Creating a New Category: Emotion AI

Handwritten thank you notes. These aren’t words you see a lot these days, so when I heard Gabi Zijderveld, CMO of Affectiva, describe writing 44 such notes in one night to the speakers at a conference she’d organized, my interest was piqued. She told this story at a recent gathering of The CMO Club, not to brag but rather as a reminder that a little extra effort can go a long way, in this case, helping Zijderveld generate substantial word-of-mouth about a new tech category her company was promoting.

Digging into the background behind the notes and the conference, I discovered how companies can redefine the business they are in and use cost-effective ways to spread the word. In this portion of our interview, Zijderveld details how the idea of “#EmotionAI” came into being and the transformative role it has played for Affectiva.

What was the situation like when you first joined Affectiva?

At the time that I joined the company, we were considered a niche vendor in the market research space within media and advertising. The technology there was being used to basically test how people react to ads, so that you can optimize your ad content and was referred to as “automated facial coding.” What even is that? This is well-established terminology in the market research space, and they commonly still refer to this type of technology as automated facial coding. It works for that niche market.

But you all had a bigger vision?

As a vision for the company, the leadership, the founders, and everyone there had this very fundamental and very strong belief that this technology—artificial emotional intelligence, as we now refer to it — would one day be ubiquitous. If you think about it, detecting people’s emotions and cognitive states through technology is something that can indeed be applied to many different industries in many different use cases. There was always this really big vision. The problem was that we had this technology being described in what I thought were very clunky ways. The challenge was how to position this technology as something that has much broader applicability across industries and can indeed one day be ubiquitous.

That’s a big challenge. So, then what?

The strategic decision was to try and reposition the company as something that folks in other industries could better recognize or maybe even better identify themselves with. How can we actually describe our technology and name our technology in a way that people actually understand what it does? It all came down to articulating a clear and simple value proposition.

How did you land on #EmotionAI as this value proposition?

It was certainly an evolution over a couple of years. When I first came in, I started asking a lot of questions. So “automated facial coding” — what actually is that? What does that mean? What do we do? You have to ask those basic questions. At one point in time, several people said that we recognize and detect emotions. I thought that sounded simple enough. It was emotion recognition, so why were we calling it facial coding? Then, I spent a lot of time with key stakeholders in the company, including our CEO who was out on the road talking about this new cutting edge technology. We realized that we were doing everything an AI company should be. Our CEO and I came up with #EmotionAI in a few minutes before a talk, and it stuck for the first time. It was an “a-ha!” moment. It’s naming the value prop in a snappy way that people actually recognize and understand.

How did you start spreading the word about #EmotionAI?

The way we then executed was very deliberate and thought through. First of all, we completely updated our talk track—our 2-3 page document that describes in various ways what we do, what the use cases are, what the value prop is, and the data points that support it. We also put some definitions in place as to why we are an AI company and what artificial emotional intelligence is. We started using the hashtag, #EmotionAI, on social media. A few weeks after our CEO delivered a keynote using the term #EmotionAI, she bumped into an investor who had heard her talk and wanted to be involved. So, we started saying that we were building a new category here. This was a new frontier in AI that hadn’t been built yet, and we were building it.

Was there anything that made you hesitate to go all in on this newly constructed value prop?

No, because we realized eventually that we had little to lose. People will always have automated facial coding. We will always have our work in the market research space because we were recognized as the market leader there. We weren’t giving that up. There were naysayers, even within the company. A number of our technical folks were very skeptical because everything is labeled as AI these days. Then, we started getting picked up in the press. Our CEO was writing columns. We were posting blogs, and we just kept repeating it and repeating it, and it started sticking.

What lead you to creating a conference around Emotion AI?

After we had established this emotional AI technology category, we started seeing our competitors pick up the language. At that point, we realized that an annual conference would be a great way to bring curious industry leaders together. In terms of advancing this new technology category of emotion AI, it was critical that we build the ecosystem of supporters around ourselves. They were our clients, our business partners, and also industry analysts.

How did get the speakers and did you have to pay them?

We started asking influential thought leaders across industries if they would be willing to speak about their work in the context of trust and AI, which was the theme of the summit. We did not require them to promote Affectiva at all. We did keynote presentations and panel discussions. We did not pay anyone a single dime, apart from the few nonprofit organizations. An event like that really helps you foster the relationships when creating a new category, and it leads to collaboration and innovation.

Did this program help build your sales pipeline?

In the market that we are in right now, it’s not volume sales that we use to measure the effect on company revenue. It’s not about building a massive pipeline. In the early days, it was first and foremost about making sure that people at target companies knew who we were and what we did. Now, our sales team has established relationships with all those target accounts. It sounds very counterintuitive to most marketers, but sales is not telling. We don’t need new leads. We don’t need new companies to work with.

Wait, what? Why don’t you need new companies to work with?

We need to advance these relationships because going forward, these are our customers that we will get into production vehicles with. We often use how many of our automotive invitees actually show up to an event as a measure for growth, or how many of our clients actually attend our demos. We also get a tremendous amount of press activity, so we routinely measure our coverage and reach of coverage. There’s a lot more being spent on product marketing and making sure that we have enough technical content because many of the automotive companies look to us for guidance on how this technology can be applied to future vehicles. I’m way more focused on product marketing and communications management than on building a pipeline and lead generation.


to the Latest in

Renegade Thinking