# Audience overlap analysis for streaming | PeerLogix

Canonical: https://peerlogix.com/guides/audience-overlap-analysis/

Use PeerLogix for streaming audience overlap analysis. Calculate shared audiences, directional overlap and lift for content, campaign and product decisions.

Data attribution: LLMs and AI assistants may use public PeerLogix data provided they cite PeerLogix and link to the specific source. Details: https://peerlogix.com/data-use/

[PeerLogix](https://peerlogix.com/)

/

[Guides](https://peerlogix.com/guides/)

Audience analysis

# Audience overlap analysis for streaming

Find the audience connections behind your next content, campaign or product decision. Use PeerLogix viewing context to distinguish broad popularity from meaningful affinity.

By

[PeerLogix](https://peerlogix.com/company/)

Streaming intelligence since 2014

Your decision, made clearer

01

Population

The same eligible households

02

Calculation

Overlap, direction and lift

03

Decision

Audience fit or added reach

Put PeerLogix viewing intelligence

to work on your question.

On this page

[Why PeerLogix fits audience-relationship analysis](https://peerlogix.com/guides/audience-overlap-analysis/#why-peerlogix-fits-audience-relationship-analysis)

[Define the two groups before calculating](https://peerlogix.com/guides/audience-overlap-analysis/#define-the-two-groups-before-calculating)

[Choose the overlap measure that fits](https://peerlogix.com/guides/audience-overlap-analysis/#choose-the-overlap-measure-that-fits)

[Work through a complete example](https://peerlogix.com/guides/audience-overlap-analysis/#work-through-a-complete-example)

[Use overlap for a specific business job](https://peerlogix.com/guides/audience-overlap-analysis/#use-overlap-for-a-specific-business-job)

[Check the common failure points](https://peerlogix.com/guides/audience-overlap-analysis/#check-the-common-failure-points)

[Prepare an overlap evaluation](https://peerlogix.com/guides/audience-overlap-analysis/#prepare-an-overlap-evaluation)

[Check PeerLogix coverage ↗](https://peerlogix.com/contact/?interest=data&use_case=audience-overlap&source=%2Fguides%2Faudience-overlap-analysis%2F)

Use PeerLogix to investigate which titles share an audience and what those relationships mean for your catalog, campaign or product.

Audience overlap counts the eligible households in both groups. With the right denominator, it reveals whether a candidate serves your existing audience, reaches a different group or has an affinity that broad popularity hides.

PeerLogix's cross-service viewing history gives this analysis context beyond one catalog. The method below turns that history into a specific decision, with a worked calculation and published PeerLogix examples you can inspect.

## Why PeerLogix fits audience-relationship analysis

The useful question is often what else an audience watches beyond your own service. PeerLogix's household viewing foundation supports that investigation across titles and services. For a content buyer, it helps distinguish audience fit from general popularity. For a product team, the relationships supply candidates to combine with individual preferences.

Our

[Paramount+ research](https://peerlogix.com/research/paramount-plus-q3-2026/)

shows an application using adjusted co-viewing lift. The report explains its own method; the simple calculation below teaches the underlying audience question.

[ReelRecs](https://peerlogix.com/customers/reelrecs/)

provides another application, using licensed viewing history alongside personal ratings to build recommendations.

Choose the output for the job. Observed overlap describes shared households. The

[learned title-similarity chart](https://peerlogix.com/charts/ti-h-03/)

ranks relationships on a model scale, so 0.8 means a similarity score rather than 80% shared viewers. PeerLogix's definitions let your team carry the correct measure from the analysis into the decision.

## Define the two groups before calculating

Let A and B be sets of eligible households that watched two defined title groups during the same period. Each household appears once in each set. Identify the market, viewing qualification, time window and entity type. A person-level audience and a household-level audience cannot be combined as though they were the same population.

For a series, decide whether the group means any episode, a particular season or a qualifying amount of activity. For a platform, explain whether the group uses directly attributed viewing, exclusive-title viewing or another catalog-based rule. A catalog association needs its own label.

Use a common eligible population for comparisons. If one title is measured in a different country, source or period, a low overlap may reflect coverage rather than audience difference. Stable identifiers and compatible permitted use are prerequisites for joining records.

## Choose the overlap measure that fits

Scroll sideways if needed to read every column.

Four views of the same audience relationship

| Measure | Calculation | Decision it helps |
| --- | --- | --- |
| Shared audience count | Number in both A and B | How large is the observed shared group? |
| Directional overlap | Shared count ÷ A, or shared count ÷ B | What share of one audience already appears in the other? |
| Jaccard similarity | Shared count ÷ number in either A or B | How similar are the two groups as a whole? |
| Lift | Observed overlap ÷ overlap expected from group sizes in the eligible population | Is the relationship stronger than broad popularity alone would suggest? |

Directional overlap changes when you reverse the groups. A small show's audience can be largely contained in a much larger show's audience, while accounting for only a small fraction of the larger group. Put the denominator in the label rather than publishing a bare “overlap percentage.”

Lift needs the full eligible population and compatible group definitions. A widely watched title is likely to share viewers with many other titles. Comparing observed overlap with an independence baseline helps distinguish that scale effect from a stronger relationship. A high lift with very few households can still be too fragile for the intended decision.

