How Advertising Works, and Why You Should Care (Part 2)
Posted by Craig Choisser
How Do You Know If Your Ads Work?
* * *
Perhaps one of the reasons marketing professionals don't dig into the details of how ads work is simply because it's complicated and difficult to pin down solid truths. But there are some basic facts in play that we as marketers can work with. Despite the protestations of the "memory and experience only" school we have to recognize that emotion appears to be a significant player in the reception, storage and retrieval of ad messages, as well as purchase decisions. Furthermore "...neuroscience tells us [that] memory, affect and cognition are simultaneous and interactive" but "[m]ost of the conscious choices that make us pause are dominated by affect." (Ambler 2000)
We also know that when (or whether) cognition comes into play will depend on circumstances - the level of motivation and involvement, and possibly when the decision is in a B2B environment where other people will be contributing or validating decisions. Ambler (2000) came up with a model that incorporates all three elements and considers their interplay. The MAC model suggests that memory (M) dominates affect (A) which in turn dominates cognition (C). He further notes that some ad impacts and decision are a combination of elements, for example MA would be a combination of memory and affect (and I would suggest this is likely the dominant combination for advertising impact and decision processes). More recent neuroscience research would probably place more weight on affect, but the basic premise of MAC suggests that with memory dominant, routine decisions happen without wasting mental energy. This is broadly supported in marketing, psychology and behavioral economics research. And indeed simple display ads like logos in a sports stadium are likely nothing more than reminders, intended to ping long term memory and refresh it. And affect, as we know, improves retention and recall in memory, with cognition coming into play with certain high involvement decisions, but less frequently than any of us would like to believe.
What's A Marketer To Do?
The first step is to acknowledge the primacy of strategy. There's a "bias towards action" in business today that creates the sense that taking time to think is automatically akin to paralysis. There are variables that the foregoing analysis identifies that can and should be assessed relative to your product market and customer base.
First, assess level of customer involvement in purchase decisions. Are customers likely weighing purchase decisions of your products based on cognitive processes? Are you selling enterprise level software systems, or shoes?
Second, assess substitutability of your product or service. This is tough for most businesses, because we all like to think we are unique and irreplaceable. If your business vanished tomorrow, would your customers easily move on to another supplier? If Home Depot vanished, could I head to Lowes without missing a beat? (Answer: Yes.)
Third, make an honest assessment of your value proposition. This may significantly overlap substitutability, although a value prop can include elements not directly incorporated in products or service, e.g. company terms and conditions, warranty or guarantees, brand awareness in specific product market area, etc). Are you truly different or are you a "me too" company?
Fourth, are your customers repeat purchasers or one-time purchasers? Being dependent on repeat purchases means purchasing is likely dominated by experience and memory (and habit), not cognition. "Where the purchase decision is habitual, only [memory] will be engaged, and that will be bringing the long-term memory of the brand, product, advertising and usage into short-term memory processing with the choice to be made." (Ambler 2000)
Low involvement, easily substitutable good where you have no true competitive advantage requires huge amounts of affect-intensive marketing communication. Think consumer packaged goods. "Because low involvement consumers do not engage in elaborate information processing, advertising messages in such situations should emphasize peripheral, affective cues (celebrity endorsers, execution elements, and so forth) rather than factual product information." (Vakaratsas and Ambler, 1999).
Even slightly higher involvement but easily substitutable (e.g. car insurance) will still be better off with affect (think Flo from Progressive, or various humor-centric adds from Geico). Even Farmer's Insurance (who just recently started running banner ads on Premier League Soccer!) is going the route of funny. Insurance, though a seemingly serious topic with significant financial ramifications evidently depends on humor rather than careful (i.e. cognitive) consideration.
High involvement goods still need affect, because even in B2B situations the distance between ad exposure and purchase are likely to be significant. Affect increases strength of the associations in memory, making them more likely to be recalled at the appropriate time. But because there are other people involved in the decision and the individuals exposed to the ads are likely to be highly motivated, cognitive aspects are appropriate (although they may be used to rationalize pre-existing beliefs). In other words, "they will be "seeking 'rational evidence' to guide or support their existing inclinations..." (Hollis, 2014).
