Quick answer
Addressable TV is television advertising in which different households or audience segments watching the same programme see different ads, chosen from household data such as demographics, viewing behaviour and location. It works on set-top boxes and smart TVs, not only streaming.
Formula: Cost per 1,000 on-target impressions = CPM ÷ share of impressions on target; addressable wins when its on-target share > addressable CPM ÷ linear cost per thousand on-target
What is addressable TV?
Addressable TV is television advertising where different households, or audience segments, watching the same programme see different ads. IAB UK defines it as the ability to serve personalised ads to different target groups and audience segments watching the same linear TV content, using criteria such as socio-demographic profile, viewing behaviour, interests and location.
In ordinary linear TV, an ad slot in a programme reaches everyone tuned in. With addressable TV, the broadcaster, set-top box or smart TV platform swaps the ad inside that slot for a chosen household, so a car brand can show a family sedan ad to some homes and a pickup ad to others while the show itself is identical. It sits next to connected TV and OTT (streaming delivered over the internet), but it is defined by the targeting, not by the pipe: addressable ads run on cable and satellite set-top boxes as well as on streaming devices.
Linear TV vs addressable TV
| Linear TV buy | Addressable TV buy | |
|---|---|---|
| What you buy | A slot in a programme or daypart | Impressions to chosen households inside a slot |
| Who sees the ad | Everyone watching | Only households that match the segment |
| Targeting data | Programme audience demographics | Household data: demographics, viewing history, purchase or location data |
| Pricing | CPM on the whole audience | CPM on a defined segment (the sample below assumes $45 against $25 for linear) |
| Main benefit | Mass reach | Less waste, and extra reach among viewers a linear buy reaches lightly |
| Measurement | Panel-based programme ratings | Device-level delivery logs, matched to audiences |
| Main risk | Paying to reach people you do not want | Match errors, thin scale in some segments, duplicate counting across screens |
How much does targeting have to improve to justify the premium?
The price of addressable inventory is higher per thousand impressions. The question is whether targeting wastes enough less to pay for it. The comparison is cost per thousand on-target impressions: CPM ÷ share of impressions that reach the target.
Sample data: your target is 18% of households. A linear buy costs a $25 CPM; an addressable buy costs a $45 CPM and, after identity matching errors, 80% of impressions land on target households.
| Linear | Addressable | |
|---|---|---|
| CPM | $25.00 | $45.00 |
| Share of impressions on target | 18% | 80% |
| Cost per 1,000 on-target impressions | $25 ÷ 0.18 = $138.89 | $45 ÷ 0.80 = $56.25 |
| Impressions needed for 100,000 on-target | 100,000 ÷ 0.18 = 555,556 | 100,000 ÷ 0.80 = 125,000 |
| Cost for 100,000 on-target | $13,889 | $5,625 |
A rule you can reuse: addressable beats linear when its on-target share is above (addressable CPM) ÷ (linear cost per thousand on-target impressions) = $45 ÷ $138.89 = 32.4%. With a 60% match the addressable cost would be $75.00, still lower than linear; below 32.4% it would be worse. The weaker your target segment is on linear TV (the smaller that 18%), the more addressable pays; for a broad product that 80% of households would buy, a linear buy wins: $25 ÷ 0.80 = $31.25 against $45 ÷ 0.80 = $56.25.
What addressable TV is good at, and what it is not
Deloitte's analysis argues that the main value of addressable TV is extending reach, not differentiating the message by household. As linear TV reach falls by 2–3% a year in major markets (per Nielsen, August 2021 data cited by Deloitte), addressable inventory lets a buyer reach viewers who watch little linear TV. Deloitte's survey of 3,974 UK adults (aged 16–75; June–July 2021) found only about one in ten strongly preferred tailored ads, and about two-thirds either did not want customised messages or were indifferent. Take that as a reminder that targeting should lower waste, not raise creepiness.
Deloitte forecast addressable TV revenue of about $7.5 billion globally in 2022, against a global TV advertising market of $153 billion (MAGNA, June 2021). Our arithmetic: 7.5 ÷ 153 = 4.9% of the TV market. It is a niche that is growing, not the default way TV is bought.
How addressable TV is measured
Three measurement questions come up in every campaign. Nielsen frames the delivery side as three rates:
- Active TV rate: the rate at which addressable devices are actively replacing ads across networks.
- Identity match rate: how successfully the ad reaches its intended target audience.
- Ad replacement success rate: the share of ad replacements that execute flawlessly.
Multiply the two losses to see the on-target share in practice. Sample data: if only 90% of replacements succeed and 89% of the households reached match the target segment, then 0.90 × 0.89 = 80% of intended impressions land well, the figure used in the cost table above. The point: a modest-looking 90% success rate and an 89% match rate already cost you a fifth of the spend.
The harder problem is cross-screen duplication. Deloitte notes that unified measurement of TV ads across all screens and services was not available in most markets as of 2022, so one viewer who sees an ad on broadcast and on demand can be counted twice, which the authors call a possible deal killer for major campaigns. When you read a campaign report, ask whether reach is de-duplicated across screens.
Questions to ask before you buy
- What is the segment size in households, and how is the segment built (first-party, third-party, viewing data)?
- What is the identity match rate and the ad replacement success rate, reported for my campaign?
- Is reach de-duplicated against my linear and streaming buys?
- What is the CPM, and what on-target share does that price assume?
- How will I measure a response off the TV screen (see below)?
Measuring the web response to a TV campaign
TapCub does not measure TV exposure, does not match households and does not serve TV ads. It measures what viewers do on your website after they see the ad, which is where an addressable campaign has to prove itself.
- Give the campaign its own landing page or short URL, and put it on the screen or in the voice-over. Add UTM parameters if the URL is used in other places too (UTM).
- Mark the flight dates with a date annotation under Configuration › Annotations, so charts show where the campaign started and ended.
- In Web analytics › Sources, read sessions and conversions for the campaign's UTM tags. Visits from the printed or spoken URL typed in directly show up as direct traffic, so watch direct sessions and the landing page in Web analytics › Pages during the flight, compared with the weeks before.
- Set a goal for the action the ad is asking for (Web analytics › Goals) and compare conversion for the campaign's sessions with your other channels.
- If you split the media buy by segment, give each segment its own landing URL, so you can read each segment's response separately.
This gives you response, not causation: for lift you need a holdout in the media buy, which TapCub does not run (it does not do A/B tests or traffic splits).
Sources
- IAB UK, Addressable TV (ATV) (jargon buster).
- Deloitte Insights, Addressable TV ads: Targeting for reach (December 1, 2021).
- Nielsen, The secret to successful addressable TV (September 2020).
Was this explanation clear?
We use your answers to fix and expand this page.
Couldn’t send. Please try again in a minute.
Thanks for the feedback. We’ll use it to improve this page.You’ve already sent feedback on this page. Thanks.

Product Analyst
Focuses on event analysis, conversion funnels, and retention, using clear metric definitions and structured analysis to explain product performance.