Monetizing data has long been a goal of entrepreneurial CSPs (communication service providers). For example, operators like A1 in Austria, and O2 in the UK, have developed mobility insights units, enabling them to form new revenue streams by providing demographic data to governments and businesses.
Today, connecting billions of sensors and devices is a reality thanks to 5G. As a result, the volume of data generated across 5G is set to explode, with estimates of up to a mind-blowing 79.4 zettabytes by 2025, according to IDC.
That’s a lot of data with a lot of potential value, and CSPs are looking to the 3GPP Network Data Analytics Function (NWDAF) and ONAP’s Data Collection, Analytics and Events (DCAE) framework to unlock some of it. With this standardized approach to collecting and analyzing network data, NWDAF and DCAE allow CSPs to manage, automate, and optimize their 5G network operations much more efficiently.
But could it also do more than that? In the same way that Amazon turned its internal compute infrastructure into the revenue powerhouse that is AWS, could CSPs turn their internal 5G network analytics into an external money-spinner? And if so, where might the opportunities lie?
Ensuring quality of experience in cloud gaming
One potential area is cloud gaming, which requires low and consistent latency to ensure an enjoyable – and fair – user experience. If a player seated in a café shoots the in-game character of a player riding on a bus, for example, that interaction has to be instantaneous for both players. And if a building in the game world collapses, it has to collapse at exactly the same time for every player.
A CSP can optimize the experience for players on their network by using real-time analytics with predictive modelling to spot when latency might be about to increase. That prediction can then trigger automatic network optimizations. As esports become more competitive and professionalized, players may be prepared to pay a premium for that glitch-free experience.
But what if the players are on different networks? If you're a game provider hosting the game, it would be nice to make sure that everyone has the same level playing field. A way of ensuring common latency across networks, even if it’s within specific games or groups of gamers, would be quite cool!
That suggests an opportunity for CSPs to share real-time network performance predictions, so the gaming experience can be consistently managed across networks. It’s possible to imagine that being delivered as a value-added service to cloud gaming companies, as well as to individual gamers as part of a gaming-themed 5G plan.
Unlocking the potential of Industry 4.0
While cloud gaming could be an interesting use case on the consumer side, more and bigger opportunities are likely to lie in the enterprise. Network performance and reliability will be critical considerations in a huge range of 5G use cases, from autonomous vehicles to AR-mediated remote operations.
An inspection drone, for example, will rely on 5G connectivity and low latency not just to livestream video, but also to calculate its position, understand its surroundings, avoid collisions, and receive instructions from a remote operator. Real-time and predictive insight into network performance will be essential to ensuring the drone can operate safely and as expected.
It’s an area where Vodafone, for one, is focusing a lot of effort. In a recent blog, the company’s UK drone lead, Jonathan Reid, says there’s “a growing consensus” that mobile network operators have a key role to play unlocking the commercial drone industry. It’s working with a startup, sees.ai, to develop AI technology that unlocks the promise of drone flight beyond visual line of sight.
An ecosystem of partners will be key
Realistically, though, CSPs probably have limited opportunity to sell network insights as individual services. “As a standalone thing, it doesn’t make much sense,” says Justin van der Lande, Principal Analyst at Analysys Mason. “Someone running a car factory isn’t going to say, well, for this one little bit of analysis, we're going to turn to Deutsche Telekom. It's got to be done as part of a broader problem-solving issue.”
In other words, the value will be unlocked when those insights become part of a wider set of services that address a specific business issue. Here, open APIs such as the kind available from TM Forum could play a role in exposing NWDAF insights to an ecosystem of partners, who can fold them into innovative services for specific use cases or vertical sectors.
In the case of 5G drones being used for inspection, for example, it’s possible to imagine a CSP working with providers of other data-driven services to create a bundle that includes real-time collision warnings, adverse weather warnings, and 5G and GPS reliability forecasts. As an integrated suite, this would provide a high level of actionable situational awareness for the drone and its operators.
For CSPs, that will likely mean forging partnerships with ISVs, systems integrators, IoT platform vendors, and even existing customers who want to move into smart, connected products.
For example, AT&T evolved from being John Deere’s connectivity supplier to become a go-to-market partner for the tractor manufacturer’s real-time data services.
A $0.9 trillion market opportunity by 2030
Whichever route to market CSPs find most appealing, the more insights they can bring to the party, the more value they’ll be able to capture – and the more commanding a position they’ll be able to occupy in the ecosystem.
For example, CSPs may see opportunities to combine NWDAF outputs with other internal data sources – such as subscriber demographics or mobility data – to create unique and valuable propositions. They may even consider becoming data brokers in their own right, aggregating data from internal and external sources to provide rich insights directly to end-customers.
Admittedly, it’s early days to be thinking about monetizing 5G network insights. The CSPs including NWDAF in their RFPs today are primarily interested in its potential to drive network automation. But it’s definitely an opportunity to keep in mind – particularly as Bell Labs Consulting estimates that AI and ML services for 5G-enabled ICT will be worth $0.9 trillion by 2030.
Four ways CSPs can get ready for the 5G AI services ecosystem
For CSPs who want to be ready for this opportunity, it’s worth getting a few things in order now. One is mapping out the consumer and enterprise use cases that network analytics could potentially support. It’s likely that CSPs exploring NWDAF are already doing this for internal use cases, but it’s worth thinking about external opportunities at this stage as well.
Another is to ensure that analytics can be available for consumption at the network edge as well as at the core. Many of the game-changing capabilities of 5G involve very low-latency use cases at the edge, so having predictive analytics available here could be very attractive to ecosystem partners and end-customers. And as networks go increasingly cloud-native, an open analytics engine that is portable across private and public clouds will be an advantage.
A third is to allocate resources to explore potential use cases with partners. This will likely involve collaborating to identify potential market opportunities, and co-developing algorithms and machine learning models that turn raw data into valuable, actionable – and saleable – intelligence. While we’ve suggested some use cases in this article, the best opportunities will likely be identified through collaborating with partners who have deep knowledge of specific industrial processes.
Finally, CSPs will want to have capabilities in place to measure and charge for external consumption of analytics. Whether that’s a subscription service, charging per API call, or as a share of revenue generated by partners in the ecosystem, a monetization platform will be needed that allows rules to be defined easily and revenues collected promptly.
Start thinking about 5G analytics monetization today
4G-era services like the mobility insights offered by A1 Austria, O2 and others have shown that network data can be a viable revenue generator in its own right. Indeed, A1 sees so much value in the emerging ‘data economy’ that earlier this year it increased its stake in its ML partner, Invenium.
Now, as guardians of the vast volume of data that’s starting to be produced by 5G networks, CSPs are well placed to capitalize on its market potential. Thinking about the monetization possibilities today, and starting to build relevant R&D and commercial partnerships will set CSPs in good stead to become more than just connectivity providers for the emerging hyperconnected era.