of interaction with a search engine compared to a display ad. It was the difference between pull and push. The nature of engagement was completely different, and trying to measure both by the same metric would be hopeless. It would be like trying to measure how fast you can run and how much you can leg press with a stop watch. Maybe you’re using the same muscles, but you’re using them in completely different ways, so you have to measure them in different ways.
Engagement defined…kind of…
Finally, after several weeks of debate, the ARF emerged with this definition: “Engagement is turning on a prospect to a brand idea enhanced by the surrounding context.” It has the sense of compromise, or more precisely, surrender to the TV folks, and everybody walked away from the table with a sense of relief and hasn’t looked back since. At the time of the announcement, I wrote an article asking how this metric could possibly be applied to search interactions:
The problem, from a search perspective, is that there are two very different forms of engagement seen with consumers, and brand plays a very different role in each.
In most marketing, brand engagement is essential. You have to form a relationship between a brand and the latent or expressed needs and desires that lie with the consumer. Engagement is essential, because you have to form an emotional bond that can rise to the surface and express itself as top-of-mind awareness when consumers are ready to actively consider their options. In this instance, engagement is emotional, intuitive and often subconscious. It is this level of engagement that I think ARF is trying to define by somehow quantifying this emotional bond, referred to in market speak as being “turned on.”
But there is another type of engagement: engagement with the actual act of purchasing. Here, the consumer is engaged with a product, but not necessarily a particular brand. This is the typical point when a consumer will interact with a search engine. And with ARF’s working definition of engagement, I don’t think search will do particularly well in a multichannel comparison.
At the time, I knew there was something inherently wrong with the definition, but didn’t have the working knowledge to know what it was. But recently, I’ve been doing some reading on the perception and decision making mechanisms that kick in in consumer interactions as part of the research I’m doing for a book I’m working on, and I think I can shed a little more light on the inherent problems in ARF’s MI4 quest (which I now believe has died its deserved death). Even though MI4 is 6 feet under (talking about universal metrics), I think this could be illustrative of why search is such an effective channel in the one metric that counts: generating business. But first, a caveat: a little knowledge is a dangerous thing, and as they say on the ad, “I’m not a cognitive psychologist, but I play one on my blog.”
First, let’s go through the steps an interruptive, or typical ad, one must go through to match ARF’s definition of engagement.
The brand “prime”
First of all, the effectiveness of an ad depends greatly on whether there’s an existing brand “prime” sitting in your implicit memory. Is there an existing mental framework, or schema, that the brand can fit in? Is the framework positive or negative? This framework is built up of a mish mash of factors that include, in a somewhat descending order of importance, personal experience, word of mouth, exposure to non-commercial brand messaging (news stories, etc), and exposure to advertising. When it comes to “imprinting” this brand framework, the more personal and emotional the experience, the stronger the imprint. That’s why a particularly intense personal encounter with a brand, either positive or negative, will have a much stronger impact on us than an ad we see on TV. Also, the more authentic the message, the more it will influence us.
Before we ever see an ad, its success depends on the framework that we have locked in our implicit memory, much of it formed by our personal experience, or the experience of those close to us, with a particular brand. By the way, once this framework is established, either positively or negatively, it’s incredibly difficult to change. In fact, research (Burgoon, Burgoon and Miller – 1981) has shown that once a stimulus triggers a response, similar stimuli will continue to cause the same response. If we have a negative framework in place because of a bad experience with a brand and we see an ad that triggers us to remember that bad experience, repetition of the ad won’t change our minds. It will just cause us to replay the horror again and again in our minds. It’s advertising’s own variation of post traumatic stress disorder. For example, every time I see an ad for United Airlines, I relive one particularly horrific flight from Chicago to Toronto.
Choosing what we see
Now, let’s get to the second step: Do we see the ad? I introduced the idea of selective perception in the last Just Behave column. But I just exposed the tip of the iceberg. If we look at the moving parts that make up selective perception, we see why ARF’s quest for a universal metric was doomed to fail, along with 99.95% of the intrusive advertising you’re exposed to each day.
First of all, selective perception is a defense mechanism. We use it to help us keep focused on the task at hand, so we’re not distracted by the millions of other stimuli that constantly bombard us. And a lot of this barrage is advertising. Around 4000 messages a day for most of us, one every 14.4 seconds. And almost none of it is related to what we’re doing at the time. No wonder we tune it out. If we didn’t, our heads would explode and we’d drive headlong into the approaching semi. Advertising starts with the deck fully stacked against it.
