Revolutionising Experiential Campaigns With AI: A Conversation With TaggLabs Founder Hariom Seth

In the ever-evolving landscape of advertising, TaggLabs has emerged as a trailblazer, spearheading the fusion of AI technology and creativity. In an exclusive conversation with us, TaggLabs' founder delves into the inspiration behind their pioneering AI-driven campaigns and the challenges they surmounted. From the audacious use of AI in the Cricket World Cup campaign to the delicate balance between human ingenuity and artificial intelligence, this discussion provides a glimpse into the future of experiential campaigns and the transformative power of AI.

Excerpts: 

How did you come up with the idea of using AI for advertising? 

There are two parts to this answer. One is that Tagglabs as a company, we were intrigued by the possibilities of AI for almost three years before it became very big this year. We were generally curious and were tinkering with stuff. So, we're already ahead of everybody else in the market in the AI game. That was one part. Another is TaggLabs, because of its pedigree, which has been a marketing technology company for almost a decade. Early this year, when OpenAI Chat GPT, Midjourney and all these things started getting news coverage, some of our clients started asking us for ideas where AI could be used in their largest campaigns. So, it started with the ask. 

What were the challenges you faced while designing the advertisement campaign for the Cricket World Cup 2023? 

For this campaign, we had to throw our hat off the wall. What I mean by that is that from day one, from the pitch stage to the execution stage, there were numerous challenges. The primary reason it was a very challenging campaign is that people have not done this before, we had no template, nothing to refer to and learn from. For this campaign, we used a tool which was available, published, and stable, but nobody in the world had automated the whole process. The tool was like software where you had to import a video and then manually select the face you wanted to swap, and then upload another JPG. It was like editing a video. The challenge was to build a website where all these things which usually a VFX guy would do, would be automated. I would say the client was very courageous, ambitious as well as empathetic and they gave us the required time to build a whole pipeline from scratch. 

One of the most important achievements of this campaign is that there are very few people across the world who have automated a VFX pipeline using AI. We might be one of the only ten teams in the world who would have cracked this thing. 

How did you measure the success of the advertisement campaign? 

I think it is a monumental success. The fact that the client not only invested in the technology, but also invested in getting three brand ambassadors Rohit Sharma, Surya Kumar Yadav and Jaspreet Bumrah. They also spent multiple crores of rupees shooting an ad between three locations in Bangalore, Bombay and London because Rohit Sharma was not in the country while this ad was being created. 

I believe six or seven million people have seen the video online. Every day, thousands of people are creating their personalised ads. Every minute we are processing tens and hundreds of videos.

What were some of the key takeaways from this project? 

So the key takeaway was, first of all, being true to the idea. We did not craft this campaign just to use fancy technology. We wanted to do a campaign where millions of people can create personalised content with the help of technology at a very low cost. And the technology became the enabler of the vision. AI was the natural answer to the problem. However, the client was not adamant that we use AI. He had a very clear problem statement and a very clear goal. Finding a solution to the real problem rather than force-fitting and reverse engineering a problem for an existing solution.

Another is if something audacious has to be attempted which has not been done before, you need a very collaborative process and transparency. The only reason this campaign has been successful because we were 100 per cent transparent with the client, how we were building it, and what are the challenges we were getting. The transparency almost removed all the pressure of being successful. The client was supporting us all the time because nobody had done it in the past. The whole team has to have the courage and the commitment to make it happen. So we had a bunch of challenges, mostly technical, because the technology was not available. Your team has to stand by you because big things cannot be accomplished by a single person. 

How do you see AI transforming advertising in the future? 

AI is transforming advertising. We are working with some big celebrities where we are create AI-based body doubles. That means the celebrity will only give one day of shooting time and on other days a lookalike will be used and the face swap technology will reduce the number of days invested in the shoot. One of the core things AI is changing is it is reducing the cost of experimentation. That is the biggest way AI is disrupting the advertising industry. 

Creating high-quality content is a very expensive affair. What AI is doing is that it’s reducing the time it takes to complete tasks. It is also reducing the cost of production, if in the past somebody had to do the face swap, they would have spent a lakh for just one ad film. Here we are creating millions of ad films, at just under 20 dollars per ad film.

