Guest Column: Gautam Mehra, Chief Data & Product Officer- dentsu (APAC) & CEO- dentsu Programmatic, SA, explains how the 2020 decade seems to be all about artificial intelligence

Every decade has its own phrase that basically dominates the narrative. The last decade was all about “digital” and the 2020 decade seems to be all about “AI”. I’m sure all of you would have read several articles, blogs, newsletters talking about the “threats” and “potential” of AI. Artificial Intelligence almost seems to be the solution to all our problems and also the origination of almost all our problems!

“Artificial intelligence (AI) refers to the simulation of human intelligence in machines that are programmed to think like humans and mimic their actions.” This definition while accepted for a few decades in the scientific community is turning out to be problematic. The reason is that by this definition, we seem to want to “replace” certain human interventions in a process that require our cognitive skills. If machines can do these cognitive pieces, then “humans” don’t have a role and will get displaced. This has been a long-standing debate between economists, the tech giants, scientists and politicians. Some say AI will create new jobs, some say it will destroy more than it can possibly create.

That’s why a newer definition of AI and a slightly different way of looking at AI advancement is now catching on in the scientific community and I feel businesses need to adapt this as soon as possible as well.

Rather than thinking of big data that is available in marketing (Search signals, social data, 1st party data, 3rd party data, etc.) as this all-knowing wise sage that will tell us what to, we should think of it as a powerful tool that needs to be wielded with care. That it can help us spend more time on strategic and business discussions, rather than on CTR of display campaigns or cost per app install of their performance campaigns. This seems obvious, but in reality, we hardly follow this.

Let me give you a real example. At dentsu, we created something called the dentsu Marketing Cloud. This helps our teams and clients generate data-driven audience insights, translate that to an activatable digital plan and also create optimisation opportunities post going live. At first glance, it might seem, why do you even need a whole agency team if the software can do all this for you. The reality is actually quite different.

The thing is that you can put just one keyword into the planning system and the country you want to target and it can create an entire plan based on it. So, now the problem of finding the best, most efficient activable segments is solved. However, a new problem arises, what is the right keyword to begin the process with. One small change in the keyword or an additional interest added and the entire plan will change. So, instead of training people on how to use some platform-specific tool of Facebook or a DSP, we are training our people on how to ask the right questions to the client, decode the brief better and to get at the right core keyword. We can also add a filter of a demographic and that again will change the entire plan. Hence, instead of creating one plan, we can see what the effect of lowering the age range might have or vice versa. The point is that we now have intelligence at our fingertips that can do analysis in seconds, which would have taken us weeks. Consequently, it can be augmented by us to create far superior and efficient plans that just wasn’t possible before.

Coming to another example, is of data-driven creative. This has again been a heavily debated topic in our circles. Does data and AI enrich creativity or stifle it? I choose the former and here is why. I don’t think that a machine is going to spit out the big idea by itself and I think as humans, we of course can, but it is a laborious and often not scalable feat. How do we turn the problem on its head and allow our most creative people to be given prompts, tools, and techniques that can help them churn out more ideas? How can we give what we traditionally think of as not-so-creative people, a chance to impact the “creative tweaks” based on all the audience segmentation data we get. Both are possible. For example, with our tool DMC Explore, a simple brief can tell you everything there is to know about an audience. What they like, what titles they watch, which songs they listen to, where do they swipe their cards, what apps they use, what games they play all based on real-time data pulled in from every corner of the digital ecosystem.

In addition to this, you can get what questions they ask about a product/category, what their opinions and reviews are, and what are the predicted search/purchase trends are and you have all the information available to figure out the “what” (creative) and the “who” needed to come with up brilliant ideas. What’s more, is that today, with Computer Vision (a technology that allows computers to “see”), we can not only tell you which creative is working but also create a data-driven “guess” on “why” creative A is working against creative B by breaking them down into hundreds of elements and analysing their performance at an element level. Hence, you could guess that is having a celebrity working or not adding as much value in a creative. Or do outdoor panoramic shorts work better than indoor studio shots?

All in all, it is a very exciting time. The shift from artificial to augmented is a game-changer in my opinion. As marketers, we should embrace it rather than fighting it. It can augment our intuition not disregard it. It is meant to enhance our creativity not restrict it. It will make us more efficient not replace us.

Disclaimer: The views expressed here are solely those of the author and do not in any way represent the views of pitchonnet.com