It is now known that the Twitter deal of Musk was abondoned on account of underreporting of fake accounts. Twitter claims that less than 5 per cent of its accounts are fake, but according to global research agencies, it is three times the quoted number. It affects the valuation of the company. It should be noted that 95 per cent revenue of Twitter comes from advertising.

There is the use of bots. Bots are used by companies to automate some of the functions, say monitoring criticism or keeping tabs on complaints. However, bots are also used to view ads. Companies pay for real audience and not machine audience of their ads. It is a fraud. Primarily bots push some other agenda, but its unattended consequence is to scan the ads, thus increasing the audience. The technology (tech) behind creating a bot is so simple and easy that it is proliferating like anything.

Advertisers spend money on these falsified numbers, and are paying for bots who are treated as an audience of human beings.

mFilterIt is an ad fraud detection and protection company in India. It believes that Google as search engine has the lowest average fraud rate — 10-12 per cent. It means that of their total viewership, 10-12 per cent is generated by machines. It means advertisers lose a similar percentage of money.

Google Ads Network and Facebook Audience Network tie-up with other digital platforms to insert impressions on revenue sharing basis (70% to website and 30% to social media company). Here the ad fraud rate is higher — 18 per cent. On YouTube, it is 17 per cent.

On click-on ads, the rate is between 15 and 20 per cent. Some e-commerce sites, publishing agencies, sports platforms, the rate touches 30-35 per cent. In India, the percentages are higher than the global average.

Over 45 per cent fraud happens through the bot route. Then there are other methods — click scams, say 25 per cent. Here malware is inserted in the device to generate ad clicks, and domain spoofing.