AdBridge contextual ads.
Every ad break sold
on what just aired.
Contextual advertising places ads according to the programme a viewer is watching. Point Media Tech's AdBridge is a contextual advertising system for CTV (connected TV), FAST and OTT. It turns the content descriptions MAM produces for every programme segment into IAB content categories and brand-safety ratings, and hands them to your ad stack before every break. Broadcasters and OTT platforms can then package the same inventory by content, while advertisers can buy the break after a running show for running shoes and stay clear of content that does not suit their brand. No viewer data is involved.
IABThe industry-standard content taxonomy
4 leversContext premium, unblocked content, sponsorship, direct sales
Drop-inWorks with your SSAI, ad server and DSPs
No PIITargeting by content
Every segment on the timeline
is a label buyers already read.
This is the context timeline for one episode of a running-coaching show. At the top, AI recognises the stadium and the track markings and reads the heart-rate caption on the broadcast graphics as text. In the middle are the segments MAM cut on scene boundaries, each already carrying IAB tags. The lanes below show categories and brand-safety ratings for the whole episode, with the 90-second window before each break in orange. The right panel details the selected break: IAB 3.x categories with compatible 1.0 codes, atmosphere, brand safety, and the products it suits, from running shoes to sports drinks and wearables.
Where the revenue
comes from.
Contextual advertising does not create new inventory; it makes the inventory you have sell better. AdBridge adds revenue in four places.
| Context premium | Breaks carry IAB content categories, so buyers know what aired before them and can pay more for relevant context, such as a running-shoe brand buying the break after a running show. |
|---|---|
| Content that was unsellable, sold | Brand safety is rated per segment, so buyers no longer block whole genres. Weather, food and lifestyle segments inside news can be sold to brands that used to avoid news altogether. |
| Sponsorship and placement premium | Competitive separation guarantees no same-category ad follows a placed brand, and brand exposure reports let the placement close with proof, giving sponsors a reason to pay more. |
| New products for direct sales | Inventory packaged by context, such as a "fitness and exercise" package, that the sales team can pitch to advertisers without waiting for programmatic buyers. |
Results depend on whether buyers bid on context, on your content mix and on the number of buyers. AdBridge has holdout measurement built in: keep a share of breaks without context, verify the effect on your own eCPM and fill rate, then set pricing and scale up.
Why CTV needs contextual advertising:
a break is worth what aired before it.
Most CTV inventory is still sold by channel or by show. Buyers cannot see what aired before the break, so they fall back on demographic guesses or bid low; cookie loss and shared living-room viewing make demographics less reliable still. Brands that cannot see the content block whole genres. AdBridge turns the content before each break into standardised context signals, so the same inventory can be packaged by content and priced by context.
Point Media Tech has already proven content-timed delivery in live sport. During a major sports channel's 2026 WBC coverage, ScoreCast-AD delivered 2.4× the contextual precision of conventional inning-break ads (WBC sports AI advertising case). AdBridge brings the same approach to every genre and to the whole content library.
The break carries a channel name and nothing else
Ad requests carry only the channel and show title. Buyers estimate the audience from demographics. Brand safety means blocking entire genres. A sponsor's competitor can air right after the sponsor's placement. The platform cannot show which content deserves a higher price.
Every break carries the scene context before it
Breaks carry IAB content categories, atmosphere and brand-safety ratings. Buyers bid on what is on screen. Safety is rated per segment, so no genre needs a blanket block. Sponsors' competitors are excluded automatically. A holdout comparison shows what context signals do to revenue.
Six capabilities,
from scene tags to the ad request.
All six capabilities share the same MAM video structure, and each can be switched on as your ad stack and sales model require.
IAB taxonomy mapping · Taxonomy
MAM cuts each title into segments on scene boundaries, and every segment carries start and end timecodes, a face count, a scene description and on-screen text. AdBridge maps each segment to IAB Content Taxonomy 3.x, outputs compatible 1.0 codes where needed, and extracts brands, objects, actions and atmosphere.
- IAB 3.x categories, 1.0 compatible
- Cut on scenes, not fixed time windows
- On-screen text read as context
- Local vocabulary for Asian programming
Brand safety & suitability · Brand Safety
Rates every segment against the Brand Safety Floor + Suitability Framework and flags alcohol, tobacco, violence, adult content and gambling, so buyers choose whether to avoid or prefer it. Statutory warnings on screen count as evidence: when a "no drinking under 18" caption appears, the segment is flagged for alcohol.
