Transparency considerations and Supply Path Optimization (SPO) are driving ad tech consolidation and a broader industry shakeout. These pressures are driving advertiser spend to slowly move towards a reduced set of trusted SSPs.

With a focus concentrated heavily on matters of transparency, recent years have seen a series of IAB Tech Lab initiatives including Sellers.json and the Supply Chain Object that try to address these concerns. These initiatives are having a positive impact, and further highlight the need to find the most transparent and optimal route to high-quality supply.

However, what is often overlooked, are the connections between SSPs and DSPs. This base-level connection is an essential point in the process and pivotal in understanding how efficient bidders are configured to secure impression opportunities for buyers at scale. Over the years, these connections have become increasingly complex and convoluted. In order for advertisers to employ effective SPO practices, they need a solid foundation without inefficiencies and unnecessary waste.

An optimized supply path benefits both the advertiser and their ad tech platform as reduced infrastructure and internet costs free up essential resources for product development and innovation. Added need for efficiency and more effective resource allocation makes adopting a strategy built on rock-solid and proven foundations essential.

To help advertisers and technology partners, we have compiled a set of Programmatic Best Practices that we recommend for improving bid and win rates while operating a healthy revenue per QPS (Queries Per Second) for each connection. Adopting this input and these best practices will help drive success for buyers and their advertisers. We believe these are high impact actions that tech vendors should take into consideration and prioritize. As a marketer or advertiser, this list can be used as a way to drive an empowered discussion with your ad tech vendors to gain an understanding of their foundation for your SPO strategy.

Below we will take a look at ways to optimize win and bid rates (success KPIs for bidders) and how to reduce the overall bandwidth and computational weight required for platforms to drive efficiencies at scale.

Publishers’ blocks and non-transparent advertiser domains are a common reason for bids to be rejected by SSPs which leads to waste and reduced efficiency. In order to reduce the publisher block-rates and simultaneously increase win-rates, it’s important for DSPs to always respect publishers’ block settings and pass transparent advertiser domains.

Publishers can configure blocks to filter for specific IAB content categories, advertisers or creative attributes. One way to help mitigate publisher blocks is through multi-bidding. This is a tactic where instead of a single bid response, a DSP submits multiple bids for a single bid request resulting in more opportunities to win the desired placement and improved efficiency.

To condense traffic flow and optimize the connection between DSP and SSP the following best practices should be considered:

• Utilize bid request/response compression to reduce bandwidth and internet costs
• Weed out unauthorized sellers and fraud by using Ads.txt and App-ads.txt filtration (as explained here, and here). Leverage Sellers.json and Supply Chain Object for increased transparency and enhanced SPO mastery
• SSPs should refrain from purposefully duplicating bid requests. For example, by sending multiple requests with different floor prices to the same DSP
• SSPs should pass floor prices with each request. In a 1^st price auction world, not sending floor pricing is no longer justified and leads to a lot of wastage in terms of bandwidth and compute time
• Support multiple impression objects in a single request. This means instead of one request per placement, a single request should encompass 1:n impression opportunities
• Support multiple types of traffic objects. For example, one request might contain display, native and video opportunities when available instead of issuing a different request for each traffic type
• Support multiple-ad sizes or formats in a single array. Instead of one request for a single ad size, pass multiple sizes as an array.

These optimizations are key to maintaining a healthy revenue per QPS. An unhealthy revenue per QPS is when QPS received and DSP revenue on media transacted through a specific SSP connection does not cover the listening and datacenter costs. These steps help identify which connections can be further maximized, are operating at peak efficiency, or are unsustainable and should be discontinued.

Smart Traffic Shaping solutions optimize the traffic volume being passed by the SSP or processed by the DSP based on historical campaign success data. For example, traffic from placements or countries that have not seen any campaign spend in the last 24 hours can be suppressed to prioritize more valuable impression opportunities.

Traffic Shaping solutions have been deployed by DSPs and some SSPs in order to manage the increased traffic load and duplicated requests resulting from header bidding and added intermediary layers.  Shaping allows technology vendors to continue listening to the maximum number of available and relevant impression opportunities without adding massive investments in new hardware and a related ballooning of infrastructure costs.

Ultimately, traffic duplication makes it difficult to judge the optimal supply path and leads to other concerns including buyers bidding against themselves (please read our thought leadership) on request duplication for more information). With the recent release of Sellers.json and the Supply Chain Object, we will be able to provide even more transparency into the number of “hops” or intermediaries present in the supply path, further reinforcing trust and empowering buyers at the same time.

In addition to algorithmic Traffic Shaping solutions, DSPs should start by making sure that the pre-targeting settings on the SSP side are properly configured to pass the most relevant requests for given advertisers demand. This means suppressing traffic from low demand regions, countries, formats, ad sizes and cutting  (re)sellers that do not add value. Setting these configurations will help ensure that the buyer is not missing out on valuable impression opportunities in key markets due to QPS caps or related blockers (if any). Pre-targeting settings are still mostly manual solutions that are only supported by larger SSPs. This is a potential area of improvement by the sell-side in order to keep DSP listening costs in check and we expect it will become increasingly common across all SSPs in the next 24 months.

The above, whilst not exhaustive, are relatively straight forward programmatic best practices, but when setup effectively these recommendations can ensure that buyers are using an efficient and cost-effective buying platform that is providing quality, reducing overhead and extracting maximum value for buyers. This setup also readies DSPs to adapt to further increases in QPS due to advances in header bidding for video and apps (‘unified auctions’) without needing to absorb or pass on potentially crippling increased infrastructure costs. Applying this guidance also further empowers advertisers and positions them to effectively employ SPO practices, with added insight and transparency into the path to an optimal supply path.

Throughout 2019, Adform built highly optimized connections with our trusted SSP partners, to set our clients up for programmatic success in 2020. To learn more, please feel free to get in touch.