Yet, despite its obvious benefits, in an overcrowded marketing space, everybody wants to be seen. To demonstrate value to their advertisers, these platforms must prove that they are providing the type of audience engagement and return on investment that justifies the cost of advertising. As such, the manipulation of metrics has become a point of concern for many in the industry.
The impurity of digital metrics
Manipulating the system is not a new concept, but in a digital world it has taken on a more advanced hue. Most people will know principal-agent theory, an economic and political framework that describes the relationship between a delegator (principal) and the decision-maker (agent). In an ideal world, their priorities would be aligned. However, increasingly in a digital world, the principal and agent have different priorities and, more importantly, different information. This means that while the platforms offer an array of metrics for marketers to analyse, the agent and principal are working with different data and misaligned agendas.
In research conducted by the United Kingdom’s Competition and Markets Authority (CMA), it was noted that, historically, there have been concerns over the misreporting of data by the big media platforms. The research found that advertisers are unable to independently assess the relative merits of advertising across platforms and are forced to rely on the metrics being provided by said platforms. The CMA was concerned that this practice prevented fair competition and could see advertisers overpaying for their advertising.
The trend was even more prevalent in “walled garden” platforms – closed ecosystems where tech companies control the ad inventory, user data, and campaign measurement – such as Google, Meta, Amazon, and Apple. These platforms have the ability to obstruct or place unnecessary restriction on the advertiser’s ability to access data needed to carry out independent evaluations, thereby introducing a degree of opacity onto an already complex process.
While the platforms provide metrics, behind the scenes the numbers are being manipulated. Examples of irregularities include platforms that, through a complex array of analytics tools, know which consumers engage through clicks and likes and which consumers bypass adverts altogether. As these platforms get remuneration from engagement, they will only show adverts to the people who engage with them. Therefore, there is a misalignment between reported vs actual engagement of the advert.
The platforms are boosting advertising reach through “targeting” consumers who either use adblockers or who actively pay not to see adverts. What the platforms are doing is that, as an advert starts to load but is blocked, in the nano-second when the ad is triggered, it is registered as a full load. This means that every advert, blocked or not, is recorded as a view.
Another tool being used to boost numbers is the use of bot farms – automated software programmes (bots) designed to manipulate online activity at scale – to generate traffic. While bot farms can be used to push up the engagement numbers for a site, a news article, or a platform, they are also used by companies to:
- Improve product reviews. Research that Prof. Morris Mthombeni, Dr Avi Ramphal, and I did on the sharing economy found that it is impossible to find a recommendation on Airbnb under three stars and Amazon has long been accused of hiding negative reviews – meaning platforms and business are bot farming their way to positive reviews.
- Give companies the ability to sabotage their competitors’ campaigns by having bots register engagement numbers, which eats their marketing budget but has zero impact as the campaign is not reaching anyone.
Now, with AI, we are also seeing platforms bringing even more ambiguity into the system, by getting AI to regularly update terms and conditions. This keeps marketers off balance because the foundation of their relationship with platforms is fluid.
More spend, less impact
Digital marketing’s flaws do not stop there. The system is becoming increasingly complex. Today, the number of intermediaries who are required to facilitate a single transaction have grown to include a publisher, ad server, supply-side platform, ad network, ad exchange, demand-side platform, multiple data management platforms, third party verifiers, ad agency, and finally the advertisers. This “ad-tax” can be as much as 58 cents of every dollar spent on advertising, almost four times the traditional 15% agency commission. PwC and the Incorporated Society of British Advertisers noted that 15% of the ad-tax spend termed “unknown delta” – meaning it was not attributable to any vendor – was simply absorbed into the system, by hidden margins, currency arbitrage, and data discrepancies.
In South Africa, for example, the digital advertising market is valued at around R17.8 billion annually, representing 40% of the country’s total advertising spend. A recent report by TrustList found that 40% of South Africa’s programmatic ad spend was wasted, compromised, or was actively funding misinformation. It went on to say that R600 million was misallocated annually.