- Floor, High, Medium and Low tiers
- Segment-level sensitive flags
- On-screen warnings as evidence for local rules
- Conservative defaults, adjustable per platform
Intelligent break placement · Placement
For library titles with no break points, suggests where breaks belong from MAM scene boundaries and the emotional arc: at natural scene ends or just after a cliffhanger, never mid-dialogue or mid-climax. Ad Insertion then writes frame-accurate SCTE-35 cues.
- Scene boundaries and emotional arc
- Avoids dialogue and climaxes
- Scores how sellable each pre-break context is
- Aligned to HLS and DASH segment boundaries
Pre-break context · Context
An ad lands on whatever the viewer just watched. AdBridge uses the 30 to 90 seconds before the break as the primary signal, and the AI summary of the whole title and the series as secondary signals, weighting the moments closest to the break highest. Context for linear schedules and VOD is computed before air.
- Adjustable pre-break window
- The new scene wins after a scene change
- Primary and secondary signal tiers
- Computed ahead of air
Competitive separation & exposure reports · Separation
Once MAM recognises a brand in the programme, later breaks automatically exclude competitors in the same category, protecting sponsors and placement clients. The same data counts how many seconds each brand is on screen across the library, and in what context, for placement reporting.
- Category-level competitor exclusion
- Sponsor and placement protection
- Brand screen time and context
- Exportable campaign reports
Lift measurement · Lift
Holds out a random share of breaks without context as a control group, then compares eCPM, fill rate, win rate and completion rate by channel, genre and context category, so pricing and renewals rest on your own numbers.
- Holdout control group
- eCPM and fill-rate comparison
- Breakdown by context category
- Report export and API
Where AdBridge sits in the ad stack:
between content and buyers.
AdBridge reads the video structure from MAM and the playout schedule, then computes the context for every break. Ad Insertion marks where the break is, SSAI carries the context into the ad request, and buyers bid on it. The whole chain shares one asset structure, with no third-party integration layer.
Broadcaster · OTT platformSells content and inventory
- 01MAMCuts titles into scenes with descriptions and on-screen text
- 02AdBridgeTurns them into IAB categories, brand-safety ratings and a context ID
- 03Playout · Ad InsertionAirs the programme and writes SCTE-35 plus the context ID at each break
AStream and SCTE-35 breaks, with the context ID riding each break
Streaming & deliveryStitches ads into the stream
- 04SSAIReads the break and its context, requests ads and stitches the winners in
- 05The viewer’s screenCTV, FAST channels, OTT apps
B · CAd requests carry the context: key-values or OpenRTB 2.6
Ads chosen on context return to SSAI
Advertiser · agencyBuys breaks by content
- 06Ad ServerDirect-sold campaigns scheduled on context key-values
- 07SSP · DSPProgrammatic buyers bid on or exclude IAB categories
Three signal paths, no DSP integration required
The same break context can leave on any of three paths. Pick the one that fits your ad stack, or run them side by side.
| A · Inside SCTE-35 | A context ID is written into the URI-type segmentation_upid of the SCTE-35 cue and travels with the stream; the downstream SSAI reads it into the ad request. Suited to platforms on a third-party SSAI or packager, with no extra API integration. |
|---|---|
| B · Ad server key-values | SSAI appends context key-values to the VAST ad request, and the ad server targets on them. Suited to broadcasters and platforms that sell mostly direct, so direct campaigns can be scheduled by context. |
| C · OpenRTB 2.6 | Context goes into standard bid request fields: content.cat with cattax, keywords and data.segment, readable by buyers without custom integration. Suited to programmatic trading through SSPs and DSPs. |
Operator view: what one break actually sent
Open any break to see the payload on all three paths and the order in which that break was delivered, from cue-out to a filled pod.
Engineering specifications.
AdBridge takes the MAM video structure as input and outputs context signals in three formats: SCTE-35, VAST key-values and OpenRTB 2.6. Full specifications and integration details are available from our solutions engineers.
| Input | Point Media Tech MAM video structure: scene segments (start and end timecodes, face count, scene description, on-screen text) · AI summary of each title · playout schedule · SCTE-35 break data |
|---|---|
| IAB Tech Lab standards | IAB Content Taxonomy 3.x (1.0 compatible) · OpenRTB 2.6 · VAST ad requests |
| Content categories | IAB Content Taxonomy 3.x · 1.0 compatible codes · local context sub-labels · brand, object, action and atmosphere segments |
| Brand safety | Brand Safety Floor + Suitability Framework · Floor / High / Medium / Low · sensitive flags for alcohol, tobacco, violence, adult content and gambling · on-screen warnings as evidence |
| Break alignment | SCTE-35 splice_insert · time_signal · pre-break window 30–90 s (adjustable) · aligned to HLS / DASH segment boundaries |
| Signal output | SCTE-35 segmentation_upid (URI) · VAST ad request key-values · OpenRTB 2.6 content.cat · cattax · keywords · data.segment |
| Rules & reporting | Competitive separation rules · brand exposure reports · holdout lift reports |
| Integration | REST API · webhooks · shared asset structure with Ad Insertion, SSAI and Commercial Detection |
| Deployment & licensing | Cloud or on-premises · licensed by channel count and content volume · software licence plus annual maintenance |
Three steps from content library to the first contextual break.