If you truly have a competitive advantage, incorporating that advantage into ads is reasonable, but don't lose sight of affect. Otherwise, emphasize "distinctive assets" (Romaniuk and Sharp, 2017), brand assets like logos, slogans, etc. to derive what has been termed a "meaningless distinction" (Flo for Progressive, "We are Farmer's, bum bah dum, bah bum bum bum"). Though you may have nothing that truly you sets you apart from the completion (and businesses are loath to admit this), you still need to generate awareness, create positive affect in customer minds, and build critical associations in customer minds so your brand is retrieved from memory at the appropriate time.
Why You Should Care
The key issue is that the increasing sophistication in digital marketing and its superiority in measurement have created a marketing mindset that is potentially harmful to long term brand sales. In the process of developing our marketing plans we set up goals and objectives, but without understanding how advertising works we can easily end up with the wrong KPIs, and ultimately evaluate programs based on the wrong criteria.
There's nothing in marketing communications research that emphasizes direct response as the primary pathway for sales; in fact, it likely accounts for about 20% or so of communication. What this points to is the importance of building mental associations in the minds of consumers over the long term. "Because of [advertising's] repeated 'nudging' effect, advertising achieves its best results on market share when it maintains a continuous presence and a sufficient weight relative to competition." (Feldwick, 2016). This suggests a long term commitment, as well as persistent and consistent marketing communication.
So there's this drunk guy in the gutter under a lamp post ...
It's late at night and there's a drunk guy fumbling around in the gutter under a lamp post. A police car slowly pulls up and the cop approaches the man. "Everything Ok here?" the cop asks. "Yesh," the drunk guy says, "I'm jus' lookin' for my car keys." The cop looks across the street and in the shadows sees a parked car. "You dropped your keys right here?" the cop asks. "No" says the drunk guy, "I dropped them over by my car." "So where are you looking over here?" says the cop, puzzled. The drunk guy stops, points up at the street light and says "Because the light is better over here."
What this parable lacks as a joke (which is considerable) it more than makes up for in marketing insight. The result is an all-too-common pitfall, namely the temptation to avoid marketing tactics that are hard to measure and encourages us to look only "where the light is better." This is due in part to our inability to incorporate years of solid marketing, psychological and neuroscience research into our marketing discussions. For example, we know that the marketing principle of frequency is based on the mere exposure effect, one of the most well-known cognitive biases (and one relentlessly exploited by politicians). Regardless of the message content, repeated exposure leads to familiarity, and to belief, which leads to higher levels of trust and lower levels of risk...and trust is typically a key element in making a sale. And yet in most applications we are unable to put a solid ROI number on the marketing effect of these exposures, leaving us in the difficult position of trying to defend what we cannot be certain of.
And so in a desperate bid to avoid the obviously uncomfortable idea that we can't adequately measure ROI for all our marketing activities, we typically do a couple things. First, we alter the conversion metric to something we can measure, whether or not we can ultimately tie it revenue (i.e. downloads or retweets become a proxy for revenue). In the language of cognitive bias, when we are faced with a difficult and complex cognitive challenge we choose to put it aside and "answer an easier question" (Kahneman, 2011).
Second, we often exclude those marketing tactics that are hard to measure from our analysis. We'll report out on clicks, but leave other less tangible efforts out of the analysis.
Third, and worst of all, we decide to only engage in marketing tactics that we can directly measure, and ignore the critical elements of brand marketing. In this case we might have a better idea of ROI, but because we have whittled down our marketing tactics to only those that are measureable, we are without question "under-marketing" and without a doubt leaving money on the table. Worse, by missing touchpoints with our customers and potential customers we may be potentially harming the long term viability of the brand.
As the somewhat exhaustive (and no doubt exhausting) research above shows, we have a pretty good idea of how advertising works - but pointing to a specific campaign, even a purely digital campaign, and providing a relevant ROI is extremely difficult. Research from Teradata suggests that online purchases "...are preceded by five to ten different brand interactions" (Lineate Whitepaper, 2018). In the language of the drunk guy parable, there's just no getting around the fact that the keys are over by the car.
I should hasten to add, that this is not to suggest that we should throw up our hands and not make every effort to measure the impact of marketing programs. Rather it is to note that some worthwhile tactics or programs may require different criteria or proxies, or, at the very least, the confidence that comes from understanding how marketing and advertising really works. So while the fast food company that put an image of a burger on the side of the bus or included a printed flyer in weekly mail circulars may not know the exact ROI of that ad placement, they can at least feel confident that the decision was based on solid marketing principles (and that it absolutely worked on me).