In addition, the efficacy of digital advertising is being questioned. In 2025, Michael Farmer did an analysis of the market in a blog entitled “Madison Avenue Media Madness”, where he found that from 1960 to 2010, fast-moving consumer goods brands P&G, Unilever, Nestlé, and Colgate-Palmolive grew at around 8% compound annual growth rate (CAGR). Since digital advertising took off after 2010, the four companies have grown, collectively, at less than 1% per annum. This is a stark warning that brands’ overreliance on digital advertising may be enriching platforms at the expense of brand growth.
We need a change in approach
To boost their figures, everyone is looking to manipulate the system, through likes, shares, and comments, all to game the algorithm. By doing this, however, they are not marketing to people, they are marketing to AI and are treating the algorithm like its own principal. As one commentator said, all the world’s companies are trying to please one company, Google, to be on the first page. The end result of this is click-baiting to farm engagement.
While it is tempting to game the system, as you do, the rules keep changing. The rules are opaque, and these platforms are a black box, which can change based on platform interests. There is zero transparency. They are not telling advertisers how they get the metrics, what they are measuring, or how they test for accuracy. So I would argue that marketers need to reassess how they measure marketing success.
When it comes to digital advertising, they need to reconsider the Peter Drucker adage: “You can’t manage what you don’t measure”. Instead, they need to refer to Goodheart’s law, which states that when a measure becomes a target, it ceases to be a good measure. This is because the measure then informs behaviour, so behaviour is shaped towards the measure rather than adjusted in accordance with what the measure indicates. As the gap between goal and measure increases, the inefficiency of resource utilisation does too, and the temptation to fudge the numbers becomes greater.
Internal performance reviews also need to adapt. Dr. Clive Corder, ex-CEO of ACNielson MRA, shared a really profound insight with me when he said, “Accuracy is the enemy.” This is because if KPIs are rigidly defined, there is no wriggle room for various marketing stakeholders to define what it means for a campaign to be successful or not.
While gaming the system is not advised, one company that has cleverly demonstrated how to do this is Swedish oat-milk company Oatly, which is using offline media to win the digital game. By using clever, witty adverts painted on the walls of the busiest London tube stations, Oatly gets people to take photos of the adverts and upload them. This generates a lot of online traffic for the brand, while successfully targeting consumers through its out-of-home campaign. In our current times of dynamic digital billboards in South Africa, however, this “gaming” is now limited, as the competition for attention is transferred to outdoor spaces. Good for media owners, less so for brand managers.
The path forward
To move forward in this environment requires companies to start being in control of their data. In South Africa we are seeing the emergence of super-apps with a lot of digital convergence. For example, companies such as Discovery and FNB own their own data, which they mine through their own technology, thereby measuring their clients’ life points. They are not relying on media platforms, or the pollution of those platforms with second- and third-hand data generated by AI. Brands such as Oatly, Red Bull, and Real Madrid have also created their own channel to market, giving them invaluable first-party data and dedicated consumer attention.
Ultimately, research finds that there has to be a unified demand for change. Everyone knows what these platforms are doing but there is no collective effort to demand primacy in the principal-agent relationship. Perhaps attention is best captured by owned and earned media, rather than paid media. In a time where algorithms become even more mediated by LLMs, owning the data and having sight of the models is perhaps the best way to link marketing activity to marketing performance.
The time has come for businesses to stop playing a game designed by the opponent. The future belongs to marketers who build value they can measure and control.
KEY TAKEAWAYS
- Digital marketing has democratised the advertising space and offers many benefits. However, it is not as transparent as it is made out to be.
- In a bid to enhance engagement metrics, the system is being manipulated by media platforms, as well as businesses and their competitors.
- In an increasingly complex ecosystem, around 58c of every dollar spent on advertising goes to intermediaries.
- Marketers need to change their approach to how they view and use metrics.
- The way forward is to rely on first-party data and not consider these large media platforms as growth partners.