Earlier, due to budget constraints, some ideas were not possible and because advertising has been a slightly expensive craft, only very seasoned professionals like Ogilvy were producing great ideas. AI is democratising the tools. Now you will see a lot of groundbreaking creative work coming from smaller agencies. Small teams of just four or five people will deliver some of the biggest campaigns in the world. 

How does Tagglabs ensure the quality and reliability of its AI systems and solutions? 

The short answer is commitment. Once we commit to a job, we don't care too much about the profit and loss aspect of the project. For instance, for the World Cup campaign, the original plan was to just use one big mega server for the automation process, but this campaign is becoming viral and the demand is three times to what we were expecting. Without even taking any official commercial approvals, TaggLabs itself is deploying three times the AI infrastructure at our own cost, we are multiplying the servers deployed to churn this campaign out. That is the kind of quality commitment we want to bring to the campaign because the client has taken such a big risk by investing so much money in celebrities, ads, film production, and PR, and we take that very seriously. 

How does Tagglabs balance the human and machine aspects when dealing with AI, such as creativity, emotion, or ethics? 

We think of ourselves as Iron Man. What is unique about Iron Man is that it’s not just a man and it's not just a machine. It is an integration between man and machine. Jarvis and the suit become part of Tony Stark and that's the closest metaphor I can think of in terms of our approach when it comes to AI. Jarvis is useless without Tony Stark and Stark cannot have those superpowers without Jarvis. 

The people in the TaggLabs team are very important because they are prompting the AI engines, they're designing the pipelines and without them, even the brightness tool cannot produce great results.

AI is going to remove all the mediocrity from the system because all the mediocre stuff can now be automated. You only need brilliant people who can use AI tools with their critical thinking and design thinking to produce brilliant results.

How do you ensure that your clients’ data is secure and protected? 

This is a question which is not only led by the client, but we also make sure that we ask these hard questions whenever we are dealing with the data of consumers. Whenever possible, we like to not store the data. If the technology allows us to complete the delivery of the content to the recipient without storing the data, we would always like to prefer this route. That means the data is only consumed to generate your content and deliver it to you, but no copies of your data are stored in the database. Whenever the technology is such where storing data is required, we have a very strict protocol where whenever dealing with hyper-personalised campaign data, as soon as the content is delivered to the customer, the data is destroyed. We don't like to store data in our systems beyond 24 hours. 

What advice would you give to other companies looking to use AI for advertising? 

One is that the simple advice would be curiosity. Any marketer needs to be very curious. The curiosity will lead to an in-depth understanding of the logistics and the details of something like this. 

Another one would be heavily networking with the upcoming engineering companies. Right now, big brands only interact with very big agencies. Sometimes innovation happens in very small startups. So, anybody who's looking to deploy cutting-edge stuff should welcome startups to be part of their campaigns. Because sometimes these disruptive campaigns will only come from people who are doing it for the first time. 

Lastly, don’t follow the fads. Usually what happens is, that whenever there is a buzzword in the market, we want to get the story out. We want to be the first people to have used AI in a certain thing. In this pursuit, we sometimes kill the golden goose. We deliver such a shoddy job that marketers become scared of technology. A lot of marketers and brand managers are very scared of technology because they have interacted with certain companies that have not delivered good results. What happens is, in the long run, it becomes difficult for companies like us to convince customers to make very big investments. I would like the practitioners to be very responsible and double down on their homework and their efforts to make stuff happen. 

What’s next for Tagglabs in terms of innovation and technology?

I think we have just crashed the tip of the iceberg with AI for advertising. We might be one of the first people in the world to have cracked it. We would like to double down on our efforts and learn this craft in detail and depth. We’ve been working on one of the first completely AI-generated music videos, for the song ‘India India,’ which is sung by Vishal Dadlani. We are releasing it in the next two, or three days.

We have something coming up with a big cricketing celebrity. As we get all the success and the fame, the idea is to not stop learning.

We want to be the torchbearers of AI and advertising. That's our goal. We don't just want to enjoy the fruits of being famous, but we want to contribute too. If anybody wants to do something relevant, we are happy to share how it's done and what the best practices are so that more of such work can come out in the industry.