Memory has been and remains a favorite topic of Hollywood, for good reason. It can act as a powerful and dramatic plot device, setting up severe obstacles in films such as The Bourne Identity and Memento. And it functions simultaneously as a character study. It dramatizes the question, “If I can’t remember myself then who am I?”

Consider this exchange, from the film Regarding Henry:


Henry: No thanks, I don’t like eggs.

Rachel: Eggs are your favorite!
Henry: Okay, then give me lots of eggs.

In Regarding Henry, Harrison Ford’s character – his way of behaving and responding to life – takes a dramatic shift when he loses his memory. Not knowing who he was, he does not know who he is. He is left without a story about himself and he must now write a new one – with nothing to go on.

We tend to think of memory as a kind of record-keeper of the past. However, it is actually memory that informs our image of who we are in the present and has a tremendous impact on decision-making that affects our future. It is up to memory to enable the brain to make predictions about the possible consequences of future behaviors. From an evolutionary survival standpoint, memory is all about the future.

And, as it turns out, it’s also about the future of brands.

Brand success is often connected to well-used measures like share-of-market and share-of-voice. And they are, indeed, useful measures. Share-of-market is a measure of brand results, an output. Share-of-voice is a measure of brand effort, an input. The value of a share-of-voice metric lies in its ability to predict share-of-market.

But exactly how does one aggregate and calculate the “share-of-voice” for your brand’s advertising investment if it’s spread across Amazon, Google, Facebook, Instagram, Twitter, television, print, banner ads, outdoor, radio and all the other emerging brand touchpoints available to modern marketers?

And how does one account for the contribution to sales from the increasing variety of advertising formats deployed across all these media choices – from six-second Facebook ads to three-minute videos on an Amazon sales page?

Most importantly, share-of-voice calculations often leave out the impact of the quality of the content, i.e., the creative – the key variable which, according to many quantitative marketing models, explains more than half of the variability of in-market results.

This brings us to the importance of measuring that quality and the role of memory in how researchers might differentiate the most effective brand communications from those that have little or no impact.

Shift the demand curve

When a company invests in advertising it is not just trying to drive short-term sales transactions. It is buying a stream of future profits. The best of this kind of advertising builds brands that shift the demand curve to a higher level of profitability.

At its best, evaluation of brand advertising tries to help brands predict future sales. All advertising researchers have developed numerous ways of measuring success, whether evaluating it against another execution, campaign or an average, but understanding branded communications cannot stop there.

Since memory has such an impact on the “present” self who is making decisions, it follows that memories that self has of a brand represents a kind of equity – for better or worse – impacting decision-making and, thus, a brand’s future. Understanding more about the effects of brand memory on brand-building must be part of our understanding if we as researchers are to reflect the real processes of the human mind.

Our view is that marketing a brand is, at its core, about memory.

Like the Japanese game of Go, the object of which is to capture more territory on the board than your opponent, the object of building brands is to capture more positive memory space devoted to you in the mind of the consumer than do your opponents.

Professor Byron Sharp, author of the successful marketing book How Brands Grow, writes about a similar concept in his discussion of mental “availability” as the key to brand success.

The idea that the story people carry in their minds about a brand, or even that they had given precious memory space to a brand at all, has been thought of before. In fact, it was the basis for a widely-accepted measure of success in the beginning of ad-testing days: recall. It was abandoned as lacking predictability to sales.

This is not surprising when understanding how memory works – a subject we researchers know a great deal more about today.

We now understand that the ad memories researchers want to access in our visual-drenched world are non-verbal and often difficult, if not impossible, to put into words. Asking someone if they remember an ad is not at all the same to the human mind as showing them an image from an ad and asking if they recognize it. This visual cue taps into a memory retrieval process far more graceful and reliable than the verbal “search” function of the brain.

Perhaps most important, however, is that now there is well-established science that demonstrates memory is far more nuanced then we previously understood. There are three major memory systems in the brain, not just one: semantic, episodic and procedural. When asking about an ad verbally, one is querying the semantic memory system, the one of the three systems weakest in its connection to emotion.

The episodic memory stores felt experiences, the “episodes” of one’s life that one draws on for decision-making – from the smallest to the largest things. Accurate emotional measures are needed to understand what made an impact in any brand’s communications, as most advertising researchers acknowledge.

To understand procedural memory, think of how you know how to open a door or eat a meal or perhaps drive a car. This is the place that physical rehearsal resides and it’s vital to our survival. It turns out that rehearsing a brand experience through advertising activates mirror neurons in our brain and can become a part of our procedural memory. This has been studied intensively and used in behavioral psychology to help humans heal.

Therefore, with our practice of querying memory through visuals and seeking to understand which of the three systems are triggered, we were led to what we felt was a next-wave question, and study, of branded communications:

If in advertising research one can identify the most powerful visual branded memories, and can attribute them to a particular memory system, can we then confidently predict that those are the visual moments, whatever the memory system, that are building brand equity for the long game?

