Connected TV Advertising: What CRM Leaders Need to Know
There are few things that represent a powerful shared experience as impactfully as iconic television commercials. Back in the era of three networks, consumers were served the same commercial messages to both entertain and persuade. Classics like Wendy's "Where's the Beef?," the pink Energizer Bunny and, most recently, Geico's gecko are mainstays of the American psyche.
Today, there are literally hundreds of TV viewing choices, from network and local channels, subscription channels, and streaming, and despite a very digital and disconnected world where entertainment and programming is generally enjoyed asynchronously, television advertising still represents a significant percentage of overall ad spend.
Coherent Marketing Insights estimates current TV ad spend at about $260 billion and expects it to expand to $377 billion by 2033. But the nature of television commercials has changed significantly as marketers seek to gather better metrics to help them measure the value of the big budgets they allocate.
Increasingly, they're turning to connected TV (CTV), which has risen in use from 15 percent in 2020 to an estimated 38 percent of total TV today. In 2024, CTV advertising spending in the United States reached almost $30 billion and is projected to reach $46.89 billion by 2028, surpassing traditional linear TV advertising.
Three forces converged to make this happen, according to Amit Sharan, senior vice president of marketing at Tatari, a TV ad platform provider. First, access improved dramatically. Programmatic platforms now let any marketer tap CTV inventory without signing exclusive contracts with networks; no more guaranteed TV placements at premium prices. "CTV can be bought through programmatic platforms a performance team already knows how to operate" Sharan explains, "and the measurement got good enough that television stopped being a faith-based line item.
Second, supply exploded. Faced with the plateauing of subscription growth, streaming services platforms like Netflix, Disney+, and Amazon Prime Video launched lower-priced, ad-supported tiers to drive new demand.
Third, creative barriers collapsed. "A brand shooting video for social every week already has raw material," Sharan notes, "and AI is now speeding up production times or allowing for different variations to be made in a fraction of the time. The six-figure production commitment isn't the price of entry anymore."
How CTV Is Different
Advertisers soon found, though, that while CTV might look like television, it isn’t. Treating it like regular TV will result in lack of clarity on ROI.
According to Tavares Beverly, president and CEO of Beverly Boy Productions, "marketing and CRM leaders need to stop treating CTV like a bigger version of linear TV. Someone watching connected TV has a remote in hand and their full attention on the screen, not half watching while scrolling their phone."
And unlike digital advertising, CTV doesn't offer a click. Viewers might watch a 30-second spot, then, hours or days later, they might search for the product on Google, visit a website, or purchase. That gap is where the measurement problem ives.
CTV does offer targeting precision that linear TV never had. "We can match household viewing data against a client's CRM lists and serve ads only to in-market buyers," explains Thomas Oldham, founder of WebMotion Media, a digital growth agency. That represents significant value to marketing leaders trying to get the most out of their advertising budgets.
Creative strategy also comes into play, says Ryan Stone, founder and creative director of Lambda Films. "Measurability should influence how the commercial is produced in the first place," he says. This means building measurement into creative development from the start so you can iterate quickly based on performance data, rather than treating production and measurement as separate functions.
"There's a danger in treating CTV too much like digital advertising," he says. "It's still appearing full-screen on a 65-inch television alongside professionally produced programs. Production value matters."
Why Companies Should Invest
CTV brings value in the form of reach, credibility, and demand creation. As traditional linear TV is reaching fewer viewers and internet-connected TV is on the upswing, CTV also offers credibility in a subtle way, Sharan says. "The placement itself is doing work that no other video buy does," he says. "A CTV ad runs full screen with the sound on, in the room where the household actually gathers, and across most CTV inventory it plays through rather than being dismissed."
Meg Prejzner, founder of Hackett Brand Consulting, sees other benefits in CTV advertising, calling it a hidden opportunity to drive real business results. "A brand can pull together an ad and have it placed directly in front of its target audience with hyperlocal specificity and track its impact almost in real time," she says. "The best marketing teams and brands track CTV impact from end to end, measuring not just advertising but real return on ad spend through conversion."
But measuring the impact and return on ad spend on CTV has its own unique challenges. Understanding how to measure CTV effectiveness is fundamentally different from measuring linear TV or digital advertising. The gap between ad exposure and conversion is where most confusion and opportunity lie.
Attributed vs. Incremental ROAS
When a CTV platform shows return on ad spend (ROAS), it's showing attributed ROAS—the conversions that happened after someone saw your ad. That's not the same, though, as incremental ROAS. The challenge lies in identifying the additional, or incremental, sales that the campaign might have created vs. those that would have happened anyway.
This distinction is especially important for retention teams, Sharan says. "Most CTV vendors will hand you a ROAS against a first purchase," he says. But "for a retention team, that's the least interesting number in the report. What you want to know is whether the customers CTV brought you are worth more at 12 months than the ones paid social brought you, and no media vendor's dashboard can tell you that."
Most vendors, Sharan adds, "give you exposure and conversion data at the household level, as raw events, exported into the environment where your [lifetime value] and churn numbers already live."
If a vendor is only showing numbers inside its own UI, he says, "you can measure their campaign but you can't measure your customers."
"A ROAS figure is only as good as the counterfactual behind it. If a vendor can't tell you what they believe would have happened without the ad, they've handed you a correlation with a dollar sign attached," Sharan notes.
