Everyone knows Connected TV (CTV) advertising gets you in front of a ton of engaged viewers, but proving its effect on upper-funnel goals like brand awareness has always been a real problem for marketers. This teardown looks at how a big telecom provider, we’ll call them “TelcoCo,” finally cracked this nut. They used a sophisticated CTV attribution setup with their mobile measurement partner (MMP) to see what their video spend was *really* doing, and the results completely upended their media buying strategy.
Key Takeaways
- Exposing audiences to the CTV campaign produced a 12% lift in unaided brand recall compared to unexposed control groups, tying the CTV budget directly to awareness gains.
- Switching to an impression-based attribution model with a 30-day lookback for CTV campaigns uncovered a 2.5x higher ROAS than their old last-touch mobile model showed.
- The data showed a frequency cap of 3 to 4 exposures per user per week on CTV gave them the best brand lift before ad fatigue started setting in.
- By linking CTV impression data with their mobile app install data, TelcoCo managed to drop their effective cost per qualified lead (CPL) by 18% over six months.
- Creative A/B testing proved that punchy 15-second spots, each focused on one product benefit, were better at driving initial brand recognition on CTV than the longer 30-second ads.
Campaign Overview: Driving Brand Recall and App Installs
Back in Q3 2025, TelcoCo kicked off a major brand awareness campaign, and a huge piece of the $1.8 million budget went to CTV. They had two main goals: make sure people remembered their new unlimited data plan and get new users to download their mobile account management app. The whole thing ran for eight weeks, from September 1 to October 26, 2025, with the budget spread across CTV, social, and programmatic display.
For years, TelcoCo couldn’t figure out what their CTV spend was doing besides generating impressions and frequency. Their attribution setup gave all the credit for app installs to mobile channels, so CTV’s role was always a question mark. This campaign was designed to finally connect the dots.
Strategy: Bridging the CTV-Mobile Attribution Divide
The heart of TelcoCo’s new strategy was a multi-touch attribution model built to see CTV impressions. We all knew that just looking at last-touch attribution for app installs was painting a completely wrong picture of CTV’s influence at the top of the funnel. So, we set up a system to connect a CTV ad exposure to a mobile action, even if that install happened days or weeks later.
Our approach had three parts:
- Impression-Based Attribution for CTV: We had the MMP capture CTV ad impressions and give them credit when a user later installed the app, no click needed. This depended entirely on a solid device graph to link the TV to the phone in a household.
- Control Group Testing: This was non-negotiable. We picked specific Designated Market Areas (DMAs) to be a control where no CTV ads ran. This gave us a clean A/B test to compare brand lift and install rates against the DMAs that saw the campaign.
- Sequential Messaging: The CTV ads were all about brand storytelling and big-picture benefits to build awareness. Then, we’d retarget those same viewers on their mobile devices (social and display) with direct-response ads to push the app install, capitalizing on the brand familiarity the CTV spot created.
Creative Approach: Storytelling for the Big Screen
TelcoCo came up with two creative concepts for CTV, cut as both 15s and 30s. The 15-second ads were all about a single, punchy benefit of the unlimited data plan, with fast visuals and not much text. The 30-second versions told a broader story about lifestyle moments made possible by great connectivity. The creatives were all made for a lean-back TV experience, trying to build an emotional connection instead of screaming for a click.
- 15-second spot: “Freedom to Connect”, which just showed flawless video calls and streaming in cool outdoor places.
- 30-second spot: “Your World, Uninterrupted”, a little story about a family using data for everything, work, school, fun, without any buffering.
Once we retargeted on mobile, the ads got way more direct. We used screenshots of the app’s UI and had obvious calls to action like “Download Now” or “Manage Your Plan.” This sequence let each channel do what it does best.
Targeting: Precision Across Screens
On the CTV side, we targeted using a mix of demographics (households with income over $75k, ages 25-54), psychographics (tech early adopters, heavy streamers), and context (placing ads in news, entertainment, and sports programming). Critically, we filtered out any households we could identify as existing TelcoCo subscribers to keep the focus on new customers. The campaign ran on the big streaming platforms, Hulu, Roku Channel, and Peacock, using their first-party audience data.
For the mobile part, we built retargeting audiences from users who’d seen the CTV campaign but hadn’t installed the app. We also used lookalike audiences modeled off their best existing app users to expand our reach.
Performance Analysis: What the Data Revealed
The results really showed how much CTV can do when you measure it correctly. Here’s how it all broke down:
Brand Awareness Metrics
We hired an independent research firm to run post-campaign brand lift studies, and the results were clear. Surveying both the exposed and control groups showed us:
- Unaided Brand Recall: A 12% lift in the CTV-exposed group versus the control. That’s a solid win.
- Ad Recall: 28% higher ad recall among people who saw the CTV ads. They remembered the creative.
- Brand Favorability: A smaller, but still positive, 5% increase in brand favorability in the exposed group.
CTV Campaign Metrics
CTV Campaign Performance (8 Weeks)
- Total Impressions: 45 million
- Average Frequency: 3.8 impressions per user per week
- Video Completion Rate (VCR): 92% (for 15-sec spots), 85% (for 30-sec spots)
- Cost Per Completed View (CPCV): $0.03
- Total CTV Spend: $900,000
Attribution and Conversion Metrics
This is where the lightbulb went on. By connecting CTV impressions to mobile app installs over a 30-day lookback window, the entire picture of channel performance changed.
