Home Content Studio Getting It Right: Leveraging Data Science To Hit CTV Budgets

Getting It Right: Leveraging Data Science To Hit CTV Budgets

SHARE:

This article is sponsored by Xandr.

We have all encountered the problem: What’s promised is not always what’s delivered. 

In the past decade, advertisers have seen a seismic shift in how consumers receive media. With more on-demand services added every year, the future of entertainment is becoming evident. This change comes with many challenges, but it also inspires innovation.

One area ripe for creative solutions is how connected TV (CTV) advertisements are purchased, specifically, how CTV budgets are paced. Traditional budget pacing systems take a buyer’s budget and distribute that spend evenly throughout a day so that a constant stream of consumers is exposed to a brand’s messaging. With CTV budgets, however, spending within a day can go awry because of the unique challenges CTV impressions pose for demand-side platforms (DSPs). Some challenges mirror those encountered when building capabilities to buy display and others are entirely new, representing a complete paradigm shift. 

  • First, the timing of CTV advertisements will vary. For CTV line items, an impression may be won at the beginning of a show and play almost instantly after the auction takes place, while for other auctions, the ad may be served much later in the ad pod. Think about an auction taking place at the beginning of a 45-minute television episode streamed on Hulu. The ad could be served instantly during the commercials before the episode, or the ad may be served nearly 20 minutes later during a commercial break midway through the episode. The greater length of time between an auction and served impression as well as the inconsistency calls for a budget pacing system that can accommodate this unknown and volatile lag.
  • Second, CTV buyers still want to see DSPs algorithmically find cost savings for them. Buyers expect to save media cost dollars through bid shading performed by their DSP. Buyers are accustomed to this spending behavior since this capability has long been available for more traditional formats such as display. As a result, buyers also expect this cost-saving mechanism for new and emerging formats like CTV. 
  • Third, CTV is largely bought through deals. This is typical buying behavior for most emerging formats. As a result, the system needs to be built in a way that controls  a rapid influx of available impressions when line items bid at, or above, set deal floors. Without such controls, buyers may inadvertently buy a vast number of impressions quickly, which can lead to overspend. 

Given the complexities of these challenges, technology platforms must build a system that goes beyond addressing each hurdle individually and instead provides a holistic, interoperable solution to control budget spending. To overcome the constraints of what represents a classic optimization problem, solutions rooted in data science are needed. 

Xandr recognized the potential risks that buyers face from CTV budget pacing and committed to building a holistic solution. As a first step, the team modified the functions responsible for distributing ad spend. Realizing a more advanced solution was needed, the team – comprised of engineers, data scientists and product experts – went back to the drawing board and took an out-of-the-box approach. Rooted in science and experimentation, Xandr developed hypotheses, simulated models, tested in live environments and eventually engineered a second layer of control alongside its existing shading algorithm. This layer, which adjusts bid rates based on signals that predict overspend, has already proven to reduce CTV overspend by up to 70% since its implementation.

In this new digital video-driven media landscape, buyers should feel confident that their CTV budgets are allocated properly. Ensuring that a promised media spend is delivered is paramount to the team at Xandr and should be for all technology platforms. Focusing efforts on innovation to solve for today’s challenges is clearly important but building flexible solutions that can easily anticipate and meet customer needs of the future will make all the difference for buyers in a fast-moving and dynamic industry.  

Must Read

Comic: Shopper Marketing Data

CPG Data Seller SPINS Moves Into Media With MikMak Acquisition

On Wednesday, retail and CPG data company SPINS added a new piece with its acquisition of MikMak, a click-to-buy ad tech and analytics startup that helps optimize their commerce media.

How Valvoline Shifted Marketing Gears When It Became A Pure-Play Retail Brand

Believe it or not, car oil change service company Valvoline is in the midst of a fascinating retail marketing transformation.

AdExchanger's Big Story podcast with journalistic insights on advertising, marketing and ad tech

The Big Story: Live From CES 2026

Agents, streamers and robots, oh my! Live from the C-Space campus at the Aria Casino in Las Vegas, our team breaks down the most interesting ad tech trends we saw at CES this year.

Privacy! Commerce! Connected TV! Read all about it. Subscribe to AdExchanger Newsletters
Monopoly Man looks on at the DOJ vs. Google ad tech antitrust trial (comic).

2025: The Year Google Lost In Court And Won Anyway

From afar, it looks like Google had a rough year in antitrust court. But zoom in a bit and it becomes clear that the past year went about as well as Google could have hoped for.

Why 2025 Marked The End Of The Data Clean Room Era

A few years ago, “data clean rooms” were all the ad tech trades could talk about. Fast-forward to 2026, and maybe advertisers don’t need to know what a data clean room is after all.

The AI Search Reckoning Is Dismantling Open Web Traffic – And Publishers May Never Recover

Publishers have been losing 20%, 30% and in some cases even as much as 90% of their traffic and revenue over the past year due to the rise of zero-click AI search.