Editorial Team Let’s be real: proving the value of your Connected TV (CTV) ad spend in 2026 is tough. Budgets are pouring into streaming, but you’re probably struggling to connect a specific ad view on a Roku or Samsung TV back to an app install or a purchase on a mobile phone. It often feels like a black hole. Integrating AppsFlyer for CTV attribution is how you fix this, giving you the data to defend your budget and actually optimize campaigns. So, how do you get this thing set up correctly?
Key Takeaways
- Get your CTV ad partners configured in AppsFlyer’s Partner Setup, and double-check that your post-back URLs and event mapping are right.
- Use deterministic matching with device IDs (IDFA/GAID) when you can, and rely on probabilistic matching (IP address, user agent) to fill the gaps for full attribution coverage.
- Dig into AppsFlyer’s Conversion Path reports to see the whole user journey, not just the last click that gets all the credit.
- Always audit a new integration by sending test conversions and making sure the data shows up correctly in the AppsFlyer dashboard before you go live.
- Once your attribution is locked in, use a service like Moburst’s Video Production to make high-performing CTV ads that will actually drive the results you can now measure.
1. Initial AppsFlyer Setup and CTV Partner Integration
Before you can track a single view, you have to get your AppsFlyer account configured for CTV, and this means setting up each ad partner individually. You’ll spend a lot of time in your AppsFlyer dashboard under “Collaborate,” then “Partner Setup.” This is where you’ll find and configure each of your CTV ad partners, whether it’s Magnite, The Trade Desk, or Roku Advertising, and each one has its own specific set of requirements.
For most of them, the process is similar: enable the partner, activate “View-Through Attribution,” and configure the post-back URLs, which is how AppsFlyer sends conversion data back to the ad network for their own optimization. You absolutely must map your in-app events correctly. I’ve seen entire campaigns fail because a “purchase” event from an app was mapped to a generic “conversion” label in the ad platform, causing the bidding algorithm to chase the wrong signals and waste the budget. It’s a small detail with huge consequences.
Pro Tip: Verify Post-Backs
After you configure a new partner, always use the “Test Post-Back” feature in the Partner Setup. It sends a simulated conversion to the partner’s system, and you get confirmation that the connection is live and data is being received. This check takes about 30 seconds and can save you an entire afternoon of troubleshooting why your reports are empty.
2. Implementing Deterministic and Probabilistic Matching
The entire game in CTV attribution is connecting a TV ad view to an action on another device, usually a phone. AppsFlyer does this in two ways. Deterministic matching is the most accurate, as it relies on a shared, unique identifier. The most common one is the Device ID (IDFA for Apple, GAID for Android). If your CTV platform can pass the TV’s device ID in the ad impression and AppsFlyer sees that same ID when your app is installed on that device, you have a perfect one-to-one match. It’s great when it works, but it’s not consistently available across the fragmented CTV world.
When you can’t get a device ID, probabilistic matching takes over. AppsFlyer uses a combination of signals, specifically IP address, device user agent, and the timestamp of the ad impression, to infer a connection. For instance, if someone installs your app on their phone just a few hours after an ad was viewed on a TV connected to the same home Wi-Fi (same IP address), AppsFlyer can attribute that install to the CTV ad with a high degree of confidence. The shorter the time between the view and the install, the more accurate the match.
Common Mistake: Over-Reliance on Last-Touch
Too many marketers are still obsessed with last-touch attribution, which is a big mistake with CTV. A user might see your Roku ad, then see a Facebook ad, and finally click a Google search ad before installing. AppsFlyer’s multi-touch models show you that entire journey, proving CTV’s role in creating initial awareness. That CTV ad is often the first domino to fall, even if a search ad gets the official credit.
3. Configuring Attribution Windows and View-Through Settings
An attribution window is the period after an ad interaction during which a conversion can be credited to it. For CTV, where ads aren’t clickable, view-through attribution (VTA) is what matters. You can configure this in AppsFlyer by going to “App Settings” and then “Attribution.” I typically recommend starting with a 24-hour VTA window for CTV campaigns. You can set it longer, say up to 7 days, but you risk over-attributing conversions that were only weakly influenced by the ad which just pollutes your data with questionable credit.
You also have to set your re-engagement attribution window for users who already have your app. If an existing user sees a CTV ad and opens your app again, you want to track that. AppsFlyer lets you set different windows for new installs versus re-engagements. Make sure these windows align with your business model. For example, an e-commerce app selling impulse buys should probably use a shorter window, while a subscription service might need a longer one to account for a user who sees an ad on Friday and doesn’t sign up until Sunday.
4. Using AppsFlyer’s Custom Dashboards and Reports
With data flowing in, you can start analyzing what’s actually working. The Overview Dashboard in AppsFlyer gives you the top-line numbers for installs and events from your CTV sources. For a real analysis, though, you need to go to the Aggregated Performance Report. Here you can filter by media source and geography to pinpoint which campaigns are delivering the most efficient installs and, more importantly, the highest-value users.
