

CTV platforms match website visitors to TV households through two mechanisms: IP address matching and CRM-based deterministic matching. Your IP address is the unique number that identifies your home's internet connection — every device on the same Wi-Fi shares it, including the TV in the living room. When a visitor hits your site, the CTV platform captures that IP address. When a streaming device on the same home network later loads an ad, the match completes. No cookies involved — CTV identity resolution runs on IP signals and first-party data, which is why it works in a post-third-party-cookie environment where browser-based targeting has degraded.
CTV apps — Hulu, Peacock, Tubi, and thousands of others — are native streaming applications, not browsers. There's no cookie jar to read or write. (A cookie is the small file a website leaves in your browser so it can recognize you when you come back — streaming apps don't have that.) Identity in the streaming environment is built on different signals: the IP address of the household network, the device ID of the streaming hardware, and authenticated login data from the streaming service itself.
When you place a CTV pixel on your website, it fires on page load or conversion events the same way a Meta pixel does. The difference is what happens next. Instead of writing a cookie to a browser profile, the pixel logs the household IP address of the visitor. When that same IP later appears in a streaming request — a viewer starting a show on a Roku or smart TV connected to that network — the match completes. The household that visited your site has been identified as an active streaming household, and your ad can serve to that TV screen.
This is a household-level signal, not a user-level signal. One IP address covers everyone in the house. That's a meaningful distinction from Meta's individual-level targeting, and it affects how you interpret frequency caps and conversion attribution — a converted household doesn't tell you which person converted, only that someone in the household did.
Deterministic matching is the second identity resolution mechanism, and it's more precise than IP matching. The word 'deterministic' just means it's based on something you know for certain — in this case, a confirmed email address from your CRM — rather than an inference from network behavior like an IP address.
The process: you export a hashed email list from your CRM, upload it to the CTV platform, and the platform matches those emails against its identity graph. 'Hashed' means each email address has been converted into a scrambled code — it can be matched against another database without exposing the original email address, which keeps customer data private during the transfer. Identity graphs are built from aggregated first-party data (authenticated logins on streaming services, retail loyalty accounts, purchase histories, and device registrations). Think of an identity graph as a lookup table that connects different pieces of data about the same household: it knows that an email address in your Klaviyo list and the Roku TV in that person's living room belong together. When a match is found, you can serve ads to that specific household on CTV.
This is the CTV equivalent of uploading a customer list to Meta. Match rates are lower than a broad IP retargeting pool in raw volume, but the quality of each matched household is higher: these are people you already know, reached on a channel where they're actively watching.
The two mechanisms complement each other. IP matching captures recent web visitors who aren't in your CRM yet — people who browsed but didn't convert, or first-time site visitors. Deterministic CRM matching reaches known customers and segments with higher confidence. An effective CTV retargeting setup uses both.
CTV without first-party data matching isn't performance TV. It's brand advertising with a shorter production timeline. Demographic-only targeting — age, income, geography — is what the legacy linear TV model always ran on. The data connection is what separates a performance CTV channel from an awareness play.
On Vibe.co, a co-managed self-serve CTV platform with dedicated account support built in, the Klaviyo integration syncs audiences directly to Vibe's Identity Intelligence ID graph. The setup works the same way as Meta's custom audiences — except it targets streaming households instead of social profiles. Klaviyo segments (non-purchasers, lapsed customers by LTV tier, email subscribers who haven't converted) map directly to CTV audiences. No manual CSV export. No separate upload workflow. The segment you built for email retargeting becomes a streaming TV audience through the same integration.
Shopify purchaser suppression works the same way as Meta exclusion audiences: existing customers excluded from prospecting so impressions go only to households that haven't converted yet. High-LTV buyer lookalikes extend reach to new households that pattern-match against your best customers across the 120M+ profile ID graph.
Branded Bills, a custom headwear and apparel brand, used Klaviyo IP matching on Vibe to retarget warm audiences across premium streaming channels. The result: 311% ROAS, a peak retargeting day of 2,209% ROAS, and $0.33 cost per session. The Klaviyo integration was the mechanism — a live sync that kept the retargeting segment current rather than a one-time audience upload.
Match rates in CTV retargeting are lower than Meta's pixel-based audience, and it's worth being honest about why. IP matching depends on that household IP staying consistent between the site visit and the streaming request — if a household's IP address changes (which happens on mobile networks, with ISPs that rotate addresses periodically, or when someone uses a VPN), the match fails. Deterministic CRM matching is limited by the quality and recency of your email list; a clean Klaviyo list with validated emails matches at a meaningfully higher rate than one with stale or synthetic addresses.
Industry estimates from identity graph providers put household-level IP-based match rates in the range of 40–60% of web visitors. In other words: if 1,000 people visited your website last week, roughly 400–600 of them can be matched to a streaming TV household and served your ad on CTV. CRM-based deterministic matching typically runs lower on raw volume but higher on confidence per matched household. *(Verify current figures against Vibe platform data or ID graph provider benchmarks before publish.)*
