DAIVID – our AI saviour from bad advertising
In this latest series we speak to tech leaders who offer solutions within the research, data, and insights space. To shine a spotlight on disruptive technologies and better understand the value they offer.
At Day One Strategy we combine technology with human intelligence and pioneer innovative AI led solutions. In this instalment Day One Founding Partner, Abigail Stuart, and Senior Research Director, Ben Lorkin, spoke with Ian Forrester, CEO and Founder of DAIVID – a company which takes AI led automation to the next level to reveal what’s happening in a creative and how it’s capturing attention, evoking emotions and how that can predict behaviour.
Ian, tell us a little bit about DAIVID…
DAIVID is a creative effectiveness business, and we focus on attention, emotions, and AI. For a piece of content to be effective it needs to capture attention, because without attention nothing else matters. Once you’ve captured attention, you need to evoke an intense, positive emotional response.
And why is evoking an emotional response important to advertising effectiveness?
The emotional response creates a memory structure, which drives action. The problem with many ads is that from an emotional point of view they’re average. Two or three seconds later you’ve forgotten the ad. The memory structure is not created, and the action is not driven. But when a brand can elevate that emotional response, the viewer remembers the way the ad made them feel and attaches that feeling to the brand.
It’s interesting that you also used the term ‘positive’ emotional response. I’m curious about that because some advertising, such as government campaigns, can often leverage negative emotions such as fear and shock
And not just government, some brands use shock and disgust, for instance, which can be very effective. It’s a great point, as the emotional response doesn’t have to be positive. Negative emotions can be used to amplify positive emotions. But in most cases, I would recommend that there is a positive resolution at the end, because that’s what gets remembered and attached to the brand.
Tell me more about your unique approach?
By combining facial coding, eye tracking and survey data with computer vision we’ve created a system which allows us to predict the attention captured and emotions evoked by content. We can ingest an ad, understand what’s happening in it automatically and predict performance without that ad ever needing to be viewed by respondents.
How have you trained the AI?
The AI is trained on thousands of videos, each viewed by 350 nationally representative respondents. By gathering very deep response data from each respondent – second-by-second eye tracking and facial coding and survey responses – we get a profound understanding of how people are responding to content. We then understand what is happening in videos by sending them to computer vision APIs (e.g. Amazon Rekognition). We then have two datasets – what’s happening in videos and how are people responding to them. Our system then correlates the two, understanding which creative attributes resulted in certain responses.
How culturally relevant is this technology?
In each country we need to collect a unique training dataset. We launched in Japan in the summer, and you can’t apply a UK algorithm to Japan. So, we’re now collecting Japanese training data by showing Japanese content to Japanese people to see how they respond.
How long do you think it will be before you don’t need human-beings?
We are always going to need humans to train the system. People’s response to content isn’t static. It’s influenced by world events, and cultural norms shift more quickly than they used to.
What are your next steps?
Soon we will be in a position where clients can go onto the platform, upload creative content, press a button and within seconds they’ll receive their attention, emotions, and memory data.
Wow, you’re blowing my mind! And of course, clients have content beyond just ads…
Nowadays every client and brand team under the sun is creating so much content. Entire in-house teams are making content and the number of assets per campaign has gone through the roof. And yet, only the top 0.1% gets tested in the traditional way because it’s too expensive and slow.
So, your goal is to help more clients deliver more effective content?
Absolutely, it means they avoid pushing out content which doesn’t have the right emotional response. There are all kinds of potential uses in government, public service messaging, UX design, web design, product creation etc. We’re building a system which can tell us how to evoke emotions. And emotions are critical to behaviour change. So anywhere someone is trying to influence behaviour our system could be useful.
DAIVID uses AI led automation to reveal what’s happening in a creative and how it’s capturing attention, evoking emotions and how that can predict memory
Day One Strategy is a specialised healthcare market research agency that combines the latest technology with human intelligence to better understand customers’ wants, needs and behaviours. We start every day as if it’s our first, full of enthusiasm and a desire to try new things. If you feel the same and would love to leverage the combined power of tech and human intelligence for smarter, faster insight get in touch: hello@dayonestrategy.com