From the very early days of the internet, the content recommendation has been a major driver of traffic for websites and online news platforms. \r\n\r\nOne of the earliest companies globally to become a recommendation juggernaut was Taboola Inc. The US-based \u201ccontent recommendation\u201d engine this week announced the expansion of its India operations to meet the growing content discovery demand here. \r\n\r\nTaboola partners with advertisers, publishers and mobile carriers to help reach their target audience online. It claims to reach over 1.4 Bn people globally with one billion unique users every month and will open up new offices in Gurugram, Mumbai and hire 80 employees to lead its India and global efforts. \r\n\r\nIn terms of interactions in India, Delhi leads the charts for Taboola in monthly unique views with 12 Mn followed by Bengaluru and mumbai with \u00a08.7 Mn and 7.3 Mn respectively.\r\n\r\nThe New York-based company, which entered the Indian market six years ago, currently has a strong relationship with most, if not all, of the top news media outlets in the country. \r\n\r\nInc42 sat down with Ran Buck SVP at Taboola, who not only set up the company\u2019s India operations but has also led the company\u2019s efforts in the South-East Asian region. Buck opened up about the company\u2019s strategy for the growing trend of Indian regional language content and how it plans to tackle the problem of fake news and misleading articles that have brought content discovery platforms under increasing scrutiny. \r\nTaboola Drives India\u2019s Online News \r\nFor many years, Taboola has been a serious player in India in the news industry as content discovery drives traffic to news websites and protects the revenue model, which has been drastically affected by aggregators such as Google News, Apple News and social media sites Facebook and Twitter. \r\n\r\nWith the global news industry still struggling to figure out the best business practices when it comes to monetising content, discovery platforms are giving them the visits and growth that keeps them going in the meantime. \r\n\r\nTaboola does not charge publishers any fee for its discovery platform, but instead enters into a revenue-sharing agreement based on the ads it gets for them. With the advertisers, Taboola charges a fee based on cost per clicks (instead of impressions) and for a video it charges a fee only when a minimum watch time of 30 seconds is achieved. Taboola has about 1,100 employees worldwide and has raised $157 Mn in funding.\r\n\r\nThe company recently signed a massive five-year, INR 300 Cr deal with NDTV. "The five-year deal ties NDTV Convergence exclusively to Taboola and involves a minimum guarantee of more than INR 300 Cr for NDTV Convergence, making it one of the largest deals not just for digital content but for the media space in its entirety," the news of this tie-up had sent NDTV\u2019s stock soaring by as much as 10%. \r\n\r\nTaboola has also partnered with the likes of the Indian Express, The Hindu, Zee Media, NDTV, India Group, Times Now, and ABP. Further, it has partnerships with electronics original equipment manufacturers (OEMs) like Sony, Vivo and Huawei to bring personalised news and content from its network of publishers to users of their devices.\r\n\r\nSome have described Taboola to be an alternative to the likes of Google and Facebook whereas the truth is that it is much more of a hybrid company that combines content aggregation and digital advertising but is also an analytics company driven by huge sets of user data which it has accumulated over the years.\r\n\r\nTaboola\u2019s artificial intelligence (AI) and machine learning (ML) algorithms are used to help increase the effectiveness of the content it promotes on partner websites, and the analytics from this helps brands tailor their digital content.\r\n\r\n \r\n\r\n\r\n\r\n\r\n\r\nCustomisation Vs Vernacular Content \r\nIndia is seeing an explosion of regional language content. According to a KPMG-Google study, by 2021, an expected 201 Mn Hindi users \u2014 38% of the Indian internet user base \u2014 will be online. The same report said that the growth momentum of Indian languages is likely to continue with the regional language internet user base growing at a CAGR of 18% to reach 536 Mn by 2021. In comparison, the English internet user base is growing at just 3% annually, and will reach 199 Mn by 2021. \r\n\r\nWhile Buck does admit that the company is heavily investing in the vernacular content category, the larger trend he chooses to focus on is customisation of content for the user. \r\n\r\n\u201cThere are nuances with users from different countries but there are some universal trends like we have observed such as finance people globally will click on a sports-related content in their second or third click. We aim at tailoring content for consumers based on their age, what time of the day they are accessing it, and even what mood they are in,\u201d said Buck. \r\n\r\nTalking more about the regional language side of the business, Buck adds that for every network Taboola signs, it also gets to work on the publisher's regional language entity like for example- India Today (Aaj Tak), Hindi editions of Indian Express and Hindustan Times. Apart from this, the company works with the likes Malayalam Manorama, ABP News, which are pure non-English news outlets. \r\n\r\n\u201cFor me, every person who would like to promote content is a customer. We have a self-service checkout service that we launched about six months ago through which a customer can transact with us using just a \u00a0credit card. There is a small and medium business team (SMB) team for the small ticket size players that included the publishers as well as media business\u2019,\u201d said Buck. \r\nFake News & Fake Promises \r\nIndia has been on the frontline when it comes to the menace of fake news with people losing their lives over rumours spread via WhatsApp. The problem is not only limited to social media and messaging platforms but extends to news websites too. \r\n\r\n\u201cThe issue of fake news is very broad and what we decided was to organise a policy team which is very rare among content companies and we currently have 50-100 people sitting in Israel who go through the content on the platform and each one of them is given a designated language,\u201d Buck told Inc42. \r\n\r\nThat number seems incredibly small given the tremendous amount of data that Taboola gets but Buck insists that Taboola is taking enough steps to keep track of this issue and stated that over 60% of the content that Taboola gets from all kinds of advertisers are rejected because they don\u2019t comply with its standards. \r\n\r\nBut wouldn\u2019t it be easier to have a deeper reliance on AI and automatically detect troublesome content rather than scan manually. \r\n\r\n\u201cI can use technology to autodetect images and title but the ability to identify contextual aspects related to the article is something we already have,\u201d Buck said, referring to the manual checks. \r\n\r\nThis strategy runs contrary to what Taboola\u2019s peers are doing which is increasingly rely on AI models to help battle violent and misleading content. Facebook (which has 15,000 human reviewers) \u00a0has had great success in detecting and blocking pornography and nudity but its efforts are still lacking which was recently at display in the aftermath of the New Zealand and Sri Lanka terror attacks. \r\nWe work with only the premium publishers who are serious about the news business and are into serious journalism, unlike crowd-sourced network and so that helps \u00a0- Gomti Shankar, country manager, Taboola India. \r\nAnother way Taboola hopes to combat fake news or misleading content is through attribution. Every piece of content that the platform promotes has clear attribution of the source which has been especially helpful in dealing with political advertisements. \r\nQuality Over Quantity?\r\nOf course, content discovery is also about staying up to date with trends. Whether it is social media or news, the general approach by tech companies, after a whole host of privacy and data-related controversies over the past couple of years, is quality over quantity. \r\n\r\nIn March, Apple launched a revamped \u2018Apple News\u2019 subscription service which pulls in news stories from the web through RSS\/Atom feeds, displayed in a new Apple News Format. The service is integrated directly into the News app and has the likes of \u00a0National Geographic, Popular Science, New Yorker, Vogue and Rolling Stone on board. \r\n\r\nThe company runs its own news service called Taboola Feed, a social media-inspired, curated content feed that lets users scroll through a list of articles. This feed is curated by a mix of AI and human moderators. \r\n\r\nDiscovery platforms such as Taboola have been taking note of this trend and are trying to take steps to make sure their recommendations are relevant to the user who is increasingly fickle when it comes to choices in an especially competitive space.