Automotive revenue analytics is what separates an OEM that knows which connected services pay for themselves from one that finds out at renewal, three years too late.
Most manufacturers have solved the connectivity problem. Vehicles ship connected, services activate, data flows. The harder question is commercial: which of those services earn their place on the bill of materials, which markets convert, and which features are being paid for by nobody who uses them.
That question is answerable. The data already exists inside the vehicle population. What it needs is measurement built for revenue decisions rather than network diagnostics.
Where automotive revenue analytics finds new income
Subscription models analysis
General Motors offers the clearest public picture of what mature connected-service monetisation looks like. The company recorded $2.7bn in realised software and services revenue in 2025, alongside $5.4bn in deferred revenue, and projects deferred revenue to rise to $7.5bn in 2026. Super Cruise alone generated roughly $234m in 2025, with GM guiding to almost $400m this year. Around 70 cents of every subscription dollar stays with the business, a margin profile vehicle sales can’t match.
The number that matters most, though, is the attach rate. GM reports that 30% to 40% of drivers continue paying once the three-year included period ends. That figure is a measurement problem before it is a pricing problem. Knowing which cohorts renew, and what those cohorts did in months one to 36, tells you what to change for the next model year.
The wider market is moving the other way. S&P Global Mobility’s 2025 consumer research found willingness to pay for connected services fell to 68%, down from 86% a year earlier. A fifth of non-subscribers said they were never made aware the services existed. We know customers resist recurring fees, but a fifth of that gap is a merchandising issue showing up in activation data, and it is fixable without changing the price.
Usage-based insights
Feature-level usage is where the unprofitable parts of a connected offer become visible. A streaming bundle with high activation and low weekly sessions behaves differently at renewal than one with lower activation and daily use, and each needs different treatment.
Usage data does the same work on the cost side. Data consumption patterns across a live vehicle population show where connectivity plans are oversized, where roaming arrangements cost more than the original business case assumed, and which markets consume far more in-vehicle data than anyone forecast.
The same telemetry supports pricing outside the vehicle. Insurers are already using connected vehicle data for commercial auto risk pricing, which turns an operational data stream into a business-to-business revenue line.
The analytics tools that turn data into decisions
Market trend forecasting
Berg Insight’s OEM telematics research puts embedded connectivity in 83% of cars sold worldwide during 2025, up from 79% a year earlier, and forecasts 91% by 2030. The active installed base of OEM telematics connections grows from 338 million to 583.1 million across the same period. Average data consumption per car is forecast to reach 10.1 GB by 2030, up from around 6 GB in 2025.
Those numbers size the opportunity well enough. The planning value sits underneath them, in the variance the global average conceals. Berg Insight puts the North American attach rate at roughly 94% in 2025, 11 points above the worldwide figure, while several regions sit well below it. A service that stalls in one market can convert strongly in another for reasons that surface in usage data months before they surface in revenue.
Forecasting from your own fleet beats forecasting from industry averages. Early adoption signals from a live vehicle population give product teams a window to act while the model year is still open.
Adoption curves for the same service diverge sharply between markets, and a feature that stalls in Germany may convert well in Spain for reasons that show up in usage data months before they show up in revenue. Early adoption signals from a live vehicle population give product teams a window to act while the model year is still open.
Customer behaviour analytics
Generic packaging produces generic conversion. Persona-level analysis on a per-vehicle, per-fleet, and per-region basis shows which buyer groups gravitate to which services, letting pricing and bundling follow observed behaviour rather than assumption.
Behavioural data also gives you warning. Declining session frequency, abandoned activations, and unopened OTA update prompts all precede churn. Catching those signals during the included period, rather than reading them in a renewal report afterwards, is where retention is won.
- Activation rate by market, trim and dealer group
- Session frequency and duration by feature and cohort
- Conversion at the end of the included period, split by prior usage
- Data consumption against plan allocation by region and network
- Update acceptance rates as a proxy for engagement
Strategies for OEMs before September 2026
The regulatory clock changes the calculation. The EU Data Act has applied since 12 September 2025, giving vehicle users the right to access data their vehicles generate and to nominate a third party to receive it. From 12 September 2026, design obligations follow: connected products placed on the EU market must be built so that users can access that data directly.
The Commission’s accompanying vehicle data guidance draws a line that matters commercially. Raw and pre-processed data sit in scope. Inferred and derived data, including driver scoring and predictive analytics, sit outside it. Manufacturers must give users access to available vehicle data on the same terms they give their own networks and authorised dealers.
Read that as a strategy brief rather than a compliance burden. If raw signals become broadly accessible, the defensible commercial asset is the layer above them: the models, the segmentation and the derived intelligence built on top. OEMs that treat analytics as an operational cost centre will find their differentiation thinning. Those that treat it as product will still own something.
Practical priorities for the next model cycle:
- Build a data inventory that classifies every signal as raw, pre-processed or derived, so scope questions are answered before an access request arrives
- Instrument the included period, not just the renewal point, so churn is visible while it can still be influenced
- Unify network, service and user data in one place, because fragmented sources produce contradictory answers to the same commercial question
- Measure connectivity cost per vehicle per market alongside service revenue, so margin is visible at the level decisions are made
Cubic3 built Explore3 vehicle analytics to do exactly this work, bringing network performance, service adoption and user engagement into a single view. It sits alongside DriverConnect3, where activation and in-car commerce data close the loop between what a driver does and what it earns.
Find out what your vehicle data already knows
Every connected vehicle you have on the road is generating the answer to your next pricing decision. See what Explore3 surfaces from your fleet.
