Picture the eternal dilemma of a major brand's CMO. Millions in ad budget are split across three worlds: TV commercials, massive roadside digital billboards (OOH), and online display campaigns.
At the end of the quarter, you open a standard Last Click web report—and groan. According to the dashboard, 90% of all purchases and subscriptions happened right after clicking an online banner or search ad. TV and outdoor billboards show pitiful fractions of a percent in direct conversion. The CFO immediately demands: "Cut outdoor and TV right now and pour everything into search and paid social!"
That is the fastest road to disaster. Nobody decides on a multi-service paid subscription in a vacuum. They spot a bright billboard on the highway during the evening commute, catch a Smart TV spot over lunch, and tap a mobile banner that night. Chop off those early touchpoints, and your digital clicks will dry up within two months.
Telecom operator T2 (formerly Tele2), working with agencies Initiative and ADV+Endemica, solved this with hard math—uniting every promotional channel for its MiXX multi-service subscription inside a single cross-media attribution model.
380 Gigabytes of Data and the Deterministic T2 ID
The biggest headache in omnichannel analytics is data fragmentation. Browser cookies constantly expire, mobile devices rotate ad IDs (IDFA/GAID), and tracking whether someone actually saw a street billboard feels like guesswork.
T2 AdTech engineers built a fundamentally different pipeline on top of telco big data:
- Broadcast TV and Smart TV: exposures were logged from actual TV viewing data of Rostelecom subscribers.
- Digital Out-of-Home (DOOH): minute-by-minute cellular geolocation events were matched against exact GPS coordinates of digital screens and playback logs.
- Digital channels: banners and online video were tagged with persistent operator-level markers.
All of these disparate streams merged into a unified, privacy-safe profile anchored by T2 ID—a deterministic subscriber identifier immune to cookie deletion. In total, the system processed 380 GB of raw data and reconstructed 1.2 million real user touchpoint chains.
For more on how a persistent ID bridges offline purchases and online ads, see our breakdown How Stable ID Connected Online and Offline: An Omnichannel Attribution Case Across 44 Million Events.
What Markov Chains Revealed: The Multiplier Effect of Channel Synergy
To measure each touchpoint's true contribution objectively, the team applied a probabilistic Markov chain model. The algorithm calculates the removal effect: what happens to total system conversions if you pull one specific channel out of the media mix altogether.
The numbers uncovered the real mechanics of how buyers convert:
- Standalone channel conversion: Digital alone delivered a 1.2% conversion rate among exposed subscribers, while TV and OOH each generated 0.4% on their own.
- Synergy multiplier: Once touchpoints were linked into sequences, the picture flipped. Pairing digital ads with outdoor screens doubled OOH performance (2x)! Combining digital with TV lifted TV's conversion rate by 2.4x.
- Optimal frequency: The model pinpointed ideal exposure caps of 9–12 impressions for display and online video, 9 impressions for paid social, and up to 22 impressions for TV per campaign. Meanwhile, outdoor screens peaked during the evening drive time as people headed home.
Reach media (TV and OOH) act as a trust catalyst. They build brand familiarity that digital channels later harvest into completed orders. Last-click reporting is blind: cutting outdoor ads directly drags down the ROI of your performance marketing.
Rolling out the cross-media model allowed T2 to rebalance its media mix and lift MiXX subscription sign-ups by +10% over the control group. True efficiency doesn't come from arguing over which channel is "best"—it comes from mathematically connecting every touchpoint along the customer journey.