But it goes further than that. When we’re exposed to similar patterns of distractions on a regular basis, we make a habit of ignoring them. We create a “stimulus” short hand, helping us to filter out non-relevant material more quickly as blocks of distractions, rather than making filtering decisions one at a time. For example, if we know that ads always appear in one part of the page, we’ll tend to make a habit out of ignoring them. It helps keep us on task.
The intersection of our wants and advertising’s message
And that’s the other part of selective perception that’s important to understand. What we choose to perceive or ignore depends on that task at hand. In order for advertising to be perceived, it needs to hit that magic intersection of relevance. The message of the ad has to be relevant to an interest and need of yours. Otherwise, it’s automatically filtered out because it’s not on task. This is the current bane of our culture. Almost none of the 4000 ads you’ll be exposed to today will be relevant to you. Maybe, if you’re extremely lucky, one or two of them will be. That means the other 3998 are a sheer waste of time and bandwidth. That’s a .05% chance of success. And yet advertisers are willing to roll the dice with those ads over and over again. Tomorrow, you’ll be exposed to another 4000 ads, with the same odds for success. And the advertiser’s answer for improving the odds? Yell louder. Be more distracting. Your response? Filter more. It’s what Marty Neumeier, author of the Zag and The Brand Gap, calls the Advertising Death Spiral.
This idea of personal relevance is a key concept. In the last column, I talked about how contextually relevant ads performed significantly better than non relevant ads. This refutes the findings of an earlier Tacoda sponsored eye tracking study that showed that engagement with non-relevant ads was greater. A recent study we did found this to be true, up to a point. There were signs of greater physical engagement with non relevant ads, but when we tested to see if the ads made it through the 4 stages of selective perception (stimulation, registration, organization, and interpretation), the non-relevant ads quickly fell behind the relevant ads. When you look at the process we go through when engaging with ads, this makes sense. If we’re aiming to hit that magic intersection of personal relevance, the odds are significantly higher that an ad will be relevant to a personal interest if you’ve already made the conscious decision to read content about a topic relevant to that interest. If I’m reading an article that compares the features of laptops under $1000, there’s a pretty good chance I’d be interesting in buying a laptop for $1000 or less.
Matching ads against the primed framework
But, let’s say that this intrusive ad has somehow beaten those impossible odds and made it through your filters. You’ve seen it, it’s somewhat relevant to a need or want, and you haven’t filtered it out. This is where the brand framework comes in. Does the message of the ad match your belief about the brand? It may be relevant, but does it ring true to you? If not, it falls into the no man’s land that psychologists label cognitive dissonance. It doesn’t match your primed brand framework. If this happens, at best you just ignore the message. At worst, it triggers a powerful response, further entrenching your negative feelings about the brand and increasing the odds that you’ll tell someone just how much you hate it. This in turn helps create a negative brand framework in their minds (remember, word of mouth ranks just under personal experience as far as brand imprinting influencers), creating another advertising death spiral. Every time United tells me to “fly the friendly skies,” I see red and spit blood.
Is this whole engagement thing beginning to look like one big mine field? Only if an ad can successfully navigate past all these traps can it even begin to create what the ARF was trying to describe as engagement. And I won’t even talk about the ad’s success in getting its message across. I did talk a little bit about the effectiveness of ad formats in the last column. For the purpose of this column, there’s only one point I want to make. All too often, in order to increase the odds of making it through all these filters, the “volume” of the ad creative is turned so loud, it starts to lose its original message. The ad creators try to engage the senses and, in some cases, they’re successful. They catch your attention, but they’ve exerted so much effort and advertising “bandwidth” (the message carrying capacity of the ad) in battering down your defenses, there’s nothing left to actually sell the product.
So, what about search? How does it do on the engagement chart?
Well, the biggest difference is this: selective perception is there to keep you from being distracted by the task at hand. When you’re searching, that is the task at hand. The huge advantage of search is that you have no filters up. In fact, you’re filtering everything else out so you can concentrate on that task. And the area of concentration is squarely on the top left corner of the page, our much discussed “Golden Triangle.” That’s why banners and other display advertising failed so miserably in the early days of search. They were habitually “filtered out” because they were identified as distracting advertising.