Direct Response "Lite"
The desire to measure marketing efforts is also leading towards efforts to target consumers near the point of purchase (i.e. closer to direct response), because purchase intent is high and proximity to purchase enhances efforts at attribution. The effect of this, however, is "likely to reach people already predisposed to buy the brand...and targeting people with a high propensity to purchase 'has a tendency of emptying the pool of people in a market without refilling it.'" (Hollis, 2014) In other words, you seal the deal with a smaller number of purchasers who were possibly going to make the purchase anyway, but you run the risk of failing to build mental availability with a vast pool of future potential purchasers. Nobody likes to see an abandoned shopping cart, but "...while short-term 'sales activation' is both effective and necessary, only repetitive brand advertising has long-term, cumulative effects on the competitive strength of the brand." (Feldwick, 2016) "Successful advertising rarely succeeds through arguments or calls to action. Instead, it creates positive memories and feelings that influence our behavior over time to encourage us to buy something at a later date." (Hollis, 2011). In addition, Romaniuk and Sharp (2016) make a compelling argument that brands grow through penetration - adding to the existing customer base. Focusing on direct response lite, which emphasizes targeting near the point of purchase, potentially limits future growth by failing to "refill the pool" and build the next batch of customers. It is far better to tackle the attribution issue head on than to use purchase proximity as a crutch.
The Ever-Present Attribution Problem
In 2013 researchers at Google and Microsoft authored an incredibly detailed research paper titled "On the Near Impossibility of Measuring the Returns to Advertising." The title pretty much sums it up, but the essence is this:
"Statistical evidence from the randomized trials is very weak because individual-level sales are incredibly volatile relative to the per-capita cost of the campaign - a 'small' impact on a noisy dependent variable can generate positive returns. A concise statistical argument shows that the required sample size for an experiment to generate sufficiently informative confidence intervals is typically in excess of ten million person-weeks. (Lewis and Rao, 2013)
This only serves to increase our desire to measure the cost and revenue impact of every marketing lever we have available, and as a result are moving us towards digital direct response as the arbiter of ROI. But even pure ecommerce plays are likely understating the impact of ads and offline marketing since they are still unsure if it was the billboard on the side of the bus, or the logo on the soccer player's uniform that first piqued the interest (and may have led to the click in an email). I may download a coupon for Jack in the Box, but it happened because I saw, over the course of several weeks, an ad on the side of a bus. So even if you have a good measure for the ratio of white paper downloads-to-revenue, you still have the issue of determining what precisely led to the download in the first place.
For example, Hootsuite sent me an email recently and I downloaded their ROI ebook, and they required that I fill out the contact form with name, company, etc. Was this a conversion? Sure. But what prompted me to sign up initially? Not the email, not the ebook, but word of mouth - and the fact that they had worked their way into my consideration set as a well-known name in social media management. The irony is that they may have attributed my conversion to their email and social marketing efforts (ROI!), but it's actually the long and indescribably hard-to-measure marketing process that did it. In other words, what worked was all those intangible and hard to measure marketing efforts that they undertook to build awareness and brand salience, never knowing if they would directly pay off. My propensity to click on a banner ad had been created offline long before the email came.
The purpose here isn't to dismiss efforts at measurement or ROI calculations. It's simply to point out that our fear of ROI uncertainty may be pushing us towards tactics that we believe are more concrete in terms of success metrics. The danger is that these tactics alone may in themselves be insufficient to build and maintain brands. Forbes notes that the almost 70% of CMOs are removed from their position in less than four years. Is it possible that CMOs are faced with a no-win situation? Where boards pressure them towards the quantifiable, but the quantifiable is not sufficient to ensure topline sales success? Sometimes we need to lean on our understanding of how marketing and advertising work in order to feel confident that we are doing everything possible for our customers and our brands.