Book a demo
Submit the contact form. Sales responds within one business day to schedule a technical assessment.
Library and ad stack assessment
A solutions engineer reviews your content library, MAM metadata, SSAI and ad server, and selects the output path for context signals.
Pilot and roll-out
Pilot on one channel or one library, compare revenue on context-enriched breaks against the holdout group, then extend to every channel.
Frequently asked questions.
What is CTV contextual advertising?
CTV contextual advertising places ads on connected TV, FAST and OTT streams according to the programme the viewer is watching. The system first identifies what aired before the ad break, such as a running session, a food show or a travel segment, then tells buyers in IAB content categories, so relevant ads land in the right break. It needs no viewer data, which makes it more reliable than demographic targeting where several people share one living-room screen.
Does AdBridge increase ad revenue?
It adds revenue in four places. Breaks that carry IAB categories let buyers pay more for relevant context. Segment-level brand safety makes content that was blocked by genre sellable again. Competitive separation and brand exposure reports make sponsorship and placement worth more. And context-based inventory packages give the direct-sales team a new product to sell. How much depends on whether buyers bid on context, on your content mix and on the number of buyers, which is why AdBridge has holdout measurement built in: keep a share of breaks without context, verify the effect on your own eCPM and fill rate, then set pricing and scale up.
What is the IAB Content Taxonomy, and why does contextual advertising use it?
The IAB Content Taxonomy is the content classification standard maintained by IAB Tech Lab. It describes content topics, such as travel, food, automotive and sport, with one shared set of codes. DSPs, SSPs and ad servers all read inventory through it, so context expressed in IAB categories lets buyers bid on or exclude a category without custom mapping. AdBridge uses Content Taxonomy 3.x and declares the version in the OpenRTB 2.6 cattax field; where a buyer still reads the older taxonomy, it can also output compatible 1.0 codes, such as IAB7-1 for exercise.
How is contextual targeting different from demographic targeting?
Demographic targeting estimates who is watching from data such as age and gender. Contextual advertising places ads according to what is on screen at that moment. CTV is often watched by several people in one living room, so account data may not describe the person in front of the screen; content signals stay reliable and need no viewer data. The two can be combined, and AdBridge supplies the content side.
Can I use AdBridge without a DSP integration?
Yes. AdBridge can write context into ad server key-values, so directly sold campaigns target by context, and it can run competitive separation and brand exposure reporting on its own. When you move to programmatic, the same context goes to SSPs and DSPs in standard OpenRTB 2.6 fields.
How do AdBridge, Ad Insertion and SSAI divide the work?
Ad Insertion decides where the break is and marks the SCTE-35 cue. AdBridge decides what context the break carries. SSAI takes that context into the ad request and stitches the ads into the stream. All three share one asset structure. If you already run a third-party SSAI, the context can travel to it inside the SCTE-35 UPID.
How is brand safety rated?
AdBridge rates every segment against the Brand Safety Floor + Suitability Framework: Floor, High, Medium or Low, and Floor segments never carry ads. Sensitive flags mark alcohol, tobacco, violence and similar content so buyers decide whether to avoid or prefer it. The framework originated with GARM, which ceased operations in 2024; it remains the industry baseline for brand safety.
Where does AdBridge get its context data?
From the video structure in MAM. MAM cuts each title into segments on scene boundaries, and every segment carries start and end timecodes, a face count, a scene description and on-screen text; the whole title also gets an AI-written summary. AdBridge turns each segment description into IAB categories, atmosphere and a brand-safety rating. On-screen text is read too: a broadcast caption such as "Intervals 400 m × 6" adds a running-training context to that segment.
How do I start evaluating AdBridge?
Submit the contact form and our sales team will respond within one business day to schedule an assessment of your content library, MAM metadata and existing SSAI and ad server. Licensing is quoted by channel count and content volume.
Price CTV inventory
by what is on screen.
Book a demo to see AdBridge turn one episode's segment structure into context signals for every break.