Measurement: Opportunities and Methods
CTV offers advances in measurement over linear TV, but it's still not perfect. The problem is clean attribution, says John Surabian III, brand growth strategist at Clickable Impact, a marketing and venture capital firm. "CTV platforms report view-through conversions, usually on a long window. Someone watches 30 seconds of your spot on a Roku channel, buys three days later after a branded Google search, and the CTV platform claims that sale. So does Google. You end up with well over 100 percent of your revenue accounted for."
The real challenge is understanding what you're measuring. A TV ad shifts awareness and demand; it doesn't convert on the first impression. By purchase time, several touchpoints have intervened.
But there are some emerging methods that can help to isolate CTV;s true impact.
Geo Holdout Testing
Surabian advocates for geo holdout testing as the gold standard. "Pick matched designated marketing areas, run the spot in half of them, hold the other half dark, then read the total revenue difference over four to six weeks. It's slower and a lot less satisfying than a dashboard. It's also the only read that survives being questioned by a CFO," he says.
Between holdout tests, watch for proxy signals like branded search volume in exposed markets, direct traffic spikes, and inquiry volume. "CTV works by making people go looking for you later," Surabian says. "If the spot is landing, that shows up as demand entering through a different door, not as a click on the ad."
Incrementality Testing
Webmotion Media's Oldham advocates for incrementality testing. The concept is pretty basic: Use a control group that doesn't see the CTV ads and compare to a group that does. "That tells you the true lift, not just last-click attribution," says Oldham, who tracks view-through conversions over a seven-day window for an automotive client. While CTV rarely closes a sale on first impression, he says, "it plants the seed. The search ad or direct visit comes later."
Oldham cautions that "marketers who judge CTV on last-click only will underreport its value by 40 percent or more."
Media Mix Modeling (MMM)
According to Jason Marvin, performance media director at Twelve Media, a digital marketing company, "At the upper end of the spectrum is media mix modeling). This uses historical data and analysis to determine channel-specific effectiveness. This method can help identify optimum spend levels, media mix, opportunities, and waste"
MMM works best with substantial data, like more than three years of campaign history, and larger spenders, but it can tell you the optimal allocation of budget across all channels and identify waste.
Data Clean Rooms and First-Party Connections
Among the challenges and risks related to CTV is that, traditionally, advertisers would send their list of customers to the CTV platform vendor who would then match the people who saw their ads with people who purchased. But that raises potential privacy and security risks.
Aaron Whittaker, vice president of demand generation and marketing at Thrive Internet Marketing, notes that "data clean rooms and server-side conversion APIs are newer options for connecting first-party CRM records with media exposure while reducing raw-data sharing." These tools, he explains, let users maintain the data in their own systems rather than in vendor dashboards.
Instead of sending data to the vendor, the vendor sends anonymized data like "Household ID 12345 saw your ad." Advertisers' internal systems then do the matching in house, ensuring that customer data doesn't leave their control.
MaryRose Daily, vice president of performance media at Props, providers of a performance-based creator marketing platform, stresses the importance of gaining a broad view of what's happening rather than relying on a single source. "CTV creates demand that is captured most often on other screens, causing its contribution to be undervalued," she says. "A strong framework triangulates near-term attribution, controlled lift testing, longer-term modeling, and measurable creative against real outcomes."
Daily notes that measuring CTV spans audience verification, brand lift, attribution, incrementality, and marketing-mix analysis, requiring marketers to "combine these sources to tell the most accurate story about performance, focusing on data compatibility and coverage rather than a single attribution tool or ROAS calculation."
Surabian agrees: "The emerging category worth attention is incrementality platforms rather than attribution platforms. Anything that assigns fractional credit to touchpoints is asking a question CTV can't answer honestly."
There's another important consideration that many marketers might miss. How does the provider handle households with multiple viewers? One TV impression does not necessarily represent one person. Shared screens inflate reach numbers if not handled carefully. According to Surabian, "most CTV measurement connects conversions to exposures by IP address, which works in a single-family home and badly in an apartment building or an office. In one client analysis, 28 percent of impressions were delivered across communal IPs, and those same addresses produced 57 percent of the one-day view-through responses."
Whittaker suggests comparing the tool's numbers to what actually appears in your CRM. "I would verify whether attributed conversions match actual customer records," he says. "I would also check whether exposed markets produced more revenue than the controls. If both tests support the reported result, I have much more confidence in the ROAS."
The Bottom Line for CRM Leaders
CTV is growing faster than any other advertising channel. Audiences that once seemed unreachable, like cord-cutters and streaming-first viewers, are now accessible at scale.
But measuring effectively requires rethinking how you evaluate media. Morissa Schwartz, founder of GenZ Publishing, advises, "for credible ROAS, define the conversion and attribution window before launch, deduplicate reach across screens, inspect frequency, and compare exposed audiences with a holdout or geo test when possible. Without incrementality, ROAS can reward ads for reaching people who were already going to buy."
Experts also encourage companies looking to get into CTV advertising to start small. Pick one audience, one geographic market, one clear business outcome, and establish your measurement method before you scale. If the test shows incremental lift verified against your own business data, you have a much stronger argument for increasing budget than simply reporting that millions of people saw your ads.
CTV has fundamentally changed television advertising from a faith-based line item to a performance channel. The tools to prove that performance exist. They're just not where most marketers are looking right now.
Linda Pophal is a freelance business journalist and content marketer who writes for various business and trade publications. Pophal does content marketing for Fortune 500 companies, small businesses, and individuals on a wide range of subjects, from human resource management and employee relations to marketing, technology, healthcare industry trends, and more.