How can intelligent QPS Optimization multiply your revenue from DSP partners?

Sidharth Sathish
5 min readOct 3, 2022


Did you know that Facebook’s website can respond to over 10 million Queries Per Second (QPS)? The response time is mere milliseconds. It’s an astonishing benchmark if you know the Web Advertising nuances well.

Digital Marketers always tend to keep a close eye on this online metric. A high QPS measure is a solid green flag for publishers looking to run campaigns via ad exchanges. It is an indication of the traffic handling capability of a server.

By now, you may be wondering what exactly is the meaning of QPS. Let’s get that taken care of.

What is QPS?

In basic network engineering terms, QPS is nothing but the number of information requests a server receives per second. Online advertisers can quickly assess how the support infrastructure handles varying website traffic and dynamically sense the efficacy of campaigns using this parameter.

Optimizing the QPS is, therefore, vital for exponential gains in server scalability and workload handling. It is this technique that has transformed platforms like Facebook and Twitter into the online giants as we know them today. Only through a sophisticated process can you show the right Ad to the right person at the right website.

There are suppliers of Ads and others who satisfy this demand in real-time. How does this supply and demand work in the world of Online Ads? What is the role played by QPS in this regard?

Role of QPS in managing Online Ads

Assume someone who has entered a smartphone website to see the reviews of the latest iPhone. As soon as she enters the page, the strategic banner positions populate with suitable Ads based on her online behavior. How does this happen?

Yes, the publisher must leverage an industry-grade Ad stack (ranging from Ad serving platform scripts, AdManager loads to out-stream videos). Each component gathers data and begins the race to win the banner positions.

The publisher transmits the Ad parameters — size, position, website, and user id to the SSP (Supply Side Platform) as a Query. Multiple SSPs receive similar queries through Amazon TAM or Google Open Bidding tools.

The SSP can send this to a DSP (Demand Side Platform) or to other SSPs in the transmission route. The DSP qualifies and transmits the bids to the source SSPs, which Google Ads Manager collects. Finally, we have a winning query, and the corresponding Ad gets displayed to the user via an iFrame code sent on the particular website. The publisher receives remuneration for the Advertising.

The entire cycle of events is repeated for each SSP and completed in seconds. You may now be thinking, where do QPS and its optimization come into play here?

Real-time bidding technologies have seen a meteoric rise, especially since the pandemic. With businesses going online worldwide, the traffic handling between the SSP and DSP has become a bottleneck, big time!

SSP players now need more infrastructure to accommodate the multitude of requests coming in from publishers. This surge of traffic requires them to invest more in Cloud power and increased expenses.

Next, coming to the receiving end. DSPs can get bombarded with requests from the Supply-side, leading to response lags and causing a decrease in revenue. This issue ultimately affects the partnerships they maintain with the Ad publishers.

To make things even more challenging, some partnerships stipulate limited SSP queries to avoid network overload. For instance, 5,000 queries per second are all they can transmit to the DSP. If the transmission system follows a FIFO (First In First Out) mechanism, then only the initial 5,000 queries reach the destination. The subsequent ones are ignored, making them lose out on any earning potential.

Now you know why optimizing the queries per second is pivotal for dynamic Ad campaigns and their revenue generation capability.

We now present Assertive Yield’s solution that enables SSPs to scale unprecedented revenue via Machine Learning powered QPS Optimization — “Traffic Shaping.”

Unclutter the Demand-Supply bottlenecks and deliver more quality and quantity of online Ads like never before!

AY Traffic Shaping — Revolutionizing the world of Online Advertising!

Gone are the days when mere QPS sampling and ML-based algorithms were enough to handle Ad publishing on the web. At AY, we go the extra mile by leveraging our deep expertise in the niche field of User Profiling. This revolutionary software solution from AY is a future-ready tool that senses user needs in real time for efficient demand handling.

Why opt for AY’s Traffic Shaping?

We combine the dual benefits of sophisticated Machine Learning algorithms and User Profiling to identify the most suitable traffic for each DSP proactively. How does this help?

  • Up to 40% cost savings due to optimized QPS and smoother infrastructure utilization
  • 6–12% increase in fill rates due to better quality of traffic
  • Increased bid rates
  • Solidified relationships with your DSP partners
  • Eliminate up to 60% of non-responsive bid requests
  • Resource-efficient User Profiling using AY’s Artificial Intelligence component
  • Trims your Carbon Footprint via lean server infrastructure usage by over 60%

You might be wondering about how we make such stellar benchmarks a reality. We will glance into Traffic Shaping’s modus operandi in the upcoming section.

AY Traffic Shaping’s secret sauce for optimizing QPS

Based on our years of observation in the Ad request domain, we’ve seen huge volumes of unwanted queries spamming both the SSPs and DSPs on a daily basis. This ultimately increases costs and brings down publishing efficiency and revenue.

How do we at AY avoid this hurdle? This answer is Predictive Analytics in real-time.

We gather data on time taken by our partners to respond to different categories of queries. Our ML algorithm considers the previous 14 days of SSP data and predicts the probabilities of achieving a response from the DSP.

You may now ponder the challenges associated with integrating a third-party tool like AY Traffic Shaping with an SSP. However, that is not the case.

AY Traffic Shaping’s Secure and Straightforward Integration

Integrate the Traffic Shaping solution to any SSP with zero additional effort as it leverages the same HTTP protocols in the existing network endpoints.

And as far as the data security of sensitive information is concerned, we hold none. Since the requests are hashed, the only information Traffic Shaping holds is the predictive bid rates.

Optimize QPS with Traffic Shaping and witness the transformation

Assertive Yield has been the go-to partner for publishers across industries over the years. Our unmatched brand reputation has been made possible due to our relentless efforts to help them manage their demands better, leading to increased revenues.

The probability of a successful bid reaching the publishers even before the request reaches the end server. The integration of cutting-edge predictive analytics using AI and ML technologies has provided our clients with a bundle of benefits that include:

  • Better bid rates and win rates
  • Optimized QPS via improving the quality of queries reaching the DSPs
  • Optimal usage of server infrastructure
  • Remove unwanted overhead expenses

With the dynamic supply-demand of QPS optimized using sophisticated and proven Machine Learning technology, achieve maximum ROI for your Ads with AY.

Contact Assertive Yield today for more information on our revolutionary Traffic Shaping solution.