mediasmart’s team is continuously expanding in different regions and as a part of the #NewHireDiaries initiative, we bring you news of our team...Read More
Announcements, analysis and opinions on industry trends around the mobile programmatic world.
When I started working in the advertising industry almost six years ago, I was very excited to bring innovative technology to an industry that seemed so far removed from it. I naively thought at the time, that if we developed technology that allowed advertisers to buy media in the most effective way, and squeeze engagement from every single impression, everybody would want it!
I quickly learnt that the advertising ecosystem doesn’t always reward companies that invest in the best technology and want to deliver the best offering – especially when it comes to the arena I know best: media buying.
As a DSP, you have two main options when it comes to establishing a business model. You can offer your technology (as a self-managed or managed service), charging a fee for its use, or you can manage campaigns for your customers, charging for the execution of each campaign. In either of the two models, if you have chosen to focus on delivering the best technology to bring value to the advertiser, you are likely to be fighting an uphill battle.
Let’s say you are one of those players that decide to charge a fee for the use of your platform, as opposed to charging per campaign. The most common model in the market in this scenario is to charge your customers a percentage on top of what they spend through your platform. Negotiations often revolve around this percentage, with customers focusing on bringing it down as much as possible. But who benefits most from this? Companies that offer the lowest fixed percentage, of course and this creates a faulty dynamic for our industry.
Firstly, this ‘percentage’ is meaningless without the ‘base’ that it applies to. Such base varies widely depending on the DSP and its own cost structure – and so it happens that companies that offer the lowest fixed percentage can also be the most expensive taking into account absolute costs. DSPs with a lot of direct connections to supply, for example, may achieve a much lower cost base by having fewer middlemen to the final publisher, but because this is not deemed to be valuable, the reward is for DSPs that don’t invest in direct connections. The customer will pay more in absolute terms – the DSP commission and that of the middlemen – but who cares if the only thing taken into account in negotiations is the fixed percentage on top of the ‘spend’?
Secondly, this model gives no incentive whatsoever for DSPs to invest in algorithms that enhance performance. If an algorithm achieves more engagement for the same impressions, with lower effective CPCs and CPAs, customers focused on those KPIs will actually have lower spend, (which is great for them) but the DSP will make less money with the fixed percentage model. Again, the model rewards the DSPs that don’t invest in technology that maximizes engagement.
If all DSPs were to move into a fixed percentage model, and be pressured into offering increasingly lower percentages in order to compete, they’d all end up with absolutely no incentive to invest in tech to enhance campaign performance. DSPs that are set to thrive are actually those making minimal investments in differentiating technology (and they may not even have much tech at all) – as well as on direct connections. DSPs that focus mostly on marketing and sales versus technology can also offer lower fixed percentages by having lower operating costs.
Now, let’s say you have decided to take the other approach and offer a service to your customers selling campaigns that you manage for them. If you are a mobile first DSP, chances are that the biggest budgets you have access to come for app promotion campaigns. Such campaigns are often very performance focused, with advertisers looking for a highly quantifiable return on their investment, and often following key performance indicators such as the price per ‘install’ or per other ‘post-install’ events. An alternative economic model could be for advertisers to pay fixed amounts for these KPIs.
You would think that companies who have focused on developing great technology to effectively buy programmatically, with sophisticated algorithms that managed data to maximize engagement, would have a gold mine here, right? After all, they should be able to deliver the best results… If they are paid a fixed price per goal for each campaign, then the better the technology, the more money they’d make, one would expect. But that’s not the case.
Demand side companies that have found the gold mine here are those who take advantage of loopholes in attribution systems and play these to their advantage. By engaging in what can be plainly called ‘attribution fraud’, some companies find ways of generating great results for the campaigns they promote without spending much. Their campaigns seem to generate many installs and post install events, but in reality, those come from users that would have downloaded the app and used it anyway. Users who haven’t used the app as a direct consequence of the ads… This is obviously terrible for advertisers (who pay for acquiring users that they would have acquired anyway) and also for the industry because it creates unrealistic expectations about acquisition costs - making demand side companies that don’t engage in this unethical practice, unable to compete.
In summary, when agencies and advertisers working with brand campaigns focus on low percentages as the criteria to choose a DSP, the true value gets lost. Sophisticated algorithms maximize engagement, and advertisers who lose sight of this over perceived cost end up losing. Ultimately, they end up paying more without the underlying technology. By working less with companies that do invest in technology they’re also encouraging bad industry practice, which exacerbates the problem further.
And when agencies and advertisers working with direct response campaigns ignore the wider impact of potential abuses such as attribution fraud and work with companies that deliver the best KPIs without considering the technology that boosts them, shady practices will just keep on proliferating.
DSPs trying to bring the most value through innovative technology find themselves between a rock and a hard place in this environment. It saddens me to see an ecosystem that rewards those that bring the least value to it, and it infuriates me to see an ecosystem that rewards those who are plainly cheating. Awareness of these issues among advertisers will hopefully change the situation, but it may be too late for many small players who were trying to do the right thing. - Rant over!
Article from Digital Marketing Magazine