Targeting decides who sees your streaming TV ad. Measurement tells you whether it worked. Set up both before you spend a dollar, not after, because you can’t measure a result you didn’t plan to count.
This guide walks through the CTV targeting options, the metrics that matter, how attribution works across the industry, and the honest ways to measure results yourself.
Part 1: CTV Targeting Capabilities
CTV targeting works in layers: where people live, who they are and what they’re into. Start with the first layer and add the others only where they help.
1. Geographic targeting
Geography is streaming TV’s biggest advantage over linear TV, which is usually bought across a whole market. On Omnya you can draw your area in several ways:
- ZIP codes. Any of 42,789 US ZIP codes, or paste a whole list at once.
- Radius. A circle around your address.
- City or county. For businesses that serve a whole town or region.
- Metro area (DMA). Any of the 210 DMAs, for wider campaigns.
- Congressional district. Useful for advocacy and public-sector work.
Geographic targeting
- Include
Chicago, IllinoisMetro
9,712,804 reach
- Include
Chicago, ILin ILCity
2,705,664 reach
- Include
Illinois District 7IL-7District
1,712,467 reach
Or paste ZIP codes in bulk
Some platforms in the industry also offer custom map polygons or uploaded address lists. For most local businesses, a set of ZIP codes or a radius does the same job.
2. Demographic targeting
CTV platforms build demographic segments from household-level data supplied by data providers. The options vary by platform; these are the common ones.
| Demographic | Typical segments |
|---|---|
| Age | 18–24, 25–34, 35–44, 45–54, 55–64, 65+ |
| Gender | Male, female, all |
| Household income | Brackets such as under $50K, $50–100K, $100K+ |
| Household composition | Single, married, with children, empty nesters |
| Homeownership | Homeowner, renter |
| Education | High school, some college, bachelor's, graduate degree |
3. Behavioral and interest targeting
Interest segments group households by what they do and like. Omnya offers 1,000+ audience segments, so you can match the ad to people likely to care. The kinds of segments you’ll find:
- Lifestyle interests: fitness, food and cooking, travel, home and garden.
- Purchase behaviors: online shoppers, value shoppers, category buyers.
- Content preferences: sports, news, drama, documentaries.
4. In-market and intent targeting
Across the industry, in-market audiences are households showing signs they’re about to buy in a category, built from browsing and purchase signals. Common examples are automotive (shopping for a car), home improvement and travel.
Availability depends on the platform and the category. Credit, insurance and housing are restricted, so treat intent segments there with care.
5. First-party data and retargeting
People who already know you are often your best audience. On Omnya, website-visitor retargeting unlocks once the Omnya tracking pixel has seen 10,000 visitors to your site.
Elsewhere in the industry you’ll also meet customer-list matching and lookalike audiences built from your first-party data. They are worth asking about, but they aren’t required to run a good local campaign.
Layer without over-narrowing
Each layer makes the audience more relevant and smaller. An auto repair shop might target the ZIP codes within about 5 miles, plus people interested in cars and auto services. Add a third or fourth layer and there may be too few households left to deliver your budget.
Part 2: CTV Measurement & Attribution
Measurement answers two different questions. Did the ad run the way you planned? And did it change anything for your business? The first comes from delivery metrics, the second from results you set up to count.
Core performance metrics
| Metric | What it means | How to read it |
|---|---|---|
| Impressions | Each time your ad is served | Compare with the forecast to see whether the campaign delivered |
| Reach | How many different households saw the ad | Bigger reach means more new people; it shrinks as you add layers |
| Frequency | Average times each household saw it | Too low and nobody remembers; too high and you annoy people |
| Video completion rate | Share of ads watched to the end | Usually high on streaming TV; a dip can point to a weak opening |
| CPM | Cost per 1,000 impressions | What you pay per 1,000; Omnya Streaming TV starts at $18 |
| Conversions | Actions on your site after the ad: a sale, a booking, a form | Needs a tracking pixel installed before launch |
Read more on frequency, video completion rate and CPM in the glossary.
Where you see the numbers on Omnya
OmnyView is Omnya’s cross-channel report: impressions, video completion, clicks and conversions in one place. Each channel also has its own report, with a publishers tab showing where your ad ran, and you can share any report by link.
OmnyView · last 30 days
Illustrative numbers- Impressions
- 158K
- Video completion
- 93.1%
- Clicks
- 1,204
- Conversions
- 212
How CTV Attribution Works
Attribution tries to connect seeing an ad with a later action, usually within a set window of days. These are the main methods used across the industry, most of them sold by specialist measurement providers.
| Method | How it works | Best for |
|---|---|---|
| Website visit attribution | A pixel on your site is matched with households that saw the ad | E-commerce, lead generation, service businesses |
| Conversion attribution | Purchases, form fills or calls from exposed households are counted | Online sales and phone-based businesses |
| Foot-traffic attribution | Third-party mobile location data is matched with exposed households | Retail, restaurants, fitness, automotive |
| Brand lift studies | Surveys compare awareness among people who saw the ad and people who did not | Launches and brand-building campaigns |
| Sales lift (matched markets) | Sales in advertised areas are compared with similar areas left out | Any business that can split its area |
Be wary of any promise of a fixed visit rate, conversion rate or lift. Those numbers depend on your category, offer, area and season. Your own before-and-after numbers are worth more than someone else’s average.
Honest Ways to Measure CTV Results Yourself
You don’t need an attribution partner to learn whether streaming TV is working. These steps cost nothing extra and work at any budget.
Install the pixel before launch
Add the Omnya tracking pixel to your site, or connect Meta or Google conversion tracking, so visits and conversions are counted from the first day.
Give TV its own offer
A code, phrase, QR code or short web address that only appears in the TV ad (“Mention TV for 10% off”) shows exactly who came from it.
Record a baseline by ZIP
Note sales, bookings, calls and walk-ins by ZIP code for a few weeks before launch. Then compare the weeks the campaign runs.
Hold out some ZIP codes
Split similar ZIP codes into two groups and advertise in only one. If the advertised group grows faster than the held-out group, that difference is your lift.
Watch searches for your name
TV viewers often search rather than click. A rise in searches for your business name during the campaign, in Google Search Console or your Google Ads account, is a useful signal.
Ask at the counter
“How did you hear about us?” is old-fashioned and still works.
Decide what you’ll count before the ad runs. Measurement you add afterward can only guess.
Set Up Targeting and Measurement Before You Spend
Run through this list before you pay for your first campaign.
- Area chosen: ZIP codes, radius or city that match where your customers live.
- One extra layer at most: an interest or age group, if your category allows it.
- Reach checked: the forecast shows the campaign can deliver in full.
- Pixel installed: visits and conversions are being counted.
- TV-only offer in the ad: a code, phrase or web address you can track.
- Baseline recorded: a few weeks of numbers by ZIP code, with some ZIPs held out if you can.
Then launch, give the campaign a few weeks, and compare. Adjust the area, the layers or the ad based on what you see. See how OmnyView reports results.