Dynamic flooring: What is it and how do you master floor pricing as a publisher?

Usually, this process can be done in several different ways. You have algorithms that are pushing floors to prebid, you might be using GAM’s out-of-the-box dynamic flooring.  Another option is using a combination of techniques to trigger specific Unified Pricing Rules (UPRs) within GAM dynamically.  The question would be: in a  fluctuating landscape, how can you determine prices accurately?    

Dynamic floor pricing has faced skepticism among publishers, given the variations in performance among different providers. Some solutions have fallen short of optimizing revenue, merely redistributing it among demand partners. 

This approach often favors specific partners to the detriment of others, failing to drive incremental revenue for publishers.

It often leaves publishers with a legacy approach of manually monitoring a lot of different segments of traffic and managing floors manually for some of them based on assumptions and estimates. 

While this process can show better results than the built-in GAM flooring (or not having any floors at all), it’s also prone to inconsistent results and oversight, and it often hurts revenue if it’s not updated frequently enough. It also costs a lot of internal resources to maintain something that can be more precisely optimized with machine learning.

What is dynamic flooring in programmatic?

Dynamic price flooring optimizes the ad request to propose the highest price for the publisher while ensuring it aligns with the most suitable price for the programmatic buyer. Therefore, the best dynamic flooring system represents an advanced shift in ad optimization. 

By integrating the power of artificial intelligence and machine learning, publishers can predict optimal inventory prices and adapt their floors accordingly.

"This is a relatively advanced method of implementation, which can handle quick changes within traffic and  CPM changes properly, so it can adapt quickly, especially to known events but also unknown events”, explains Nils Lind, CEO/Founder of Assertive Yield.

Unlike traditional methods, the system offers unparalleled flexibility for setting floors not only on GAM but also on Amazon and Prebid. This expanded scope empowers publishers to navigate a multi-platform landscape with ease, adapting to the unique demands of each partner —all in one place.

With each impression, a good dynamic flooring solution learns and improves its ability to create the right floor at the right time by running predictions in real-time for every impression (not just for every visit, nor updating the rules every hour),  but for running the prediction for every single bid request that is happening on the display side as well in real-time.

The system also provides different prices for creative sizes for the same placement and runs those predictions within the browser. This way, you don't have additional latency, or delay the auction because you have to wait for the prediction data from the server. It is already available within the browser's cache and it can execute it within nano to microseconds.

How does dynamic floor pricing work?

A good dynamic floor solution generally uses a three-tiered approach: Control, Exploration, and Optimized Groups. The system sets smart floors on the impression level, leveraging real-time data, user-specific information, and various dimensions that are not available on GAM such as User Segments, UTMs, Layout, and Session Depth to adjust to market values dynamically. 

Here is how it works in more detail:

Control group

This group stays untouched. No changes here—it sticks to the publisher's existing floor price strategy, with no tweaking involved. 

Exploration group

This is the testing ground, it tests over 100 different strategies (or less, depending on the publisher's size and limitations). Once it spots a winning strategy, most of the traffic is given to this strategy under the optimized group.

Optimized group

This behind-the-scenes group takes over, using the winning strategies. It tailors them based on different slices of incoming traffic. For example, when the system spots a better strategy for a specific country and operating system, it switches. 

But if flooring isn't doing the trick in some markets at certain times, it automatically switches to a strategy without floor prices or one with the minimal prices set by the publisher. Effortlessly keeping up with the market's twists and turns to protect the publisher's inventory from being undersold!

The benefits of dynamic flooring

At this point, you can already see that an advanced dynamic flooring system is quite beneficial to publishers, especially in today’s programmatic market economy, where publishers seek innovative strategies to unlock new opportunities and enhance their advertising revenue.

Embracing the power of AI-based solutions has become crucial in this pursuit, as it enables publishers to optimize their operations and secure a larger share of the advertising pie. Advanced platforms utilizing AI, revenue predictions, and machine learning algorithms offer sophisticated, data-driven floor pricing strategies.

As the industry enters a pivotal moment, publishers need to reevaluate costs, make intelligent investments, and strive for the highest possible returns. 

Other aspects to consider:

Impression-level smart floors with real-time user CPM 

A holistic dynamic flooring system operates on the impression level. Utilizing machine learning predictions directly within the browser, supported by edge computing Cloudflare, the system assesses user-specific information, including previous interactions, CPM history, visit depth, and various dimensions. 

This granular approach enables smart floors that dynamically adjust to the market value, ensuring each impression opportunity is optimized in real-time.

 Adjusts floor prices to current market value

One of the standout features of the system is its ability to generate and measure uplifts in real-time. The system continually learns and adapts to market value through its exploration, control, and optimized groups, unlike on manual floors, where somebody needs to adjust them around New Year’s when the shift in demand is a dramatic one.

Due to this feature, publishers can witness improvements in CPM, RPM  (revenue per thousand impressions), and overall revenue per visit and session, ensuring a dynamic and responsive approach to ad revenue optimization.

This automated process ensures that publishers can respond dynamically to market shifts without the need for constant manual intervention.

Unprecedented flexibility

An AI dynamic flooring system introduces a paradigm shift by offering unprecedented flexibility in setting floors not only on GAM but also on Amazon, Prebid, Open Bidding and Google AdExchange. This expanded scope empowers publishers with greater control and adaptability in maximizing revenue across diverse platforms.

Exclusive metrics and dimensions

A holistic dynamic flooring system extends beyond the confines of traditional methods by optimizing for Revenue per Visit. This comprehensive data point ensures true incremental uplift for the publisher.

Save time, cost, and resources

Because everything is automated with AI, publishers can save their development, maintenance, and revenue operation teams from spending time and money manually monitoring different segments of traffic, optimizing, calculating, and updating price floors for each of them (which is prone to mistakes). It's human after all. 

9 criteria to select the right dynamic flooring solution for publishers

A good dynamic flooring solution will give you a floor price strategy to implement. The best dynamic flooring solution will work closely to analyze your data points, test different floor pricing strategies, and build innovative solutions to make your price floors more efficient and effective for your goals.

"AY's Yield Optimizer stands out because of its Dynamic Flooring's integration with GAM and efficient communication of floor values in the bidstream. Our tests show no negative impact on page or ad load speeds, resulting in swift performance and significant assistance, leading to a remarkable 63%+ increase in programmatic advertising revenue” Jan van der Crabben, the CEO and Founder of World History Encyclopedia.

Embrace the future of AI dynamic flooring

The best dynamic flooring system stands out in the dynamic flooring arena, not just for its real-time pricing adjustments but also for its innovative use of AY's powerful machine learning, with more than 70 billion monthly impressions around the world. The solution also offers 99% hands-free integrations, allowing you and your team to maximize resources, save time, and streamline processes, all while your integrations run smoothly. 

This advanced approach allows for rapid adaptation to market changes, both predictable and unforeseen, ensuring that publishers can maximize their revenue.

The AY Yield Optimizer has seen success amongst pubs with its average results being 15%. This is a result of the solution's advanced algorithms (and other features) taking into account an extensive range of factors, making it a formidable tool for publishers to combat prevalent issues such as bid shading and optimize their ad stack for maximum efficiency and profitability.

Learn more on how to refine your floor pricing strategies for ad revenue success: Talk to us!