There is a great potential for programmatic advertising in Africa, especially since the continent is seeing a growth in smartphone penetration, which grew to over 26% in 2018. As at 2019, there were 525 million internet users in Africa, with an estimated 80% of the 800 million people in sub-Saharan Africa expected to have mobile devices by the end of 2020.

Photo by Markus Spiske © from
PexelsWith such an unprecedented increase in mobile and internet access, it is understandable that the surrounding sectors would be ripe for advancement in technology. Media buying traditionally was the physical negotiating and securing of media real estate on billboards, in magazines and on street poles.
Mobile devices have shifted the attention and much valued “eyeballs” into the world of banner ads, pre-rolls and interstitial advertisements. To some extent, the same logic relating to negotiating, agreeing a price and flighting media still applies, but at a much faster, more insightful way - just like magic.
Programmatic advertising in itself is ever evolving, especially with the growth in machine learning; a piece of software can now handle the sale, placement and engagement of digital ad impressions in split seconds through an ad exchange platform. Programmatic advertising creates an online intersection for advertisers wanting to buy ad spaces online and publishers who want to sell these spaces. We believe that we will see an increased appetite for programmatic due to businesses looking for effectiveness and efficiencies out of their media investments.
Andrew Smit and Johan Walters 6 Jan 2021
More value, for less
Covid-19 has created immense constraints on consumer demand in most markets, and in response, brands will be looking to get far more value for a far reduced spend. As a direct response to market conditions, this is where programmatic can be advantageous as programmatic advertising is all about improving cost and resource efficiency throughout the media buying process. The real secret sauce of programmatic advertising is that it leverages on data traffic and other online targeting avenues to serve as impressions and efficiently reach a wide audience. For advertisers and publishers, having a system that accurately accounts for impressions and efficiently handling this at scale would translate to better return on investment (ROI).
With programmatic advertising, there may be exponential growth in sight for African advertisers; publishers and advertising agencies who run ad exchanges can now optimise the auction process and connect with advertisers in real time.
Benefits of programmatic advertising for African advertisers
So in light of these key challenges, one needs to ask the question - ultimately, what control does a brand have over a platform?
In reality, once a programmatic system has been put in place, for the most part, it would feel like a “black box” for an advertiser who is not well versed in the technicalities. Automation by its very nature implies initial control over the inputs, but very little control over the process, which is often the work of algorithms and dashboards to monitor. With artificial intelligence, there is an opportunity to streamline ad management and achieve a higher degree of control. In reality, many businesses may not have budget to invest in AI, so they partner with agencies, who become the gatekeepers to the “black box” that creates a feeling of no control.
Are there sufficient processes to detect fraud?
According to Andrew Buckman, MD of EMEA Sublime: “The programmatic ‘black box’ prevents true visibility of where ads are served, and almost 80% of advertisers who purchase through programmatic are concerned about this lack of clarity – making it all the more vital that brands understand the supply chain and choose partners they can trust.” Lack of transparency gives rise to ad fraud and many brand-related malpractices. While that in itself cannot be resolved in a day or two, it seems like the industry is moving in the right direction.
We are seeing technologies like ads.txt and sellers.json make the supply chain more transparent for everyone. It is also worth noting that Google recently made the use of sellers.json mandatory for all Ad Manager users. With that, this standard is expected to witness better adoption.
We also believe that AI can go a long way in fraud detection and, to some degree, correction as well. Some experts argue that viewability could potentially pose a bigger problem to digital advertising than ad fraud. Essentially, with programmatic buying, you bid on every impression.
This means that you are paying for every ad displayed on a screen. Whether this ad is seen, the advertiser pays. Even if you have solved the problem of fraudulent inventory, you are still unsure whether the advertisement that you are flighting is actually shown on a screen as it might be ‘below the fold’, in a slot where the user has not scrolled to. With machine learning, historical data is used to improve the future outcomes.
In developing economies like Africa, some pitfalls of programmatic can become a deterrent to its use and likely compromise on the value ultimately to be received by advertisers.
We are hopeful that some creative solutions can help turn programmatic from a potential “foe” to a “friend” using machine learning to not only have better ad targeting, dynamically adjusted CPMs and ROI predictions, but also to detect and correct fraudulent occurrences.