Connected TV has moved from an experimental line on the media plan to a channel marketers treat as core. US CTV ad spending reaches roughly 38 billion dollars in 2026, up about 14 percent from 33.35 billion in 2025, and the channel is projected to pass traditional TV ad spending for the first time in 2028. With that money flowing in, the question buyers keep asking is no longer whether to run CTV but how tightly they can aim it. Geofencing on connected TV answers part of that question, though the precision it offers comes with limits that get glossed over in a lot of sales decks.
This guide walks through how CTV geofencing works in 2026, where it earns its keep, what it genuinely cannot do, and how to set up a campaign in Display & Video 360. A worked example with a hypothetical local retailer ties the mechanics back to a real buying decision.
What CTV geofencing actually means
Traditional geofencing draws a virtual boundary around a physical place and serves ads to mobile devices that cross into it, using GPS and mobile signals. CTV geofencing works differently because a smart TV does not move and does not carry GPS. Instead, the boundary is built from the home's internet connection. Platforms map a television to a household through the IP address tied to that home's router, then layer in geographic data to confirm the location down to a region, city, or ZIP code.
The mechanism rests on three signals working together.
IP-block precision sits at the foundation. Every connected home shares one public IP address across its TVs, phones, tablets, and laptops, which lets a buyer treat the household as a single addressable unit rather than chasing individual devices. Household-level targeting groups viewing behavior across all of those screens back to one home, which matters on CTV because several people often watch the same television without separate logins.
ZIP and postal precision narrows the geographic frame. Streaming platforms resolve IP data to a postal area, and for most local businesses that ZIP-level frame delivers enough accuracy to reach the right neighborhoods while keeping the brand-building weight of a television placement. Some vendors market finer cuts using ZIP+4 or address matching through identity resolution, an approach often called addressable geofencing, which activates specific households and the devices linked to them rather than everyone inside a broad circle.
ACR signals add a behavioral layer on top of location. Automatic content recognition identifies what plays on a smart TV by fingerprinting short audio or video clips against a reference library, which lets advertisers know which households were exposed to particular programs or ads across both linear and connected TV. More than half of CTV buyers now use ACR data for targeting and rate it valuable when planning video campaigns. Combined with IP and device-graph data, ACR turns a location-based audience into one a buyer can sequence, suppress, or extend across screens.
Where it earns its keep
CTV geofencing fits any campaign where the buyer wants the reach and credibility of television without paying for a national audience they will never sell to.
Retail co-op programs use it to fund regional bursts where a brand and its local stockists split the cost, aiming the same creative at the ZIP codes around participating stores. This kind of store-level reach sits at the center of how we approach retail and eCommerce campaigns. A regional restaurant group running several locations can fence the trade areas around each one and rotate location-specific offers, so a viewer near the east-side branch sees that branch's promotion rather than a generic chain spot.
Automotive dealers lean on it because a dealership draws from a defined radius, and serving streaming ads only to households inside that radius removes most of the waste that a DMA-wide buy carries. We applied the same logic to a geo-targeted EV campaign on DV360, aiming delivery at the locations where intent actually clustered. Political advertisers value the same control for the opposite reason: campaigns map messaging to specific districts and precincts, and CTV lets them reach streaming households inside those boundaries that linear schedules tend to over- or under-serve.
The hyperlocal demand is real even in smaller markets. Search interest for connected TV advertising tied to specific mid-size cities shows that advertisers in those areas are actively looking for this capability, not just agencies in the top ten metros.
What it cannot do
The honest version of this story includes the ceiling, and apartment-level targeting is where CTV geofencing hits it. Because the technique resolves to a household's IP address and a postal area, it cannot reliably single out one unit inside a multi-dwelling building. Several apartments can sit behind shared infrastructure or rotate through dynamic IP assignment, so a fence that works cleanly for single-family homes blurs in dense housing.
The accuracy of IP-to-household matching is also weaker than most pitches admit. Independent data firm TruthSet found that matching an IP address back to a postal or email address is accurate only 13 to 16 percent of the time, and the same research estimates CTV advertisers are on track to waste 7.36 billion dollars in 2026 because of poor data quality. IPs change more often than targeting models assume, and dynamic assignment, VPN use, and mobile network switching mean yesterday's IP can belong to a different home today.
DV360 carries a hard technical limit worth knowing before you plan: targeting based on specific IP addresses is not available in the platform. Geographic precision in DV360 comes through its geography controls, not through uploading a list of household IPs. Buyers who need true address-level activation reach it through specialized addressable vendors rather than DV360's native targeting, and even then the match rate, not the promise, decides the result.
None of this makes the channel a poor choice. It means a buyer should treat ZIP-level reach as the dependable outcome and address-level precision as a probabilistic add-on that needs its own measurement.
How to set it up in DV360
Display & Video 360 handles CTV geofencing through geography targeting at the line-item level, and the workflow is straightforward once the limits above are clear.
Start by setting the line item's environment and inventory so it serves to connected TV devices, then open the geography section to add your targets. Regional targeting covers states, cities, and ZIP codes, and you can paste a line-separated list of ZIP codes directly. Each US ZIP resolves correctly when paired with the country code, so an entry reads as 11206, US. For tighter trade-area control, proximity targeting lets you build a radius around a street address or point of interest, which you can save as a location list at the advertiser level and apply across line items.
A few platform rules shape the build. Very small ZIP areas are sometimes best-effort and may not serve precisely, and you cannot target a ZIP if you have excluded the larger region, state, or metro that contains it. For a multi-location campaign, the cleaner structure is one granular geography per line item, since that gives you separate budgets, creative, and reporting for each trade area rather than one blended buy.
Layer behavioral and audience signals on top of geography only after the location frame is set, so the geographic boundary stays the controlling filter rather than competing with audience lists for delivery.
A worked example
Picture a hypothetical local retailer, Bridgewater Outdoors, an independent gear shop with two stores in a mid-size market. The owner wants streaming TV ads but cannot justify a DMA-wide buy that would reach hundreds of thousands of households outside any reasonable drive.
The build starts with two line items, one per store, each targeting the ZIP codes inside a 15-minute drive of that location, pasted in as a country-coded list. Store A's line item runs creative featuring its address and weekend hours; Store B's runs the same template with its own details. Both serve only to CTV inventory, so the spots land on the living-room screen rather than on mobile.
ACR data then informs frequency and sequencing. Households shown the brand spot three times without a follow-up signal move into a suppression segment, which keeps the budget from over-serving the same homes. For measurement, the retailer holds out a comparable set of ZIP codes as a control and compares store visits and online orders from the targeted areas against it, defining the attribution window before launch rather than after.
The realistic expectation matters here. Bridgewater reaches the right neighborhoods with high confidence at the ZIP level, treats any vendor claim of household-exact delivery as something to verify against the control group, and accepts that apartment-dense ZIP codes will carry more spillover than the suburban ones. That framing turns CTV geofencing into a measurable local channel rather than a precision promise that quietly underdelivers.
The takeaway for local buyers
CTV geofencing in 2026 gives local and regional advertisers a credible way to put television-grade creative in front of the specific communities they serve, anchored on IP-block and ZIP-level precision and sharpened by ACR-driven sequencing. The technique rewards buyers who plan around its real boundary, which is the household and the postal area, not the individual apartment, and who measure outcomes instead of trusting match-rate marketing. Set the geographic frame first, treat address-level precision as a probabilistic layer, and the channel delivers reach that linear television never offered at this level of control.
For a fuller picture of how the channel works end to end, see our Connected TV solutions overview, and if you would rather hand off the build, our breakdown of DV360 campaign management services covers what fully managed execution looks like.
