South African advertisers could be being defrauded of between R2.8 billion and R12 billion a year through programmatic online advertising, according to TruthsetsOnline.com’s analysis of local campaigns.
Marc Dhalluin, co-founder of TruthsetsOnline.com, estimates South Africa’s digital advertising market at between $2.1 billion and $2.4 billion, or about R35 billion to R40 billion, with programmatic advertising accounting for an estimated 40% to 50%.
Applied to the estimated programmatic market, that would put potential losses at between R2.8 billion and R12 billion a year.
While Dhalluin stresses that reliable industry-wide South African fraud figures were difficult to obtain, analysis by his company of multiple campaigns for local businesses found wastage in programmatic advertising commonly ranging between 20% and 60%.
“I can’t say this is representative for the industry. We have done analysis on multiple campaigns for South African businesses, and these results are typical. So far in SA, we’ve seen worse, not better.”
Dhalluin notes that digital advertising fraud broadly included non-human activity generating impressions, clicks, leads and affiliate transactions, as well as manipulation of advertising impressions themselves.
The forensic media analytics company’s analysis of advert traffic involves inserting a JavaScript tag into the ad, which then logs data for its lifetime, says Dhalluin. This, combined with a fingerprint made up of the location, IP address and device – or a fingerprint – of impressions, enables the company to flag potential fraudulent ads.
Dhalluin adds that these hits are only classified as “confirmed fraudulent” after at least three suspicious indicators. TruthsetsOnline.com’s technology tracks ads after they are served, examining where they appear, geographical location, internet service providers and devices.
In a live campaign involving a global advertiser, about 6.6 million impressions have been analysed, with 41% confirmed fraudulent and 12% confirmed legitimate, Dhalluin tells ITWeb. Among the most recent 20 000 impressions, confirmed fraud fell to 27% as the campaign was optimised.
International clicks
Fraudlogix’s analysis of a dataset comprising 105.7 billion impressions collected throughout 2025 reveals a global invalid traffic (IVT) rate of 20.64%, translating into around one in five ad impressions displaying characteristics of fraudulent or non-human activity.
“When we look at 105 billion impressions and see consistent patterns across devices, browsers and geographies, we’re not looking at anomalies – we’re looking at the structural reality of ad fraud,” says Jessica Lopez, COO of Fraudlogix.
This, the fraud and bot detection company says, is “a persistent challenge with significant financial implications for the industry”.
Digital advertising fraud has been a problem since at least 2016, with US senator Mark R Warner of Virginia last March, in correspondence with Andrew Ferguson, chairman of the US Federal Trade Commission, stating he had pointed the issue out a decade ago.
“I voiced my concerns about bots plaguing the digital advertising space by creating fake consumer traffic, artificially driving up the cost of advertising in the same way human fraudsters can manipulate the price of a stock by creating artificial trading volume,” writes Warner.
Warner adds Fraud Blocker estimates that digital advertising fraud accounts for $84 billion of the $700 billion global digital advertising industry.
Hundreds and thousands
Website traffic can be bought cheaply online. Fiverr currently lists more than 2 200 website-traffic services, with some sellers advertising thousands of targeted or “organic” visitors for as little as $5.
Netotraffic openly sells website traffic through what it calls its ad network. It offers unique IPs, country and niche targeting across more 120 geographies, mobile traffic, social traffic and bulk orders, with campaigns starting at $15 and most orders starting within hours – although it stresses this is bot-free.
Fraud rates in campaigns without fraud mitigation reached 10.9%, 15 times the 0.7% rate among campaigns using anti-fraud technology, Integral Ad Science’s 2025 Media Quality Report found.
The advertising verification company’s latest report, published in July 2026 and based on more than 300 billion digital interactions analysed daily, found advertising risk and waste increasingly concentrated in mobile web display, which accounted for 45.1% of impressions but 71.9% of made-for-advertising impressions.
“Ad fraud isn’t a technology problem anymore – it’s an incentive problem,” says Hagai Shechter, CEO of Fraudlogix. “The tools to detect and prevent IVT exist. What’s missing is the industry-wide commitment to use them consistently. Platforms need to stop treating fraud detection as a cost centre and start seeing it as a competitive advantage.”
Catching lizards
Fraudulent operators could create artificial traffic or manipulate how advertisements were served, says Dhalluin.
He explains that techniques included stacking multiple advertisements behind a visible advertisement, serving advertisements into zero-by-zero pixel spaces, or generating advertising calls that were never designed to produce a visible advertisement.
Advertisers can be charged for impressions with no realistic opportunity for human engagement, says Dhalluin.
“We’re not talking about waste through dropped connections, poor planning or misguided knowledge. We’re talking about the systematic diversion, in all online campaigns, of significant portions of an advertiser’s budget. By criminal means.”
Dhalluin argues conventional industry measurements did not necessarily detect the full range of fraudulent activity. “Our point is, if you go into the garden to look for lizards, you will not catch all the reptiles. You will catch lizards.”
Share
