Player and app
- Event schema shared by web, iOS, Android and smart TV players
- CMCD keys sent with media requests
- Heartbeat, error, ad boundary and progress events
- Session ID and content ID on every event
Media and Entertainment Analytics
Media and entertainment analytics development
Your player, ad server, billing system and rights database each report a different number of plays. Agnotic Technologies builds media and entertainment analytics that start from the player event, agree one definition for every metric and reconcile it with the systems that pay and bill on it.
Player telemetryTitle engagementAd reconciliationRoyalty reporting
Trusted by global innovators
What is media and entertainment analytics?
Media and entertainment analytics is the data system that turns player events, ad delivery records, subscriptions and rights data into trusted numbers for streaming services, publishers and studios: playback quality, engagement by title, churn signals, ad delivery and royalty-bearing plays. Agnotic Technologies designs and builds these systems for media companies.
Most of the work is agreeing what each number means. A play, a view and a completed stream can each be counted three ways, and finance, editorial and partners often use different ones.
Customer profiles and identity matching are covered on our customer analytics page, and streaming pipeline engineering on our real-time data warehouse page. This page covers what is specific to media.
Metric register
Which media metrics need one owner and one definition?
We build this register with your product, finance and partner teams before any dashboard. Every metric gets an event source, a definition owner and a system it is checked against.
| Metric | Event source | Definition owner | Reconciled against |
|---|---|---|---|
| Concurrent viewers | Player heartbeat events per session | Product or streaming operations | CDN request logs for the same window |
| Rebuffering ratio | Player stall and resume events, CMCD buffer starvation flags | Streaming operations | CDN delivery logs by region and device |
| Start failures | Player error events before the first frame | Streaming operations | Entitlement and licence server logs |
| Completion rate | Progress events against title duration | Editorial or content strategy | Title metadata from the content catalog |
| Ad impressions served versus billed | VAST impression and quartile tracking events | Ad sales and finance | Ad server reports and the advertiser's invoice |
| Royalty-bearing plays | Plays that pass the qualifying rule in each licence | Finance and rights management | Rights and licensing windows, then partner statements |
| Subscriber churn signal | Viewing gaps, failed payments and cancellation events | Subscription product team | Billing system records |
When two systems disagree, the register says which one wins for which purpose. Finance may report from the ad server while product reports from player events.
Data flow
How does player telemetry reach a trusted report?
Four layers, each owned and tested separately. The streaming engines underneath are covered on our real-time data warehouse page.
Collection
- Beacon endpoint with schema validation
- CDN log ingestion
- Ad server and SSAI event feeds
- Late and duplicate event handling
Modelling
- Sessions built from raw events
- Title, episode and season dimensions from the catalog
- Qualifying play rules per licence
- Bot and test traffic filtered out
Reporting
- Live quality of experience views for operations
- Engagement by title for editorial
- Ad delivery and royalty reports for finance
- Partner statements with the rule that produced each number
Sources
Which systems does media analytics connect to?
The sources we connect most often. The register decides which of them is the reference for each metric.
Playback
- Web, mobile and smart TV players
- CDN request and delivery logs
- Licence and entitlement servers
Content
- Content management and catalog metadata
- Rights, territories and licensing windows
- Release and scheduling calendars
Subscribers
- Billing and subscription management
- Entitlements and plan changes
- Support tickets about playback
Advertising
- Ad server decisions and tracking events
- Server-side ad insertion logs
- Ad verification and measurement partners
Failure points
Where do media metrics disagree, and how do we fix it?
The disagreements we find most often when we audit an existing media data setup.
Challenge
Player events and CDN logs report different audience sizes.
Agnotic approach
Sessions are built on a shared session ID, and the register states which source is the reference for each use.
Challenge
Smart TV apps send events late or in batches.
Agnotic approach
The pipeline accepts late events within an agreed window and restates affected reports instead of dropping them.
Challenge
Ad impressions in the ad server do not match the invoice.
Agnotic approach
Impression and quartile events are joined to ad server records per ad break, so each gap is traced to a cause.
Challenge
Royalty reports count plays that do not meet the licence rule.
Agnotic approach
Each licence's qualifying play rule is coded once, versioned and applied to every report that pays partners.
Challenge
A title is counted under different names in different regions.
Agnotic approach
Every event carries a catalog content ID, and titles are resolved from the catalog, not from free text.
Challenge
Test devices and bots inflate viewing figures.
Agnotic approach
Known test devices, internal accounts and automated traffic are flagged at collection and excluded by default.
What do we build for media and entertainment analytics?
Modules we design and build, sized to the questions your teams need answered.
Playback quality monitoring
Start failures, rebuffering and bitrate changes by device, region and CDN, close to real time for operations teams.
Engagement by title
Starts, completion and viewing time by title, episode and season, joined to catalog metadata for editorial teams.
Subscriber churn signals
Viewing gaps, payment failures and plan changes combined into signals your retention team can act on.
Ad delivery reconciliation
Impressions served, verified and billed, matched per campaign and ad break, with every gap explained.
Rights and royalty reporting
Royalty-bearing plays by licence, territory and window, with partner statements that show the rule applied.
Event schema and SDK work
One event schema across players, with instrumentation added to your web, mobile and smart TV apps.