Ambler, Tim, "Persuasion, pride and prejudice: how ads work"; International Journal of Advertising, c19, pp299-315, 2000
Ariely, Dan, Predictably Irrational, Harper Perennial, New York, NY, 2009
Bagozzi, Richard P. and Alvin J Silk, "Recall, Recognition and the Measurement of Memory for Print Advertisements," Marketing Science, Vol 2 No2, Spring 1983
Brandwatch.com, "The Hidden Value of A Social Strategy", 2016 https://www.brandwatch.com/guides/hidden-value-of-a-social-strategy/
Braun-LaTour, Kathryn A, et al, "How and When Advertising can Influence Memory for Consumer Experience," Journal of Advertising, Vol 33, No 4, Winter 2004
Choisser, Craig, Positioning Private Label Brands, Superset Marketing Blog 2016, https://www.supersetmarketing.com/2018/04/23/positioning-private-label-brands/
Feldwick, Paul, "How does advertising work?" in Advertising's Big Questions, Advertising Association and Credo, 2017 https://www.adassoc.org.uk/wp-content/uploads/2017/10/AdvertisingsBigQuestions-How-does-advertising-work.pdf
Hollis, Nigel, "Have marketers forgotten how advertising works?", Millward Brown blog, June 2014 http://www.millwardbrown.com/global-navigation/blogs/post/mb-blog/2017/06/14/have-marketers-forgotten-how-advertising-works
Hollis, Nigel, "Why Good Advertising Works (even When You Think It Doesn't), The Atlantic, 2011 https://www.theatlantic.com/business/archive/2011/08/why-good-advertising-works-even-when-you-think-it-doesnt/244252/
Kahneman, Daniel, Thinking, Fast and Slow, Farrar, Strauss and Giroux, New York, 2011
Kireyev, Pavel, Koen Pauwels and Sunil Gupta, "Do display ads influence search", International Journal of Research in Marketing, Vol 33(3) Sept 2016
Lewis, Randall A. and Justin M. Rao, "On the Near Impossibility of Measuring the Returns to Advertising," Whitepaper from Yahoo, Google and Microsoft, April 2013
Lineate Website, "The Key to Solving your Marketing Attribution Problem," March 2018, https://lineate.com/key-to-solving-marketing-attribution-problem/
Manjoo, Farhad, "Facebook Followed You to the Supermarket, Slate, 2012, http://www.slate.com/articles/technology/technology/2013/03/facebook_advertisement_studies_their_ads_are_more_like_tv_ads_than_google.html
McAlister, Leigh, et al, "Advertising Effectiveness: The Moderating Effect of Firm Strategy", Journal of Marketing Research, Vol LII, April 2016
McDonald, Scott, Measuring the ROI of Marketing: A/B Tests Vs. Market-Mix Models Vs. Multi-Touch Attribution, Forbes website, Jan 2018, https://www.forbes.com/sites/scottmcdonald1/2018/01/23/measuring-the-roi-of-marketing-ab-tests-vs-market-mix-models-vs-multi-touch-attribution/#1688b1352576
Page, Graham, "Engaging Consumers Brains, The Latest Learning", Millward Brown POV, 2007 http://www.millwardbrown.com/docs/default-source/insight-documents/points-of-view/MillwardBrown_POV_EngagingConsumersBrains.pdf
Pierno, Adam, "How Social Metrics Can Lure Brands into a Repetitive Rut", AdWeek, March 2018 http://www.adweek.com/brand-marketing/how-social-metrics-can-lure-brands-into-a-repetitive-rut/
Play Devils Advocate blog, https://playdevilsadvocate.wordpress.com/2017/09/11/how-marketing-works/
Plassman, Hilke, et al, "What can advertisers learn from neuroscience?" International Journal of Advertising, Vol 26, No 2, 2007
Reed, Peter W. and Michael T. Ewing, "How advertising works: Alternative situational and attitudinal explanations" Marketing Theory, Vol 4, 2004
Romaniuk, Jenni and Byron Sharp, how brands grow Part 2, Oxford University press, Victoria Australia, 2016
Schultz, Don, "The Future of Advertising or Whatever We're Going to Call It," Journal of Advertising, Vol 45 No 3, 2016
Sharp, Byron, how brands grow, Oxford University press, Victoria Australia, 2010
Sterling, Greg, Yelp Test Offline Attribution, Searchengineland.com, Feb 2018, https://searchengineland.com/yelp-tests-offline-attribution-using-guest-wifi-email-matching-292522
Thomas, Mark, "Showing how YouTube ads drive sales for CPG brands with Oracle Data Cloud", Google Whitepaper, https://adwords.googleblog.com/2016/04/showing-how-youtube-ads-drive-sales.html
Vakratsas, Demetrios and Tim Ambler, "How Advertising Works: What Do We Really Know?" Journal of Marketing, January 1999
Van Praet, Douglas, Unconscious Branding, Palgrave Macmillan, New York, 2012