advertising technology
For all their complexities, international politics and programmatic advertising have one thing in common: in the end, it all comes down to trade. And specifically, trading as efficiently and profitably as possible.
From our vantage point at the cutting edge of programmatic trading, where ad networks scour the market for tools that deliver incremental revenues to their business, trading efficiency can be the difference between success and failure, profit and loss. The good news is, a high degree of trading efficiency is eminently achievable, if you do it right.
What is trading efficiency?
In our eyes, this is the practice of delivering your demand partners the exact inventory they require, at the frequency they require, while understanding that their needs are constantly changing. To put it another way, it means generating the maximum revenue using the supply and the tools at your disposal.
The benefits of efficiency
For ad networks specifically, trading efficiency has three key justifications:
1. Strengthening your demand relationships
QPS is a measure of the volume of traffic your DSP partner is allowing you to send them. The trick is to use this QPS allocation to ensure you are sending your DSP as much of the most valuable supply as possible.
DSPs will use a performance KPI such as an Ad Request RPM (Ad revenue per thousand requests) or Fill Rate % (the percentage of ad requests which resulted in an impression) as indicators of whether the partner is being successful in their collaboration.
Good networks will understand their DSPs’ KPIs well, and be able to walk the line effectively between maximising scale (QPS) and achieving the required efficiency (performance KPI) to make their demand happy.
Most large-scale DSPs manage their overhead by cutting any underperforming QPS allocated to ad networks and ad exchanges. Essentially this means you need to maximise the impact of your allocated QPS through good results - or risk having the quota reduced.
Conversely, a platform like ours will offer unlimited QPS. This is because, from the outset, we want our clients to be able to freely explore trading opportunities without worrying about platform costs. If clients perform well, they should have the ability to scale that without their tech holding them back. Once they have settled on an approach, efficiency becomes more important.
2. Propelling your revenue potential
The more efficient your trading, the more revenue you can generate, as your demand partner allocates you more QPS and is more inclined to spend advertiser budget with you. Additionally, on a platform like ours, efficient traders also pay less, and a lower overhead means your profit margin will be higher, allowing you to reinvest, hire more sales people and create more business. Programmatic trading efficiencies, in other words, give you a cascading opportunity to make your business exponentially more efficient.
3. Building a more sustainable ecosystem
Being an efficient trader means that you're also being a green trader. In a time of net zero targets, when green credentials are becoming ever more important, an efficient model means less server capacity is required, with a corresponding reduction in carbon emissions.
So, how can you trade more efficiently?
The popular advice for those seeking business improvements of all kinds is to lean into AI, which many now consider a one-stop efficiency solution. However, AI does not dictate the supply you receive. Efficiency begins on the supply side - which is why supply path optimisation is such a big priority for so many now.
What really matters here, more fundamentally than the technology itself, is the fabric of trading: what the supply is, what the demand is, who your partners are. So where trading efficiency is concerned, we believe the real requirement is for transparent, interactive tools - not black-box AI - that give companies enhanced oversight and control of those factors.
If people understand what their machines are doing, and have the ability to tailor and configure them, they can not only better understand their supply chain and inventory, but act at a speed and scale they couldn't achieve alone. In our case, this can be demonstrated through two innovations:
1. Depth of data: From granular data to seat ID, SSP ID and bundle ID, the volume of data available is vast. But the kind of functionality that really empowers partners is the ability to sample their own bid requests. Rather than relying on a spreadsheet, they can see first-hand what their supply chain looks like, survey the bids they are receiving - and, in turn - use that information to create further efficiencies.
2. Custom optimisation: More radically still, our Real Intelligence engine, currently in beta, allows ad networks to set optimisation rules that automate previously manual actions. Humans have always been able to change the domains they’re sending to a buyer, block traffic that isn't working well, or increase the volume of traffic they’re delivering to a particular advertiser - all important efficiency and revenue drivers on the buy side. But the optimisation rules engine allows them to set up their data and their KPIs, then automate that process - not only saving time, but essentially allowing platform partners to build their own bespoke algorithms on top of their RTB exchanges.
In both cases, that level of interactivity gives ad networks the ability to bring their USPs, their values - the qualities that make them special - to an AI-assisted platform, combining enhanced speed and scale with the real intelligence that already powers their business.In doing so, it allows them to rediscover what AI sometimes encourages them to lose: the ownership, the power to set limits and tolerances.
Efficiency, technology and character all work together
Technology, we all know, is critical, but so is human intelligence and the distinctive character of any given ad network. If we keep the focus on combining these elements, we will allow creative ad networks to keep doing smart things in the future. And a strategy that leverages each element to maximise trading efficiency is the way to get there.
Andrew Macdonald & Dan Nelson
Andrew Macdonald - Andrew is Director of Client Success at Limelight, working with the company’s clients to help them get the most out of the Limelight platform. His role involves reactive support through calls, meetings, instant messaging and email, as well as proactive help to identify new opportunities. Client support is an essential component of the Limelight platform in order to enable the company’s clients to leverage all its many capabilities.
Andrew works alongside his counterpart, Dan Nelson, supported by a team of five. Prior to joining Limelight, Andrew spent two years as Head of Operations at a digital marketing agency. Ads360, and prior to that, he was an Account Manager at mobile advertising firm, Zapp 360.
When not working, Andrew enjoys spending time with his son, and playing strategy and fantasy role-playing games such as Risk and Dungeons & Dragons.
Dan Nelson - Dan is Director of Client Success at Limelight. His role involves supporting clients to maximise the value of the Limelight platform through email, phone, instant messaging and face-to-face consultation, both to troubleshoot issues and explore new opportunities. Dan was the company’s first Client Success director, and today works alongside his counterpart Andrew Macdonald, the pair supported by a team of five. Client support is an essential component of the Limelight platform in order to enable the company’s clients to leverage all its many capabilities.
Daniel came to Limelight after four years working in communications and support for two different Members of Parliament. Prior to that, he had stints at Agenda21 Digital and Rocket Fuel. He has also been elected as a Councillor for Southend-on-Sea and was a Cabinet Member responsible for economic growth, community safety and health. He was also the Deputy Leader of his local Political Council Group.
Outside of work, Dan is trained in coding using Python, speaks Spanish, and enjoys long Muay Thai and padel tennis.