App Install Attribution: Last-Touch Mobile vs. Multi-Touch (Including CTV)
| Metric | Last-Touch Mobile Attribution | Multi-Touch (CTV + Mobile) Attribution |
|---|---|---|
| Total App Installs Attributed | 18,500 | 46,250 |
| Cost Per Install (CPI) | $15.14 | $6.05 |
| Return on Ad Spend (ROAS) | 1.8x | 4.5x |
| CTV’s Attributed Share of Installs | 0% | 60% (27,750 installs) |
The massive difference in attributed installs and CPI shows how much conversion credit gets lost in the dark when CTV isn’t wired into your attribution. Without this view, over 27,000 app installs would have been wrongly credited to mobile channels, completely ignoring the fact that CTV did the heavy lifting to create that initial interest. So many marketers are still stuck in that siloed view, unable to see how the first touch influences the last.
What Worked Well
- Integrated Attribution: The single best decision we made was committing to a real, impression-based CTV attribution model. It finally let TelcoCo prove CTV’s dollar-for-dollar contribution to app installs, giving them the confidence to increase their investment.
- Control Group Methodology: The geo-split test was our silver bullet. It gave us undeniable proof that the CTV ads were directly responsible for the brand lift. Without that control group, we’d just be guessing.
- Creative Sequencing: Using brand-heavy CTV ads to warm up the audience before hitting them with direct-response mobile ads was the right call. It respected how people watch content on different screens. The 15-second CTV spots were particularly good at getting the brand name to stick.
- Frequency Capping: We held the average frequency to 3.8 per user per week, and it worked. Looking back at the data, we saw that going above 4.5 exposures a week gave us diminishing returns on brand lift and started driving up our CPMs.
What Didn’t Work as Expected
- 30-Second CTV Spot Performance: The 30-second spots had a decent VCR, but they didn’t really move the needle on brand recall or app installs any more than the 15s did. And their CPCV was a bit higher. For top-of-funnel goals on CTV, it looks like short and sweet is the way to go.
- Initial Budget Allocation: We started with a 50% CTV, 30% social, 20% display split. After seeing the attribution data, we realized CTV was massively underfunded for the impact it was having on app installs.
- Lag in Data Integration: We had up to a 24-hour delay getting CTV impression data synced up with mobile events. It wasn’t a dealbreaker, but it did make real-time optimizations tricky. That’s a pretty common headache with server-to-server integrations, but we’ve since found ways to shrink that delay.
Optimization Steps and Future Outlook
Armed with this data, TelcoCo made some big changes going forward:
- Increased CTV Budget: In their next campaign cycle, they bumped the CTV allocation to 65% of the total media spend. They now see it as the foundation of their acquisition efforts.
- Prioritized 15-Second CTV Creatives: From now on, 15-second ads are the default for brand awareness. They’ll save the longer formats for more specific product education or promotional pushes.
- Enhanced Data Pipelines: They put money into shortening the data transfer latency between their CTV partners and the MMP, with the goal of getting to near real-time reporting.
- Refined Retargeting Segments: They also tightened up their mobile retargeting. Users who installed the app were excluded much faster to avoid wasting spend on people who already converted.
This whole campaign proved that with the right measurement in place, CTV is an engine for actual, measurable conversions, not just vague awareness. Building the attribution framework to see the entire customer journey, from the TV screen to the app install, is how you get a real advantage.
To really see what CTV is doing, you need a full-funnel attribution strategy that connects those upper-funnel TV views to lower-funnel results like an app install. By setting up impression-based tracking and using clean control groups, TelcoCo discovered where their conversions were truly coming from, letting them move budget with confidence and get a much higher return on their ad spend.
What is CTV attribution and why is it important for brand awareness?
CTV attribution means connecting the ads people see on their smart TVs to a specific outcome, like an app install or just remembering your brand. It’s essential for brand awareness because CTV is an incredible storytelling medium. Good attribution proves how those TV stories actually build brand recall, which is the first step toward any sale, even if no one clicks.
How can marketers connect CTV ad views to mobile app installs?
You connect CTV ad views to mobile app installs by using a device graph, usually provided by a mobile measurement partner (MMP). The technology anonymously links the TV in a household to the phones and tablets in that same household. So when someone sees your ad on their Roku and later installs your app on their iPhone, the MMP can connect the two events and give credit to the CTV ad view.
What is a good frequency cap for CTV brand awareness campaigns?
A solid frequency cap for a CTV brand campaign is somewhere in the 3 to 5 exposures per user, per week range. If you go much higher, you risk annoying people and seeing your results decline. Too low, and your message won’t break through the noise. But every campaign is different, so you should always test a few different frequency caps to find your own sweet spot.
Why is a control group essential for measuring CTV brand lift?
A control group is the only way to prove your CTV campaign actually caused a lift in brand metrics. By showing your ads to one group and deliberately *not* showing them to a similar group (the control), you create a scientific baseline. Without that baseline, you can’t tell if an increase in brand awareness was because of your ads or because of a million other things happening in the market.
What role do 15-second versus 30-second CTV ads play in brand awareness?
They do different jobs. 15-second ads are workhorses for driving initial brand recognition. They’re quick, get straight to the point with a single message, and people are more likely to watch the whole thing. 30-second ads are better for telling a more complex story or building an emotional connection. For pure top-of-funnel awareness, the shorter 15-second spots usually give you more bang for your buck.
