Bidstream bloat is a widely acknowledged challenge in programmatic advertising, and one that is moving up the priority list as the industry increasingly looks to reduce unnecessary complexity and cost in the ecosystem, as well as address pressing sustainability issues. Here we unpack exactly what is meant by the term, how the situation arose, and the steps that the buy and sell side can take to mitigate it.
Cause…
At its core, the bloated bidstream problem is an issue of supply and demand: it occurs when the number of bid requests sent to demand side platforms (DSPs) significantly exceeds the amount of inventory available.
Digging a bit deeper to determine the roots causes of this over supply, the picture is - unsurprisingly - a bit more complex.
Bidstream traffic has seen a huge increase over the past few years, with a number of contributing factors. The rise of header bidding, in allowing publishers to offer inventory to multiple supply side platforms (SSPs) at the same time (to increase competition, and with that, yield), means the same impression can be sent to a DSP multiple times. The result? A huge duplication of bid requests. This is exacerbated by the plethora of intermediaries and resellers in the supply chain, which introduces yet more pathways for a single impression to reach a buyer.
The problem is further compounded by indiscriminate bid broadcasting, which sees SSPs send all available bid requests to all the DSPs they are connected to - regardless of whether these buyers are likely to bid. This traffic can be irrelevant or duplicative, lowering the potential value to DSPs, and has the consequence of congesting the bidstream.
In short, a single impression opportunity can trigger multiple bid requests, with buyers often having no chance of participating in the auction. The result is bidstream bloat.
… and effect
Bidstream bloat is a sign of the programmatic system not working as it should and could. As well as being a technical headache, it has financial, operational and environmental consequences.
Looking first at the bottom line, it’s not hard to see that processing the trillions of bid requests resulting from their oversupply puts a significant strain on media trading infrastructures. DSPs are pushed into investing in more robust servers and cloud capacity to handle the volumes; SSPs meanwhile face rising distribution spend on traffic that often doesn’t deliver a return.
The increased costs from the infrastructure overload are a drain on margins and resources, while expenditure incurred from storing, transmitting and processing irrelevant or redundant bid requests is wasted spend.
Any overcrowded environment risks the things that are of real value being missed, and the programmatic bidstream is no exception. The huge volume of traffic forces DSPs to filter out vast numbers of irrelevant or duplicated requests - and with those some high-quality signals that have got lost in the noise. BidSwitch data illustrates this degradation of signal to noise - the auction win rate has decreased 24%, indicating untapped revenue opportunities.
Too much traffic also slows down the delivery of ads; this has a negative impact on campaign performance (timeouts and missed impressions for example) and degrades the user experience.
And in processing more data (much of it unnecessary), more energy is consumed. As the industry becomes increasingly serious about its sustainability responsibilities, this inefficiency and the resulting higher carbon emissions becomes more and more of a concern.
Beating the bloat
So far, it would seem, so bad.
But bidstream bloat is not inevitable; there are already some proven tactics to make the bidstream smarter.
In practice, reducing bidstream bloat comes down to a small number of interlocking strategies:
-
Traffic shaping and intelligent filtering
-
Smarter routing based on real demand signals
-
Supply Path Optimization to reduce duplication
-
Industry-wide collaboration and standards
BidSwitch sits in the middle of the programmatic ecosystem, connecting SSPs and DSPs. As such it has a vantage point that enables it to see the big picture when it comes to industry challenges and, with that insight, develop tools and practices to address them.
Traffic shaping enables platforms to filter and prioritize bid requests so that, rather than sending everything, only the most relevant go to DSPs. SSPs leverage BidSwitch’s bidstream optimization tools, including SmartSwitch, to analyze historical bidding patterns and real-time signals to predict which inventory buyers are most likely to bid on. Coupled with SSP-own traffic shaping algorithms, these tools reduce the QPS (queries per second) load, lowers processing costs, and sees DSPs receive higher quality bidding opportunities.
Advanced systems like SmartSwitch can segment traffic by device, geography, content type and more, using machine learning to refine this targeting over time. This intelligent routing, which allows SSPs to match supply with real demand rather than falling back on ‘spray and pray’ tactics, increases the likelihood of a bid and reduces waste on both sides.
Removing unnecessary bid requests reduces traffic processing costs by up to 60%. QPS efficiency is improved because DSPs only receive traffic they are likely to bid on. And the reduction in irrelevant bids cuts down on noise, making it easier for DSPs to listen, enter the auctions that further their purpose - and therefore win more auctions. On the sell side, SSPs connected to BidSwitch and leveraging SmartSwtich, and therefore sending more high quality traffic, are more attractive partners.
Supply Path Optimization (SPO) has been a core topic for almost a decade. Now supported by industry standards such as ads.txt, sellers.json and Global Placement ID (GPID), it reduces the number of intermediaries and duplicate auction paths in favour of transparent direct routes and can minimise bloat. It also helps to increase financial returns - fewer hops mean lower fees, less latency and more accurate data for decision-making.
Solving an industry problem such as bidstream bloat is a collective responsibility that calls for collaboration. This requires SSPs, DSPs and technology intermediaries to work together to undertake initiatives such as setting auction limits and QPS caps, implementing protocols to deduplicate auctions, introducing bidstream audit frameworks, and setting best practices for traffic management.
The road ahead
Bidstream bloat is a real problem.
But it also has robust solutions: auditing the bidstream identifies unnecessary volume and duplicates; working with the right partners enables traffic shaping and filtering; and using tools such as SmartSwitch, along with adopting SPO practices, improves efficiency.