Performance TV Is Ready to Perform, but Measurement Needs to Hurry Up: Starcom’s Kristin Haarlow
CANNES, France – Connected TV has gained the audience, advertising inventory and buying flexibility needed to become a serious performance channel. Its measurement systems, however, are still searching for their boarding passes.
That was the assessment from Kristin Haarlow, chief investment officer of Starcom, during a panel discussion moderated by Mike Shields on behalf of Beet.TV at the Cannes Lions International Festival of Creativity.
Haarlow said the transformation began with viewers, who abandoned the industry’s carefully labelled boxes and simply followed the programs they wanted to watch.
“The audience does not care if they are watching a linear channel or if they are watching a streaming channel,” Haarlow said. “They’re just following the content.”
Advertising platforms eventually followed. Streaming services opened more inventory while programmatic technology gave buyers greater speed and flexibility. That combination has made connected TV, or CTV, more useful for campaigns tied to business results.
Previously, limited supply made it difficult to reach streaming audiences or adjust campaigns. Now advertisers can buy at greater scale without committing their budgets to the television equivalent of a long and slightly mysterious marriage.
Buying technology arrived before measurement
Haarlow said advertising technology has largely caught up on the purchasing side. Brands can buy programmatically, move faster and make adjustments that were difficult under traditional upfront commitments.
Measurement is another matter.
“The measurement is still very, very fragmented,” she said.
Advertisers face a crowd of measurement providers, each with its own approach. Clients often want one clear account of performance. Agencies, meanwhile, are still assembling what Haarlow called “the puzzle” of the analytics story.
That gap becomes especially noticeable when clients expect television to behave exactly like search or retail media. A chief financial officer may want the same cost per outcome from video that the company gets from search. Haarlow said that expectation does not match how the channels work.
“The cost per outcome is not going to be the same in CTV as it is in search, but that doesn’t mean it’s not relevant for performance,” she said.
CTV can now be evaluated against more specific results than impressions, reach or brand lift. Haarlow said advertisers can track metrics such as cost per store visit, cost per quote and even cost per purchase.
Still, comparing those figures across channels requires context. Television may contribute awareness and incremental demand that a tidy short-term metric does not fully capture. The spreadsheet may want one answer, but consumers continue to be inconveniently human.
Smaller brands get a ticket to television
The expansion of CTV inventory has also lowered the financial barrier for smaller businesses and direct-to-consumer brands.
“You don’t have to lock up hundreds of thousands of dollars or millions of dollars to get into the CTV space anymore,” Haarlow said.
Brands can begin with a test budget, learn from the results and adjust their spending. More supply has also pushed prices lower, a development Haarlow acknowledged with an apology to anyone in advertising sales.
The lower cost does not make CTV identical to search or social media. It does make experimentation more practical for brands that once viewed television as a club with an intimidating cover charge.
Haarlow said video retains an important advantage as more advertisers enter the market: scale. Audiences are already there and television can provide mass reach that brands would otherwise need to piece together through several channels.
The next challenge is connecting more data sources to that viewing activity so advertisers can understand how exposure contributes to business outcomes.
Chasing one metric could leave brands short
Budget pressure is encouraging companies to focus on a single key performance indicator. Haarlow warned that this can push advertisers to remove too much money from broad-reach channels and direct it toward tactics that produce cheaper immediate results.
That approach may look efficient in the short term. It can also ignore the “halo effect” created by upper-funnel advertising.
“My fear is that partners or clients will just cut too much from the more mass channels and go too much into the very specific performance-based channels,” Haarlow said.
Brands that stop investing at the top of the funnel may later see weaker returns in marketing mix modelling or other long-term assessments. Rebuilding lost awareness can take considerably longer than cutting the budget did. Competitors may also use the opening to claim market share, because they rarely observe a rival’s retreat and politely leave the space untouched.
Measurement needs to pick up the pace
Asked what improvement she most wanted to see, Haarlow gave a direct answer: “I think measurement’s gotta get faster.”
Campaign results can still take four, six, eight or even 12 weeks to arrive. That delay clashes with expectations that advertisers should optimize toward precise business outcomes in real time.
Haarlow said the problem is partly technological and partly organizational. Clients must become comfortable with newer measurement platforms and models instead of relying only on slower marketing mix modelling or return-on-investment studies. Technology providers must also process enough data quickly enough to support useful campaign changes.
Shields joked that CFOs and marketing mix models had emerged as the two villains of Cannes. Haarlow noted that the dynamic was “pretty much the same as 10 years ago,” only with a different topic.
CTV, in other words, has acquired abundant inventory, flexible buying and measurable outcomes. Now it just needs its reporting to arrive before everyone has forgotten what the campaign was selling.
You’re watching Beet Talks at Cannes Lions 2026, presented by Roku. For more videos from this series, please visit this page. You can find all of our coverage from Cannes Lions 2026 here.
