When I began my career in communications and brand research, I thought of my previous careers as trunks to be stored away in the mind’s attic—perhaps pulled out in a far-distant future when I had time for reflection, meandering beach walks and writing all those letters to all those editors out there.

Now, as a senior researcher, I realize there is no such thing as life-trunks, understanding today how often I’ve drawn upon my past as a teacher and editor of creative writing in the study of advertising. One of the best parts of my days—and, gratefully, it makes up a good chunk of my actual job—is to share how the writing and visual construction of the advertising we study is inextricably connected to the data. As understanding story is in the core skill set of everyone in our research consultancy, I have a lot of conversation partners.

This was brought home to me this week when I introduced, for the newer members of our team, our seminal work studying emotional response in advertising. Yes, it’s true. The idea that emotion was a vital measurement in ad effectiveness was once under debate. We were on that debate team—also known as the AAAA/ARF Emotional Response in Advertising Committee—and firmly on the side of emotional measurement, as we had been doing it since our inception, now 25 years ago.

I’m proud to say the eight-member co-committee worked hard and made progress, beginning as all good researchers do with base-lining the measurements themselves through studying the same set of beer advertising. Once that was done, we each got to select the next case study. The time between when I received the email and called the committee chair was roughly equivalent to the pause between late-stage delivery contractions. I already knew what I wanted. I wanted to study the BMW online film phenomenon, what was then labeled “branded entertainment.”

This was the case study I brought to our internal team this week, revisiting the success of the overall effort, and taking a close look at two of the films we studied, to understand why one was indeed branded entertainment and one was simply entertainment. In order to bring current this work, I demonstrated how the exact same principles applied to one of our recent tests of digital Facebook advertising.

What’s the unifying principle that worked then and works now, whether a 7-minute film or a film on your feed? A story construct called The Controlling Idea.

A phrase used by screenwriting guru Robert McKee, the controlling Idea is simple but not easy. It is the story’s ultimate meaning, as expressed through the story’s action and emotional climax. As a poet and essayist, my personal process has never included an outline or even something I could refer to as a direction—rather, it’s been more a directive…to simply write. That said, understanding the ultimate meaning of what I’d made on paper or on screen was a critical step in having it see the light of day, much less an audience. That bit just came later and was the compass which directed all revision.

In understanding my work, my former students’ work, and the examinations of advertising work we study, across all forms of media, we examine the effectiveness of the controlling idea by working backwards.

We do not begin by searching for the story’s ultimate meaning. We don’t start there, because we don’t know what it is—we can only examine the story that was shown and felt, and draw conclusions on meaning. Fortunately, we have data to support those conclusions. In fact, we have visual data using the story itself to calibrate action, feeling, and meaning. That helps. A lot.

Each time, we start with the “story’s action” as it is expressed in the advertising. In short, what happened? We humans do this all time, whenever we describe to someone else a movie we’ve seen or an experience we had. We tell that person what happened. That’s the action. Using our Attention graphs of the film as parsed by the audience’s collective mind, we can see what they saw.

Then, one can move on the the “emotional climax.” How did it end? When telling a story, everyone is moving to the “high” or “low” of the end. For our research, we can see that high or low by the graph of the film that charts accompanying levels of emotional response, moment by moment as the story ratchets forward.

Only then can one look at what happened and how it ended to discover what the ultimate meaning of any piece of creative, verbal or visual, truly is. And that’s where it gets tricky.

With my students, I would play back what I saw on the written page—the action of the story and its ending. I would ask questions. As beginning writers, there were often a lot of those. And in our meetings I would almost always get into a discussion with the writer that sounded something like “well, that’s not what I meant!”

They would go on to explain a back story to me, about who the character is, what her motives are, and so on. I would listen, waiting until they took a breath and ask them one last question: can you put that on the page?

And this remains true in advertising evaluation. The controlling idea is not what you meant to be the ultimate meaning of the ad. The controlling idea ends up being the ultimate meaning of the ad you actually made.

All the meetings on brand strategy, KPI’s, emotional levers, and whatever else, all of that and more are hidden behind the curtain. A very thick light-blocking curtain that the audience cannot see through. That’s why examining every story, after it is created and before it meets its public, needs to be examined with the discipline of looking only at the action and its ending. Then, using your most grown-up voice, ask what the story you just examined must then mean.

If that’s not what you wanted, revise and edit. And if it is exactly what you wanted, and it matters to your public, the data will support that. And, I hope you’ll trust me on this. We’ll be first to applaud.

Amy Shea is a Senior Research Consultant at Ameritest