Advertisers trying to understand how the current programmatic workflow affects their marketing budgets are faced with a daunting task. With a vast array of ad technology services, platforms and cost models on offer, it is becoming increasingly difficult for advertisers to cut through the noise. One of the areas garnering much attention is the supply chain. While third-party tools have created a level of control for advertisers on the buy side, most sell-side technology was not built to transact in real-time bidding (RTB) auctions.
This means it’s not always clear what takes place between when a bid is submitted and a clearing price is returned. It can also cause uncertainty for publishers over how to price inventory and who is buying it.
To overcome these operational complexities, it is important to understand how programmatic RTB auctions work, if there are any hidden supply side fees and where the ads are appearing.
Historically, RTB transactions have been conducted using a second-price auction. This means that the winning bidder doesn’t actually pay what they bid, but pays US$0.01 above the second-highest bid.
When programmatic began, publishers used a ‘waterfall’ set up with multiple SSPs (Supply Side Platforms), stacked based on priority. If the first SSP did not sell the available inventory, it would be offered on the second, and so on. Over the past two years, header bidding – where publishers make their inventory available to multiple exchanges at the same time instead of sequentially – has fast become the new normal. While header bidding can create higher CPMs for publishers, it forces SSPs to compete with one another on price in a way that they had not in the past. SSPs raise the closing price of their auctions to increase their probability of winning, and adjust their fees or implement variable price floors, which are not always disclosed to the buy side platform. Buyers are unable to detect these floors and often must bid higher to win the same amount of impressions.
The use of these tactics has led many to argue that there is no longer such a thing as a fair, or true, second-price auction. While fees are an inevitable and important part of the programmatic ecosystem, buyers and sellers need to have better transparency into when and how the fees are applied.
SSPs and exchanges have started adopting a first-price auction model, where the winning bidder pays exactly the price they bid. Ideally this delivers more working media to the publisher, squeezing margins of the ad tech players in the middle, but it comes with its own set of challenges. The rollout has been inconsistent and confusing, with different SSPs approaching it in different ways. Some have gone 100% first price, while some have experimented with only part of their inventory. Some have switched over to first price without alerting the buy side at all.
Major DSPs have addressed the change in auction dynamics in various ways. Some are focused on adjusting their algorithms to automatically account for market norms and prevent overpayment. Others are bidding specifically on first-price or second-price inventory. DSPs are also starting to provide information about buy-side fees, either in their platforms or in custom reports to provide more clarity into where the working media dollar goes.
To maximise the value of their programmatic buys, brands need to be aware of how the transactions are conducted. Varying the bid price and comparing it to price paid can give you information that allows you to improve bid efficiency.
Consider moving more transactions to private marketplaces, which offer better results, better transparency and less risk
Auction Dynamics Best Practices
Monitor CPMs by exchange to identify and address any major pricing variations
When monitoring eCPMs and performance by exchange, compare the bid and the price paid to identify if a first-price auction or dynamic floors are affecting pricing. This will help create specific bid strategies by tactic and channel to avoid ever-creeping CPMs.
Campaigns that easily hit spend targets with ‘Pace Evenly’ should consider lowering their bid prices a small amount each day until they begin to underspend, then they should raise the bid price accordingly to spend the budget. This will inherently improve CPMs and mitigate problems imposed by the fragmented auction dynamics system.
Private marketplaces provide a structured environment with lowered risk
Private marketplaces, especially fixed-price deals, help with rate savings. They also offer priority positioning, viewability goals and brand-safety guarantees that are easier and cleaner to manage. Grouping publisher deals by auction type will also provide a structured, transparent environment with lowered risk.
Ask DSPs what products are available in their self-serve user-interfaces
This will help identify, target, and optimise against inventory running on first-price models. Ensure there are no pass-through costs from SSPs/exchanges without disclosure, which will shine a light on previously undisclosed fees.
Push for universal standards that provide a more transparent marketplace
Participate in industry certification such as TAG in the US or DTSG in the UK, and evaluate new technologies such as blockchain to broaden your view of the potential future landscape.