The re-emergence of contextual targeting in digital marketing

Up till a few years ago, all buzz was around keywords. When brands did a campaign that came on Google search, as the basic go-to for any marketer, they knew that they had to use keywords for the campaign around their product or service for it to appear across websites and mobile devices. Indeed, Google’s Adwords was arguably the first contextual advertising network, its algorithms targeting consumers based on the context of certain keywords they use, and displaying advertising relevant to the same.

While earlier the focus was around the positive or negative context of the keywords and ads were shown accordingly. That has now significantly changed, not least because words can be used in myriad ways. If we just take shot as an example, it can be used to mean a vaccine shot (which has become very significant of late) or a gunshot, and there are many examples of such keywords with multiple meanings, thereby making the context of the keywords ever more important.

Technology players like Xapads Media, and others, have built our own Artificial Intelligence tools to read all the content of websites and mobile applications on a real-time basis, understand its context and only then display those ads, which make sense at that time and space.

And while earlier this was text-based, contextual advertising has evolved and advertisers want to utilize it in video and image formats as well. For instance, where a publisher like Times of India has its gallery of images and videos, brands want to understand the content and context that can be seen in them and use that as a means of determining where they want their ads positioned.

Technology is moving rapidly, and companies like Xapads are creating tools that will enable us to run ads before videos, based on the context of the videos. We have recently started a campaign for a Middle Eastern market where, during the Ramadan period, we’ll be running content on the canvas only when the video is on Ramadan.

So, while earlier it had to be done manually, today’s contextual ad technologies understand what kind of videos are being played on different publishers and the tools will then showcase advertisements accordingly.

This technology makes all the sense in the coming years. The cookieless-future promised to us, when cookies will be phased out by the end of this year, means that one of the biggest tools used by advertisers to target and re-target consumers will be gone. And companies like ours are finding ways around that, as once cookies are no longer there, there will be no option except to target ads based on the context of the content.

We built our first AI-model in 2019, and we work with over 75,000 publishers, who reach out to more than 1.5 billion people globally, meaning we are helping cater to close to 20 per cent of the world’s population. Over time we have built up huge databases based on the information coming on our platforms, our partners’ platforms as well as the OEMs that we work with, and we have built our data sets around that.

We have used those terabytes of data to help inform the AI with predictive analysis on how a particular user will behave with a brand, using different signals around a user’s data, including likes, dislikes, demographics, and other indicators to help understand which individual is more likely to convert to which brand.

And that is why contextual advertising is coming to the fore again, as the AI tools that marketers have developed and are developing is understanding content, and the context of that content to help advertisers reach out to the most relevant consumer to them.    

(As told to Shantanu David)