CANNES, France – Connected TV has gained the audience, advertising inventory and buying flexibility needed to become a serious performance channel. Its measurement systems, however, are still searching for their boarding passes.
That was the assessment from Kristin Haarlow, chief investment officer of Starcom, during a panel discussion moderated by Mike Shields on behalf of Beet.TV at the Cannes Lions International Festival of Creativity.
Haarlow said the transformation began with viewers, who abandoned the industry’s carefully labelled boxes and simply followed the programs they wanted to watch.
“The audience does not care if they are watching a linear channel or if they are watching a streaming channel,” Haarlow said. “They’re just following the content.”
Advertising platforms eventually followed. Streaming services opened more inventory while programmatic technology gave buyers greater speed and flexibility. That combination has made connected TV, or CTV, more useful for campaigns tied to business results.
Previously, limited supply made it difficult to reach streaming audiences or adjust campaigns. Now advertisers can buy at greater scale without committing their budgets to the television equivalent of a long and slightly mysterious marriage.
Buying technology arrived before measurement
Haarlow said advertising technology has largely caught up on the purchasing side. Brands can buy programmatically, move faster and make adjustments that were difficult under traditional upfront commitments.
Measurement is another matter.
“The measurement is still very, very fragmented,” she said.
Advertisers face a crowd of measurement providers, each with its own approach. Clients often want one clear account of performance. Agencies, meanwhile, are still assembling what Haarlow called “the puzzle” of the analytics story.
That gap becomes especially noticeable when clients expect television to behave exactly like search or retail media. A chief financial officer may want the same cost per outcome from video that the company gets from search. Haarlow said that expectation does not match how the channels work.
“The cost per outcome is not going to be the same in CTV as it is in search, but that doesn’t mean it’s not relevant for performance,” she said.
CTV can now be evaluated against more specific results than impressions, reach or brand lift. Haarlow said advertisers can track metrics such as cost per store visit, cost per quote and even cost per purchase.
Still, comparing those figures across channels requires context. Television may contribute awareness and incremental demand that a tidy short-term metric does not fully capture. The spreadsheet may want one answer, but consumers continue to be inconveniently human.
Smaller brands get a ticket to television
The expansion of CTV inventory has also lowered the financial barrier for smaller businesses and direct-to-consumer brands.
“You don’t have to lock up hundreds of thousands of dollars or millions of dollars to get into the CTV space anymore,” Haarlow said.
Brands can begin with a test budget, learn from the results and adjust their spending. More supply has also pushed prices lower, a development Haarlow acknowledged with an apology to anyone in advertising sales.
The lower cost does not make CTV identical to search or social media. It does make experimentation more practical for brands that once viewed television as a club with an intimidating cover charge.
Haarlow said video retains an important advantage as more advertisers enter the market: scale. Audiences are already there and television can provide mass reach that brands would otherwise need to piece together through several channels.
The next challenge is connecting more data sources to that viewing activity so advertisers can understand how exposure contributes to business outcomes.
Chasing one metric could leave brands short
Budget pressure is encouraging companies to focus on a single key performance indicator. Haarlow warned that this can push advertisers to remove too much money from broad-reach channels and direct it toward tactics that produce cheaper immediate results.
That approach may look efficient in the short term. It can also ignore the “halo effect” created by upper-funnel advertising.
“My fear is that partners or clients will just cut too much from the more mass channels and go too much into the very specific performance-based channels,” Haarlow said.
Brands that stop investing at the top of the funnel may later see weaker returns in marketing mix modelling or other long-term assessments. Rebuilding lost awareness can take considerably longer than cutting the budget did. Competitors may also use the opening to claim market share, because they rarely observe a rival’s retreat and politely leave the space untouched.
Measurement needs to pick up the pace
Asked what improvement she most wanted to see, Haarlow gave a direct answer: “I think measurement’s gotta get faster.”
Campaign results can still take four, six, eight or even 12 weeks to arrive. That delay clashes with expectations that advertisers should optimize toward precise business outcomes in real time.
Haarlow said the problem is partly technological and partly organizational. Clients must become comfortable with newer measurement platforms and models instead of relying only on slower marketing mix modelling or return-on-investment studies. Technology providers must also process enough data quickly enough to support useful campaign changes.
Shields joked that CFOs and marketing mix models had emerged as the two villains of Cannes. Haarlow noted that the dynamic was “pretty much the same as 10 years ago,” only with a different topic.
CTV, in other words, has acquired abundant inventory, flexible buying and measurable outcomes. Now it just needs its reporting to arrive before everyone has forgotten what the campaign was selling.
You’re watching Beet Talks at Cannes Lions 2026, presented by Roku. For more videos from this series, please visit this page. You can find all of our coverage from Cannes Lions 2026 here.