Viewing privacy
Which privacy rules shape viewing data?
Viewing history is sensitive data. Your legal team decides how each rule applies; we build the controls into the data model.
01
Video Privacy Protection Act
Example: Sharing which titles a named subscriber watched with an ad partner or a social platform
Approach: Viewing records are kept apart from identity data, and any sharing that names a person runs through a consent record your legal team has approved.
02
COPPA
Example: A kids profile, or a service directed to children under 13
Approach: Child profiles collect only the events the service needs, third-party tags are off, and consent status is checked before any personal information is collected.
03
State privacy laws and opt-outs
Example: A viewer opts out of the sale or sharing of personal data
Approach: Opt-out and consent signals are stored with each profile and applied in every export, audience segment and partner feed.
04
Partner and licence terms
Example: A studio contract limits what viewing data can be shared or how long it is kept
Approach: Contract terms become data rules: what each partner may receive, at what level of detail and for how long.
Who it is for
Who builds media analytics with us?
Media businesses whose numbers drive payments, programming and advertising decisions.
Streaming services
Subscription and ad-supported services that need playback quality, engagement and churn data from one model.
Publishers with video and audio
News, sports and lifestyle publishers reconciling ad delivery across their own players and partners.
Studios and distributors
Rights holders who need royalty-bearing plays reported consistently across every platform that licenses their titles.
Live sports and events
Live streams where concurrent viewers and start failures must be visible while the event is running.
Media entertainment analytics for startups
New streaming and creator platforms that want the event schema and metric register right before the audience grows.
Scope and cost
What drives the cost of media and entertainment analytics?
We do not publish prices. These factors set the scope, which we confirm on the scoping call.
Player platforms
How many players and app platforms need instrumentation, and whether they share an SDK.
Sources to reconcile
CDNs, ad servers, billing, rights and catalog systems, and how clean their exports are.
Latency
Which reports must be live during an event and which can be refreshed daily.
Licence rules
How many partner agreements define qualifying plays, territories and windows differently.
History
Whether past data must be backfilled and restated under the new definitions.
Privacy controls
Kids profiles, consent records, opt-outs and partner data sharing rules.
Engagement
How does a media analytics engagement run?
From metric register to reports that finance, editorial and partners accept.
1.
Scoping call and data audit
We review your event data, systems and current reports, and list where numbers already disagree.
Audit2.
Metric register and schema
We agree definitions, owners and reference systems with your teams, then fix the event schema.
Define3.
Fixed-scope first release
We build the first reports end to end, usually playback quality or ad reconciliation, and check them against the reference systems.
Build4.
Extend and hand over
We add royalty, churn and partner reporting, then hand the code and runbooks to your team or keep supporting it.
ExtendStandards
Standards and rules we build media analytics to
Public specifications and laws behind the event schema and the privacy controls.
CTA-5004IAB VAST18 U.S.C. 271016 CFR 312
Facts and sources
Facts behind media and entertainment analytics
Specifications and laws that shape player telemetry, ad tracking and viewing privacy, each linked to its primary source.
CTA-5004-B, Common Media Client Data, lets a media player attach data to each media object request as query arguments or request headers, and adds an event mode that reports on triggers such as play state changes, errors, ad boundaries and a heartbeat interval.
Source: Web Application Video Ecosystem, Common Media Client Data (CTA-5004-B), CTA WAVEIAB Tech Lab's Video Ad Serving Template (VAST) is a universal XML schema for serving ads to digital video players, and VAST 4.1 (November 2018) added ad verification with Open Measurement.
Source: Digital Video Ad Serving Template (VAST), IAB Tech LabThe Video Privacy Protection Act treats information that identifies a person as having requested or obtained specific video materials as personally identifiable information, and allows disclosure with informed, written consent given in a separate form for up to 2 years.
Source: 18 U.S.C. 2710 (govinfo.gov)The COPPA Rule defines a child as an individual under the age of 13 and requires an operator to obtain verifiable parental consent before any collection, use or disclosure of personal information from children.
Source: 16 CFR Part 312, sections 312.2 and 312.5 (eCFR)
Frequently Asked Questions
There is no fixed price. Cost follows the number of player platforms, the systems to reconcile, how live the reports must be, how many licence rules apply and whether history must be restated. You get a written proposal after the scoping call.
It depends on the sources and reports in scope, so we agree a fixed first release on the scoping call. That release covers one report end to end, so your teams can check it against their current numbers early.
Yes. We add instrumentation to the players you run and read from your current ad server, CDN, billing and rights systems through their exports and APIs. We do not ask you to replace them.
Packaged player analytics tools are often the right choice for playback quality on their own. A custom build pays off when you must reconcile plays with ad billing, royalties and subscriptions, which packaged tools rarely model your way. Many clients use both.
Yes. Live events need concurrent viewer and start failure views that refresh within the event, while on-demand reporting can run on longer windows. We design both on the same event schema.
Yes. A common first release is the metric register, the shared event schema and one reconciled report, such as ad impressions served versus billed. Later releases add royalty, churn and partner reporting.
Engineers may use AI coding assistants for routine code, but metric definitions, licence rules and privacy controls are written and reviewed by an engineer and tested against reference data. Nothing reaches your main branch without code review.