Burning Questions: What is AI’s impact on engagement?

shutterstock_363790325

One of the goals of our technology and innovation coverage is to answer the questions our readers have about what is changing in that space. We’ve asked marketers from across Canada to pose their burning tech questions directly to us, which we have in turn posed to experts from the industry for answers.

“How will the rise of AI impact media consumption and advertisers, and when will this happen?”
– Paul Cuaso, marketing manager, General Mills Canada

jamieOur expert: Jamie Anderson, SVP and CMO at SAP Hybris. As a company, SAP is probably best known for providing enterprise software related to commerce, marketing, billing, service and sales. But at the Hybris division, the company is exploring how innovation and new tech is impacting not just those areas, but consumer experiences. Anderson is giving a talk today at The Gathering conference in Banff about “The Sci-fi Future of Customer Engagement,” which will, among other things, tackle how artificial intelligence will redefine marketing and the customer experience.

strategy: I think we can break this question down into three parts, so let’s start with how AI could impact the media consumption habits of consumers.

Anderson: I think it already has. The tracking capabilities exist to capture what users are doing and to identify them. Maybe not at the level of a specific individual, but I can build an anonymous profile of a customer through digital engagement and map that profile to other customers who followed similar patterns. Those patterns of behaviour are what’s picked up by machine learning and intelligent algorithms that will drive the AI that promotes the response. It’s not happening as widely as it could be, but the tools and products and power to make it happen does exist today.

And what kind of impact will those tools have once they get more sophisticated?

We’ve seen the statistics about the ridiculous amount of digital ads that get ignored, and that’s because they’re not relevant. Relevance in a digital world means ads shouldn’t be just about the interest a customer has had in the past. Yes, that’s part of how you build a profile around a customer, but it’s also about what they’re doing at a specific moment in time. Are they trying to solve a problem? How and why are they interacting with you? What device are they using and where are they physically? Using AI to help understand all these things will help drive much better contextual engagement. There’s still great merit in traditional, mass brand awareness if that’s your desired outcome. But if you’re trying to achieve a more specific engagement with a customer, then you have to be more contextual.

Is there a cycle that gets created? As AI becomes more prevalent and sophisticated, will consumers have higher expectations for how relevant ads are to them?

I think it does, and there has to be slightly more work done to make the algorithms more intelligent and decipher the differences between things like someone who is shopping for a new baby in the family and someone who is actually pregnant, for example.

But you have to think beyond that and to what the outcome you want is, and that is when AI really comes into its own. What makes an ad successful is when you actually get someone to engage. And when you think about something like chatbots as an example, we’re talking about things that can change and configure an online experience in real time. They’re collecting information and using that information to change what gets fed back to a customer, based on what it has used AI to determine will get them to engage the most.

So the real big exciting change in consumer behaviour is less in targeting and more in engagement.

Correct.

So how is the rise of AI going to impact advertisers? What should they be doing or thinking about to use AI to better achieve those outcomes?

Attribution is the big topic people are talking about right now and linking dollar spend to results. It’s not good enough for a brand to appear in a Facebook feed because of a cookie someone has picked up on another site. It’s about how you make that ad get an engagement, tie that back to the original advertising and prove that it was effective. That, to me, is the next step.

What’s also largely untapped is the ability to marry the data from within your enterprise with the data outside it. Some of the richest data about your consumer is outside the four walls of your organization. Finding a way to blend that and engage with the consumer on a more regular basis and get them to identify themselves to you – by creating a profile or account or something like that – is when you start to build a much richer and real time view of the consumer.

This is where it get interesting. You can process an incredible amount of data, in real time, aggregated from multiple sources to create a real time view of the customer at a given moment. So you can target digital ads to the customer because you know its when they’ll be on their device, it makes it much more effective. But you’re collecting so much data all the time that the way in which you message when you do, for example, your big out of home campaigns should be more intelligent as well. Because you’re able to adapt faster to changing trends, you can change while it’s in flight.

You already said this is already happening, so I’ll augment the last part of the question slightly. When is AI going to become a standard, the kind of thing that powers most of the advertising a consumer sees?

I think this is about advertisers and agencies as well to transform the way they look at engaging with customers and understanding what they are looking for. The power of the networks and connectivity and the technology that’s powering it, it all exists today and is hear. It comes down to how comfortable the companies themselves are with digital transformation. It takes innovators to change the markets, because then it becomes the standard. It hasn’t yet, but it’ll take someone to do it consistently well and intelligently. When it happens is difficult to predict, and it might end up happening in a sector-by-sector basis. I can see it happening in retail and CPG first, though.