Still smarting over the global brand safety crisis, which had multinational corporates pulling advertisements from Google and YouTube mid-March, Google is to introduce more ad filters that would identify and remove violent extremism and terrorism-related content.
Given that digital trust has been shaken by a proliferation of malicious content and data breaches, it has posed significant consequences for brands that tend to regularly use these platforms. Machine learning has come to the fore to tackle brand safety.
Two categories
The search giant is set to introduce two new categories of advertising filters, powered by machine learning, that can be used to block automated ads.
The first one will be an sensational ad filter, which will block ads that often look like click-bait. Often, these tend to feature celebrities, who bait-the-web viewer. Though these types of ads do not necessarily violate Google’s policies, the company is eager to hand over control to content companies.
Then there is the suggestive ad filter, which aims to block ads that are risque or feature spicy, playful and racy images. The company is to use machine learning to identify any obscene connotations. The new safeguards are set to come into force by the year-end, the company said in a blog post.
Already an integral part of business, machine learning has begun to thrive in advertising. By gathering and analysing consumer data, and utilising machine learning techniques, advertising companies are able to optimise messages and tailor them directly to consumers.
At a recent meet in Mumbai, Sapna Chadha, Head of Marketing, Google India and South-East Asia said that explosion in consumer data is the main force behind machine learning and artificial intelligence, enabling marketers to better target their consumers.
With marketers also calling for more accurate measurement and attribution standards, companies have decided to step up to the plate and bring in several checks and balances, especially in the case of digital advertising, “since consumers tend to do so many things online.”
Some 180 million Indians tend to watch YouTube, Google’s video sharing platform, every month on their mobile. With 400 hours of content uploaded to YouTube every minute, “finding and taking action on violent extremist content poses a significant challenge,” the company said. Google used machine learning to tackle the number of videos it has removed for violent extremism.
New buzzword
The company pointed out that as technology advances, data collection and consumer engagement tactics are set to become even more nuanced. With digital consumption in India growing 400 per cent year on year, attentive reach has become the new buzzword for advertisers.
A survey conducted by the company with Majestic Research, using eye-tracking technology, showed how the first ad in a TV is watched actively by 50 per cent of people, and how the last adv is watched actively by only 13 per cent of viewers.
(The article first appeared in The Hindu BusinessLine.)