Google is pivoting from being a mobile-first company to an AI-first company, declares Barney Pierce, Director, APAC Platforms and Display GTM for Google. He points to how Google is deploying artificial intelligence everywhere – in Gmail which now suggests replies, in its Photos app which helps you search, archive and even suggests which snaps can be junked, in Maps which is parsing local searches and has now significantly improved due to machine learning.

But what he doesn’t say is that perhaps the most lucrative deployment of AI for Google, which is expected to make $73.8 billion in net digital ad sales in 2017 according to eMarketer, is in its ad tech solutions. Pierce, who was speaking at Google’s Think Platform event in Gurgaon, does describe how machine learning is virtually powering the digital advertising landscape, especially media buying which, thanks to the rise of programmatic, affords huge scope. AI, which Pierce describes as the science of making things smart, is not just greatly improving efficiency but also the relevance of ads.

Homing in on targets An example that Pierce offers is how machine learning models can help hotels sift through customer data and group profiles for targeted and personalised messages. “Advances in machine learning have improved similar audience performance by double digits year on year. This helps companies find and reach people with similar interests,” he says.

It’s not just Google. A host of digital marketing and ad tech companies including Oracle, IBM and Adobe are now offering AI-powered solutions to improve the performance of campaigns.

Adobe, for instance, has launched Adobe Sensei, a machine learning platform. At the Adobe Symposium in Mumbai earlier this year, Adobe President & CEO Shantanu Narayen described Sensei as the magic behind the digital experiences it creates today. Sensei is now integrated with all the Adobe clouds – from creative to advertising to marketing to experience. One of the most powerful presentations at the Symposium in Mumbai was how creative professionals could play with photos a lot better thanks to machine learning. You are given suggestions on which photos to pick from Adobe Stock, and features such as semantic segmentation, which labels each region in an image, help you change just a part of the picture. So just the sky can be changed with one command.

Says Gautam Mehra, Chief Data Officer, Dentsu Aegis Network India, “AI is at a stage where it is great for specific tasks. You will see a lot of areas of creativity – such as fetching the best images, giving music recommendations, stitching video snippets, giving copy suggestions – where it will do well.”

However, on a larger role for AI in creatives at this stage, Mehra has some reservations. More on that later.

At the Symposium, several shared stories of how machine learning algorithms had helped them use behaviour and purchase data to test ad templates in real time for effectiveness. The messages could be changed to increase the possibility of conversion. Essentially, AI is enabling advertisers to be more persuasive through intent prediction, response measurement, personalisation, better targeting and messaging.

For example, take something like Adobe Target which can read consumer intent by grouping data and quickly coming up with insights that help it predict which content will do better. If a customer watches a video of an eco-friendly detergent, the retailer can quickly send a recommendation of such a laundry product. According to Adobe, data science advancements yield a 60 per cent improvement over other algorithms.

It’s already showing immense benefits in media buying and selling. According to Juniper Research, machine learning algorithms being used to enable more efficient ad bids over rea time bids will generate some $42 billion in annual ad spend by 2021, up from $3.5 billion in 2016. According to D Rajappa, Founder and Managing Partner of brand consultancy Aamrass, in the Indian context, AI-powered advertising is nascent, though it has phenomenal potential. “The entire space will get reinvented,” he says. The reason is the far more tangible metrics it offers as delivery led by the communication can be measured. He also thinks the behavioural aspect in real-time consumer engagement is being analysed better.

Ads, by man or machine? Gautam Mehra says AI and creative have a love/hate relationship. While many stalwarts have said data-driven creatives cannot replace human insights and inspiration, others have embraced it. A classic case is The Next Rembrandt, which won 16 Lions at the Cannes festival of creativity last year.

However, Mehra thinks AI in creatives cannot replace humans in creating ads from start to finish. When it comes to general intelligence, AI is not there yet. At the moment, he says, a lot of focus is on “iterative improvement” of existing creative by AI to increase scale and efficiency.

Whither privacy? Rajappa, meanwhile, raises a question about AI’s use. “The whole privacy issue comes into play. Your entire lifestyle is up before a bot,” he points out.

Thanks to AI, the advertiser probably now will get to know you better than you may know yourself. The rumblings over ethics have begun. But that’s another story.

Related Topics
comment COMMENT NOW