The research

To advance the body of knowledge in the field of branded memory, Ameritest and our sample partner, Research Now, operating under the aegis of the Advertising Research Foundation, recently conducted a study of long-term advertising memories in the quick-service restaurant category.

Nine years had elapsed between the time these fast-food ads were aired, and tested, in 2009, and when we went back, in 2017, to measure them again. We wanted to discover which, if any, advertising memories from that time still remain in people’s memory. And, importantly, to see if those powerful visual moments identified in that 2009 research had any connection to what was remembered.

2009: QSR category ad sweeps

In 2009, Ameritest was operating a syndicated ad-testing service in the quick-service restaurant (fast-food) category. That year we tested all of the 30-second TV commercials that aired nationally in the QSR category. There were 268 adult targeted ads for 17 national brands, including McDonald’s, Burger King, Pizza Hut, Sonic, Wendy’s, etc.

The timing for this, our baseline year for our new study, was serendipitous. That year was shortly after the birth of social media, and Facebook and Google advertising were not yet significant factors in the media mix for fast-food advertisers.

Each commercial was tested among 100 fast-food consumers the week it first aired – so the sample for this baseline study, recruited by Research Now, was 26,800 consumer interviews.

Each ad was tested with a standardized survey, lasting 15 minutes and including the topline measures of attention, branding and motivation. Importantly, as well as verbal diagnostics, it included our visual diagnostics: frame-by-frame measures of memory (recognition) and associated emotion (felt experience), which were collected about 10 minutes after ad exposure.

2017: Share-of-memory tracking

In May 2017 we conducted this new follow-up study among 5,500 respondents, again using Research Now, to match the category usage and demographic sample of the 2009 study (adjusted for aging of the original target audience).

The stimuli used in this long-term memory research study were the same 4,050 (unbranded) still images originally pulled from all 268 ads for the 2009 ad tests and used in the original visual diagnostic testing. Each respondent saw a sub-sample of these visuals, drawn at random.

Respondents were asked if they recognized seeing each image before in any fast-food advertising, and if so, with which brand they associated that image.

Requires more study

This does not pretend to be anything other than the beginning of our mutual deep-space journey, exploring what impacts longer-term brand memories. This work, as we knew it would, requires more study to be carried forward by the industry as a whole.

What we did learn is that nine years after exposure, the average recognition level of images that appeared in fast-food advertising in 2009 was 12.5 percent. Importantly, we also found that images remembered and in the top quartile for positive emotion in 2009 were recalled at an even higher rate (17.2 percent) than the average in 2017.

This indicates that emotion plays a role in what makes it into long-term memory, contributing to brand equity. This is not surprising, given the significant amount of work done by the research community in the study of the power of emotion in advertising to drive decision making, and fits with our own data as well.

When examined through the lens of brand, we calculated the share of ads on air each of the QSR brands had in 2009 (a surrogate for media spend, as we didn’t have that data) and then examined the share of memory by brands – in essence, the brand attribution of those remembered images.

The big winner was Sonic. The brand has kept to its visual iconography in its advertising through the years, featuring two people conversing and eating in the front of a car. While the people, products and dialogue in the ads changed, and thus the ads are different, this consistent approach became an equity for Sonic. This indicates that consistent visuals can be kept in memory and help create stronger brand connections.

McDonald’s suffered the most from poor branding, with its share-of-memory less than its share of the ads it had on-air.

The study outcomes have some strong indications, which can contribute to hypotheses to explore in ongoing research:

  • This study indicates that the effects of powerful advertising do not decay as quickly as we may have imagined.
  • Quality communications – as measured using a comprehensive visual, verbal and emotional research methodology – appear to have a long tail when it comes to being worth keeping in memory.
  • Poor branding in the short-term appears to create misattribution to other brands in the long-term. Even if those category memories are kept, other stronger brands may get the credit.
  • It appears that two memory systems dominate the longer-term space: the emotions of the episodic system and the rehearsal or physicality of the procedural memory. The semantic, or verbal, memory system may contribute the least, depending on the category.
  • Branded communications that decay more slowly in memory can be reasonably hypothesized to make a greater contribution to sales over a longer period of time. This, in turn, would result in a greater return in the original advertising investment.

Understanding success

When focusing on aggregating images and their memories, we are no longer restricted to focusing on traditional “ads.” Visual-level assessment is what is required, as a very first step, to understanding success and branded memory deposits.

The long-term brand memory study reported in this article was about television advertising. Today, that view would be far too limited and unrealistic. For advertising running this year, the individual images used in the study would come from any kind of visual branded communications format, running on any platform, in any country.

This initial work makes the case for the closer examination of a shift to visual-based tracking systems if one is to understand what brand memories are being filtered for and found meaningful in the long game of brand-building.

Images are now the “facts” of what your consumer is remembering. This is the visual language we are speaking today and, as it turns out, we are all wired for it.