But (and this is huge) if sponsored ads are relevant to the query, they are very much “on task,” so rather than being filtered, they’re being focused on. In fact, in past eye tracking studies, we’ve consistently seen that the north, or top sponsored ads, are the first place we look and the first listings we read. The success of the search ad comes from the fact that it doesn’t have to get your attention (at least, not if it’s in the top north position), it just has to convince you that it’s the most relevant option for your task at hand.
Now, relevance in this context can not be a throwaway, namby pamby term. Relevance can’t be wishy washy. When I say relevance, it’s in hushed, reverential terms, as the heavens part and a single ray of golden light shines down upon it. Okay, maybe that’s a little over the top, but hopefully it gets my point across. The only factor in a search ad’s success comes from relevance. I’ve preached on this before. The more squarely the ad sits on top of that magic intersection of the prospect’s needs and wants, the more successful it will be. Period. End of story.
And because of all these inherent benefits, the search ad can devote all its advertising “bandwidth” to this one goal, establishing relevance. That’s why you only need a few words in a search ad, as opposed to a million dollar production budget for TV. You don’t have to batter down filters and trick prospects into listening. You already have their undivided attention. Now all you have to do match their intent.
Finally, let’s talk about branding and search
But what about that implicit brand “prime” I talked about? What role does that play in search interaction? Well, it still has impact in search as well. When you scan down a list of search results after launching a generic query, say “cheap laptops,” you have some existing brand primes in place and this implicit framework will be a big influencer in what listings you click on. Let’s say I use the above and the following brands come up in the top results. (This is a simulation. The brands may or may not appear for this query. I’m just using them for illustrative purposes.)
For myself, as I’m scanning down the list, here’s what my brand primes (these are my opinions only) are for each of these:
Dell – fairly reliable and dependable, but no huge brand loyalty. I have a Dell laptop and I’m reasonably happy, but I’m somewhat resentful of their huge market share and lack of customer service.
Sony – no real brand framework. No positive or negative framework, which makes me suspicious. Why haven’t I heard more about Sony?
Lenovo – I still haven’t decided about the reliability since the switch from IBM. I had an IBM Thinkpad and had mixed results, but many I know swear by their Thinkpads.
MacBook – love Mac but unfortunately we’re a Windows shop, and I am cynical about actually buying a “cheap” Mac because they’re expensive.
These implicit frameworks are the lenses through which I see and filter the results. Dell has the most fully developed framework, because of my past experience, but it’s rather ambiguous. Next comes Lenovo, but my framework hasn’t been updated since the switch from IBM. I’m relying on some word of mouth (mixed) and a few product reviews I’ve read. Again, ambiguous. Next comes MacBook. I started on Mac and love them. But for various reasons, I went away from Mac years ago. Secretly, I have been resentful about this ever since. But, when it comes to my intent, to find a cheap laptop, the search ad telling me Mac has cheap laptops doesn’t jive with my “Mac” brand prime. It triggers cognitive dissonance. Macs are good, but they’re not cheap. And Sony is a blank slate. I have no positive or negative impressions. While you think this might be an advantage, it’s not. We get suspicious going into unfamiliar territory.
So, with this, it’s highly likely that I’ll click on the Dell listing. This isn’t solely a function of messaging (we don’t really read the full listing–I may pick up on a price point in it) or position (we tend to scan the top 3 or 4 listings and then make our choice), but largely a decision based on my existing brand primes and which of the choices most closely matched my implicit beliefs and experiences with each brand.
Consideration sets and the mystery of the missing brand
That shows what happens when brands appear in the search results. But what happens when they don’t appear? HP didn’t show. For reasons I’ll lay out, this is a golden missed opportunity for brands.
When we go to search for something, like “cheap laptops,” it’s because our task is to build a consideration set. We want to compare alternatives, generally from 3 to 5 or 6, which seems to be the number of short listed candidates we can comfortably handle in our memory banks. Eventually, we whittle this down to 2 or 3, and then we make our decision. As we move down this road, we do something psychologists call “satisficing.” In the beginning of the buying process we have way more options than we can possibly consider. Much as we’d like to rationally consider all the options and make the optimal decision, we just don’t have enough time. It’s called bounded rationality (by political scientist Herbert Simon). We have to create short cuts. So, we use a “good enough” heuristic threshold to cut our consideration set down to a manageable number of alternatives. We create a list that “satisfies” that qualification threshold and provides us with a “sufficient” number of alternatives – hence the word satisficing (also courtesy of Dr. Simon).