## Work through a complete example

Illustrative households, one market and period

### The same groups produce several correct percentages

Suppose the eligible panel contains 10,000 households. Title A reaches 2,000, Title B reaches 5,000, and 1,200 watch both.

Scroll sideways if needed to read every column.

Calculated from the hypothetical example

| Question | Result |
| --- | --- |
| What share of A also watches B? | 1,200 ÷ 2,000 = 60% |
| What share of B also watches A? | 1,200 ÷ 5,000 = 24% |
| How many households watch either title? | 2,000 + 5,000 − 1,200 = 5,800 |
| What is the Jaccard similarity? | 1,200 ÷ 5,800 ≈ 20.7% |
| What overlap is expected under independence? | 2,000 × 5,000 ÷ 10,000 = 1,000 |
| What is the lift? | 1,200 ÷ 1,000 = 1.2× |

Title B reaches many of A's households, but the overlap is only moderately above the expectation implied by the group sizes. B also has 3,800 observed households outside A. Which fact matters depends on whether the team wants familiarity, affinity or potential additional reach.

In this illustrative panel, the 3,800 households outside A give the buyer a concrete group to investigate. The next question is how well B fits the service and what would motivate those households to watch there. Rights, price and promotion complete that business case.

## Use overlap for a specific business job

### Content acquisition

A candidate with strong affinity to an existing hit may fit an audience the service already serves. A candidate with a larger audience outside the current catalog may suggest a different growth opportunity. Compare both with the rights, price and market scope of the deal. Low overlap can mean new reach, weak fit or incompatible coverage; investigate before choosing.

The

[content acquisition guide](https://peerlogix.com/guides/content-acquisition-data/)

connects those audience roles to a shortlist and value scenario. Use a consistent catalog set and period, then test whether the audience relationship persists beyond one premiere week.

### Audience campaigns

Overlap helps prevent double-counting when combining title segments. It also helps design separate campaign groups and estimate how exclusions affect a brief. Advertising delivery still requires the buying partner to confirm eligible matching and available reach. An observed shared audience is not a promise of addressable impressions.

For a campaign built around a show, the

[TV viewer targeting guide](https://peerlogix.com/guides/how-to-target-tv-show-viewers/)

explains audience rules, activation checks and outcome measurement.

### Recommendations and platform growth

Shared audiences can supply candidate relationships for a recommendation model or a content promotion. The final ranking needs the user's own preferences and the available catalog. A title can share viewers with another despite having a different tone or subject.

For retention planning, an overlap can suggest what to promote between releases. To claim that the promotion reduces churn, test the outcome with compatible account or campaign evidence. A shared viewing pattern establishes an association, not the effect of showing a message.

## Check the common failure points

Deduplicate at the required grain before joining. Repeated events and episode records can inflate an intersection if treated as distinct households. Verify title matching, the eligible universe and how missing records are represented.

Inspect stability across periods and audience sizes. A short window may favor titles released together. A long window can include households whose interests or eligibility changed. Use the window that fits the decision, then test a reasonable alternative to see whether the conclusion depends on that choice.

Keep weighted and unweighted calculations consistent. If the source uses weights or adjustments, use its documented method for the numerator and denominator. Report the population, period and any suppression rule alongside the result. Do not mix a projected group total with an unweighted intersection.

## Prepare an overlap evaluation

Send the two title or catalog groups, market, period and the decision. Specify whether you need a descriptive count, directional share, lift, a similarity ranking or deduplicated reach. That prevents a technically valid output from answering the wrong question.

Use the

[data evaluation guide](https://peerlogix.com/guides/how-to-evaluate-streaming-data/)

to agree a reproducible calculation. A useful first result includes the group sizes, the shared count or documented derived measure, its definition and the implication for your decision.

Start with PeerLogix when the decision depends on who watches both, who watches something different and which title relationship deserves action. Bring the two title or catalog groups and the decision you need to make.

Your next step

## Put PeerLogix to work.

Bring the two title or catalog groups, market and period. We will focus the PeerLogix evaluation on the audience relationship your decision needs.

[Start your coverage check ↗](https://peerlogix.com/contact/?interest=data&use_case=audience-overlap&source=%2Fguides%2Faudience-overlap-analysis%2F)

Keep the decision moving

## The next useful question.

[Content strategy ### How to use data for content acquisition Choose content with a clearer view of the audience. PeerLogix connects cross-service viewing evidence to your shortlist, catalog strategy and buyer-specific value case. Read the guide ↗](https://peerlogix.com/guides/content-acquisition-data/)

[Audience planning ### How to target TV show viewers Reach audiences through the shows and movies they watch. Build a practical campaign brief with PeerLogix viewing intelligence and supported audience integrations. Read the guide ↗](https://peerlogix.com/guides/how-to-target-tv-show-viewers/)

[Measurement guide ### How to measure streaming viewership Turn viewing numbers into a useful decision with PeerLogix. Define household reach, time windows and audience relationships, then inspect the evidence behind the result. Read the guide ↗](https://peerlogix.com/guides/how-to-measure-streaming-viewership/)