New Delhi & Mumbai: Last Diwali, a big surprise was in store for both consumers and small shop owners. Movie superstar Shah Rukh Khan appeared in an ad promoting the neighbourhood kirana shop, or jeweller, or photo studio, among many others.
How could the small business owners afford it? They didn't, actually. Cadbury Celebrations' chocolate ad during Diwali leveraged artificial intelligence (AI) and machine learning to make Khan appear in multiple versions of the same ad for different shops.
To feature their businesses, shop owners had to fill a form, and name their shops and products. For Cadbury, it was a goodwill service for its vendors' benefit; shopowners only needed to give consent, and no fee was charged.
In a statement on its website, Wavemaker Global, the ad agency behind the campaign, said that it used AI to localize the campaign to 500 pincodes with 2,500 local business owners. "Working with our tech partners, Rephrase.ai and DeltaX, we used machine learning to recreate SRK's face and voice to take local store names in the ad. Every time a Celebrations ad was served, the user would see and hear SRK encourage purchases from specific stores that were in close vicinity,” the agency said.
It's early days yet, but AI has clearly begun to enable brands and ad agencies to create highly individualized advertising campaigns that grab attention, both at the planning and execution stages.
Typically, with digital advertising, companies struggle to get the attention of users who are more often than not trying to skip ads, especially on platforms that have live streaming or video games, unlike on television. So, when an ad announces a shop near you that resonates with you, it becomes easier to grab that 10-20 seconds of attention, say advertisers.
Vishal Jacob, chief transformation officer at advertising agency Wavemaker India, said AI has made advertising far easier than before.
“If we were to manually produce a thousand different creatives for one company, it would take a massive number of man hours," said Jacob. "But now we have a number of tools which can personalize content to result in multiple permutation combinations like say, ‘a model with a smiling face or one with a serious face’, helping us decide what works best for us at the concept stage. Producing the campaign then at the actual shooting or production level becomes faster and easier.”
He added that across the three broad functions in an agency—planning; how to reach out to the audience; optimize and measure effectiveness of a campaign—have been impacted in a positive way. Right from the early stages of formulating an ad plan, everything has now changed.
AI sifts through massive data sets, learning consumer behaviours and preferences on an unprecedented scale that can create hyper-personalised advertising, where campaigns seem almost clairvoyant in their accuracy. It can also study buying patterns to make campaigns that hit closer home.
In an earlier conversation, Dheeraj Sinha, FCB Group chief executive officer, India and South Asia told Mint that when users see ads online for products they recently searched for, it is personalisation. But today, companies need to do personalisation at scale or engage with users while still being creative.
“Companies realised that just uploading 20 or 100 posts a day was not actually leading to any engagement. We are now working on creating more engaging campaigns for brands using machine learning and AI for assistance,” he said.
"AI in advertising is no longer just a buzzword," said Chandrasekhar Mantha, partner and media and entertainment leader at Deloitte. "It's driving efficiency and innovation across the industry." With its continued enhancement, it can now handle vast data sets and perform complex tasks too. He added that AI enabled more efficient and large-scale task execution.
Oindrila Roy, managing director, Publicis India, said the company has spent a substantial sum of money to develop an in-house software that is at the core of its work.
"We are all getting more AI-fluent and whenever we see an opportunity, we use it. AI is going to sit at the heart of everything we do. Not sure about the next five years but in another three years, my hunch is, freshers who join the workforce won’t know what to do without it. People who are currently in the workforce will need to stay in tune with the times. Only those who grasp AI well will stay in the forefront," she added.
Globally, Publicis has set aside €300 million over the next three years in CoreAI, to put together data from across its operations under a single entity.
Food delivery company Zomato used Hrithik Roshan to endorse a local restaurant that partnered with it, too, in its campaign, "Mann Kiya, Zomato Kiya". The campaign used “faceswapping” and replaced restaurant owners' faces naming their own cuisine and restaurant's name with the actor's face to make it look like he was saying those words.
“These campaigns can have a much higher return on investment, because companies can keep changing the messaging based on the type of audience they have. Globally, the AI market in advertising is $25 billion and will grow 4x by 2028. While it is still new in India, we will likely also follow the global growth trajectory,” said Aviral Jain, managing director of valuation advisory services at risk management company, corporate investigation and risk consulting firm, Kroll.
Despite its benefits, AI's rise brings significant concerns. Privacy intrusions are a major issue, as AI's capability to delve deep into personal data can be unsettling.
In India, where the Digital Personal Data Protection Act, 2023 has recently been passed, there is significant anxiety about data misuse without stringent legal safeguards. The infamous case of Cambridge Analytica, which involved the unauthorized use of personal data for political advertising, remains a stark reminder of how data can be exploited.
But Roy of Publicis cautioned that another change will be in the implementation of guidelines to contain ethical issues. Deepfakes, privacy concerns and biases could be possible areas of concern. AI tools are still fairly nascent and some of the outputs could be inaccurate.
“Our group has a lot of guardrails in place and the whole industry will need to train its people about these; people have to learn to be more mindful when using AI tools,” she added.
Mantha agreed: “Despite some scepticism regarding issues like content originality, IP ownership, workforce rationalisation, and the misuse of AI with deepfakes, regulation and industry self-governance are expected to address these concerns in the near future. Ultimately, the benefits and transformative power of AI in advertising and marketing are likely to outweigh the apprehensions."