Satisficing is what we’re looking to do when we turn to search. We go online to create our consideration set, and search is the first place we turn. We’re just looking to fill up 4, 5, or 6 “brand” slots for future comparison. A brand doesn’t have to be a clear winner to make the list, it just has to pass the initial satisficing cut. We may have an existing brand affinity, or even a few, but chances are we need more to round out the set. And increasingly, when we use the Internet for research, it means we’re open to considering other brands. No brand has a lock on the upcoming purchase yet.
So, let’s go back to my “cheap laptop” example. I use a Dell, so chances are Dell will be a brand I consider. Beyond that, I don’t really have any slots occupied and I’m completely open to be swayed. As I said, I use a Dell but they haven’t made me a brand advocate. They aren’t remarkable. I’m looking for alternatives when I search for “cheap laptops.” Because search is the first place I turn, it also has the first opportunity to start filling those consideration slots. From the example above, I automatically include Dell because it matched an existing brand prime. I may include Lenovo, and I might even check out Sony and Mac, just to see if my initial suspicions proved to be correct. Remember, there’s little risk on the search page. If my click through doesn’t persuade me that I was wrong in my initial assessment, then I just hit the back button and head back to the results for another candidate for the available slots.
Now, if HP appeared in the results, there’s a good likelihood that I would have included it in my consideration set. It would have met my satisficing requirements. But it didn’t show, so unless I had an unusually strong brand loyalty for HP, an exceptionally powerful prime, I wouldn’t consider adding it to the set. Out of sight, out of mind. Now, depending on how long I look around before I close off the set, HP may still have a chance to make the cut through exposure somewhere else, say a review or head to head comparison. But they’ve missed the first magic intersection, the search results page.
How powerful is this branding effect in making the consideration on the SERP? We decided to test that, at Google’s request. We asked a number of panelists who were in the process of considering a vehicle purchase to do some online research about hybrids and fuel efficient cars. Then we tested the appearance of brand in various positions on the SERP, against a control group where no brands appeared on the search results page. For everyone who said you can’t brand build in search, the results showed otherwise. For our test brand, we saw a 16% increase in brand awareness when branded listings showed in both organic and sponsored locations, compared to our unbranded results control group. But even more significantly, we saw a similar drop in brand recall for brands that didn’t appear in the SERP. So, to state this in a bit clearer way, out of 10 people (the sample size was over 2700 people), all shopping for a fuel efficient car, 5 of them had an existing brand prime for our test brand before they ever used a search engine, and 4 of them had a existing brand prime for one of the competitors. Now, let’s let them go to a search engine to start to assemble this consideration set. If our test brand showed up in the top organic and top sponsored location, but the competition didn’t show up at all, 6 of them would now be considering our test brand, so they picked up one potential customer. But, now only 2 people mentioned the competition, so they slipped out of the consideration set of 2 customers. In total, we saw a 42% gap between the test brand and the competition when the test brand was highly visible on the results page.
Back to picking on “engagement”
But the biggest problem with ARF’s definition of engagement was the question of what you were engaged in. The assumption of the definition proposed was that you would be distracted and “turned on” by the ad and its brand messaging. While this wasn’t said in so many words, it’s implied through the language used. You need to be “turned on” to the brand message in order for interruptive advertising to work. It needs to make it through your selective perception filters to be recognized, organized (matched against existing schema), and interpreted. These ads have a lot of work to do, because not only do they have to get the message across, they have to get your attention when your attention is focused elsewhere. That’s why the brand message has to be hammered home with as much sensual “volume” as possible. Think of the ad shorts they now play before movies in theaters. You have a 70 by 30 foot screen dominating your field of vision, full THX sound blaring at you, and you’re in a darkened room. By the ARF’s definition, this would be a home run. But one critical element is missing. You’re not there to watch the ad.
Now, consider search. Search doesn’t have to turn you on to a brand message. In fact, again, by ARF’s definition, the average search ad would be a dismal failure. There’s no sensory engagement. It’s scanning a few words in about 1.5 seconds, on the average. But here’s the thing. Search doesn’t need sensory “volume.” You’re there to find out more about a product or service. Brands simply have to “show up” to be considered. All the brand has to do is crawl over the relatively low satisficing threshold. Why then is it that so many brands don’t even make this minimal effort?
Opinions expressed in this article are those of the guest author and not necessarily Search Engine Land. Staff authors are listed here.