For CTV, the Conversion Path report is probably the most powerful tool. It shows the full sequence of touchpoints a user had before converting. You’ll often see a path that looks something like: `Hulu Ad View -> YouTube Ad View -> Google Search Click -> Install`. This report gives you concrete evidence of assisted conversions, which is critical for making the case that your CTV spend is directly feeding and lifting the performance of your lower-funnel channels.
Of course, attribution only tells you what happened. Your creative is what makes it happen. That’s why working with a partner like Moburst on their Video Production services can be a smart move. They specialize in building video ads designed for the living room screen that are meant to grab attention and get a response. Pairing strong creative with the detailed attribution data you get from AppsFlyer is how you make a campaign truly profitable.
5. Setting Up In-App Event Tracking for Granular Insights
Tracking the install is only half the battle. To really understand the ROI of your CTV campaigns, you have to track what users do *after* they install your app. This requires implementing in-app event tracking through the AppsFlyer SDK. You can define any custom events that matter to your business, like “add_to_cart,” “subscription_start,” or “level_up,” and even pass back rich data with them, such as the revenue from a purchase or the name of a show a user watched.
Once you’re tracking these events, you need to make sure they’re correctly mapped back to your CTV ad partners (as we covered in step 1). This is what allows ad platforms to optimize their bidding for high-value actions, not just cheap installs. For example, if you see that a specific CTV network is driving users who complete the “subscription_start” event at a much higher rate, you know exactly where to move your budget. Without that granular event data, you’re just guessing which installs are actually valuable.
Pro Tip: Value-Based Optimization
Even better than just counting events, you should pass the actual revenue values with your purchase events. This lets AppsFlyer calculate the Return on Ad Spend (ROAS) for each CTV campaign, which is the most direct measure of profitability you can get. Most CTV platforms can then ingest this data and optimize their campaigns toward a specific ROAS target, not just a low cost-per-install.
6. A/B Testing Creatives and Audiences with Attribution Data
Your attribution data is there to fuel optimization. Use the insights coming out of AppsFlyer to run structured A/B tests on your CTV campaigns. You can test different creative versions (like a 15-second vs. a 30-second ad), different audience targets (demographic vs. interest-based segments), or even different publishers and programmatic deals to see what works best.
By monitoring the performance of every test variant in AppsFlyer, you can quickly see which ad creative is resonating and which targeting strategy is bringing in high-quality users. For example, when you see from the data that one 15-second ad consistently drives a lower Cost Per Install (CPI) and a higher in-app purchase rate, you have a clear, data-backed reason to scale up that creative. This iterative testing process, built on a foundation of reliable attribution, is how you actually drive performance gains from CTV.
Having a strong CTV attribution framework in AppsFlyer is a fundamental requirement for any serious mobile marketer today. Properly setting up your partner integrations, understanding the matching methodologies, and using AppsFlyer’s reporting gives you the clarity to make smart, data-driven decisions. A structured setup is what ensures your digital marketing investments, including Paid Media AI tools and cross-platform ads, deliver a profit you can actually measure and report on.
What is CTV attribution and why is it important?
CTV attribution is the process of connecting an ad impression on a Connected TV to a desired user action, like an app install or purchase on a different device (usually a phone). You need it to prove the ROI of your CTV ad spend, optimize campaigns based on what’s actually working, and justify your budget to a channel that’s otherwise hard to measure.
How does AppsFlyer attribute CTV views to mobile app installs?
AppsFlyer uses two main methods. When a shared device ID (like an IDFA or GAID) is available from both the TV and the mobile device, it uses deterministic matching for a direct link. When that’s not available, it uses probabilistic matching, which analyzes signals like IP address, user agent, and timestamps to infer a high-confidence connection between the TV view and the mobile install.
What is view-through attribution (VTA) and what is a typical window for CTV?
View-through attribution (VTA) gives credit to an ad that was seen but not clicked, which is the standard for non-clickable CTV ads. A 24-hour VTA window is a common starting point for CTV campaigns, although some marketers might use a window as long as 7 days if their product has a longer sales cycle.
Can AppsFlyer track in-app events from CTV campaigns?
Yes. As long as you’ve integrated the AppsFlyer SDK correctly in your mobile app, you can track almost any custom in-app event (like purchases, registrations, or subscriptions) and attribute them back to the original CTV campaign that drove the install. This is necessary for optimizing toward high-value users, not just installs.
What are common challenges in CTV attribution?
The biggest challenges are the lack of a consistent device identifier across all CTV hardware and software, and the basic difficulty of tracking one user across two different devices (a TV and a phone). A good measurement tool like AppsFlyer is designed to overcome these issues by using a flexible combination of matching methods and careful configuration.