The practical implication: you're not replicating your full Meta retargeting audience on CTV. You're reaching a high-confidence subset — households that both visited your site and are active streamers, or known customers whose emails resolved to streaming device IDs. The audience that doesn't match is typically non-streamers, mobile-only users without a connected TV, or households where IP rotation prevented the link.
That subset performs. Blindster, a custom window treatment retailer, retargeted warm website audiences on Vibe and measured $45 CPA on CTV vs. $89 on Meta — both in the same Northbeam setup, using identical attribution methodology. A 40–60% match rate on your site visitors still delivers a high-intent, streaming-active audience that converts at rates that justify the CPM. For teams building a full-funnel streaming TV strategy, CTV retargeting is typically the highest-ROAS layer precisely because the audience is already warm.
CTV attribution is view-through, not click-through. There's no clickable link in a streaming TV ad — a household that sees your ad and converts within the attribution window (7 days is standard) gets credited to CTV. That's a different model from Meta's click-based attribution, but it works the same way: a window of time after the ad exposure, during which a purchase gets counted as a CTV-influenced conversion.
Measurement parity is the key. CTV retargeting results need to appear in the same Northbeam or Triple Whale dashboard as your Meta retargeting campaigns. When streaming TV results live in a separate platform report, they get reconciled quarterly and treated as a brand investment. When they sit next to paid social and search in the same dashboard, CTV retargeting is managed on the same cadence — same CPA target, same weekly review, same creative testing cycle. That's what TV functioning as a real third pillar alongside Search and Social requires.
Vibe connects impressions to Northbeam and Triple Whale via the Clicks + Deterministic Views model, the same one those tools use for paid social attribution. The attribution window setup guide covers the exact configuration. The CTV retargeting ROI guide goes deeper on holdout methodology — where a randomly selected group of matched households doesn't see your ad, so you can compare their conversion rate against households that did and measure what the ad actually caused vs. what would have happened anyway.
Retargeting used to live in Search and Social. CTV extends it into the third pillar — the living room screen — with the same customer data and measured in the same stack. For teams that already run retargeting on Meta and Google in-house, the operating model transfers directly. For context on when to make the move, the Facebook audience exhaustion guide and the scale beyond Google and Meta playbook cover the channel-expansion decision in full.
CTV platforms use two mechanisms: IP address matching and CRM-based deterministic matching. IP matching connects the household IP address from a web visit — the unique number identifying your home internet connection — to the same IP seen on a streaming device on that network. No cookies required. Deterministic matching starts from a confirmed identifier (a hashed email address from your CRM) and matches it to streaming device IDs via an identity graph. The two work together: IP matching captures recent visitors who aren't in your CRM, deterministic matching reaches known customers with higher confidence.
CTV apps are native streaming applications, not browsers — there's no cookie environment to read or write. Identity resolution runs on IP address signals (the household network IP from a web visit matches to the same IP on the streaming device) and on first-party data via identity graphs that map email addresses to streaming device IDs. Because neither mechanism relies on third-party cookies, CTV retargeting works in a post-cookie environment where browser-based targeting has degraded.
Deterministic matching is identity resolution that starts from a confirmed identifier — typically a hashed (scrambled) email address from your CRM — rather than an inferred signal like IP behavior. You export a customer list from Klaviyo or Shopify, upload it to the CTV platform, and the platform matches those emails to streaming device IDs via its identity graph. The match is deterministic because it begins from something you know for certain rather than a probabilistic guess. Match rates are lower than IP retargeting in raw volume but higher in confidence per matched household.
IP-based household matching typically runs 40–60% of web visitors — meaning roughly half the people who visited your site can be matched to a streaming TV household. CRM-based deterministic matching depends on the quality and recency of your email list; a clean, validated Klaviyo list matches at a meaningfully higher rate than a stale one. The audience that matches is high-confidence: households that visited your site and are active streamers, or known customers whose emails resolved to streaming device IDs. Blindster's $45 CPA on CTV vs. $89 on Meta — both measured in the same Northbeam setup — confirms that even a partial match rate delivers results that justify the investment.
Place the Vibe pixel on your site (it fires the same way as a Meta pixel), connect your Klaviyo or Shopify account for CRM-based audience syncing, and configure your retargeting segments — non-purchasers, lapsed customers, high-LTV lookalikes — the same way you'd build custom and exclusion audiences on Meta. Attribution syncs to Northbeam or Triple Whale via the Clicks + Deterministic Views model, so CTV results appear in the same dashboard as your existing campaigns. The attribution window setup guide covers the exact configuration.


An identity graph connects email addresses, device IDs, and household data to enable cookie-free CTV ad targeting. Here's how it works in streaming TV.
How to measure CTV incrementality alongside programmatic: holdout tests, iROAS benchmarks, and the attribution stack for enterprise media teams.
When Facebook audiences saturate, the fix isn't more spend. Here's how DTC brands extend reach to CTV using the same Klaviyo data they built on Meta.
Yes — streaming TV ads are bought programmatically, just like display. Same RTB auctions, sharper targeting, zero bot traffic. Here's how it compares.
Self-serve CTV puts targeting and attribution in your hands. Agency-managed adds fees and a data layer you can't verify. Here's how to choose.

© 2026 Vibe, Inc.