- Telco-as-a-platform is mostly about harnessing telco APIs to enrich enterprise network services with advanced features and data insights, thus helping make good on the promise of 5G
- But telcos are looking to clear a series of hurdles to get into position. Can they do it?
The telco-as-a-platform model may no longer be the shiny new thing it was when the term first appeared on TelecomTV in 2021, but the concept has gained traction in the past year or so thanks to the development of helpful specifications and use cases from the GSMA’s Open Gateway initiative and the TM Forum’s application programming interface (API) initiative.
As a result, there’s been a surge of interest from telcos and major vendors, which increasingly see an API-armed platform concept as a practical way to light a fire under the elusive 5G enterprise services sector that has, until now, failed to truly ignite – see MWC24: The rise of network APIs.
So TelecomTV recently convened a two-day Telco as a Platform Summit to flesh out its pros, cons, barriers and opportunities for telcos, with the help of a panel of experts and some crucial awkward questions from our online community. Here’s what we learned.
The telco-as-a-platform concept envisages network operators assuming an important role as distributors of network functionality, digital services and bandwidth for their existing customers, other network providers and corporates. Telcos are already distributed traffic hubs and, once digitally transformed, are nicely positioned to provide access, via APIs, to functions and network data embedded in their systems through a platform that can be securely tapped into by third-party partners.
That’s a great concept but one that assumes the telecom industry plays its cards right and baits its hooks properly to appeal to enterprises and their all-important software developers.
That’s not a given and there’s a contradiction inherent with the telco approach. On the one hand, corporate customers clearly value the promise of a relatively open platform with a broad range of available APIs, bolstered by ease of integration, software stability and other services and options. On the other hand, it might be argued that an expensively developed 5G platform with data gathered and APIs developed to expose capabilities, should be seen as a potential series of monetisable assets and sources of competitive advantage that shouldn’t be given away lightly.
So it looks almost certain that openness will be attenuated by a lack of access to at least some categories of API for competitive and/or security reasons, not least because many applications rely on access to content and customer data and that runs up against national data and privacy regulations. Similarly, the fragmented nature of the global telecom industry – hundreds of telcos with only slightly compatible technologies and architectures, many of which are competing with each other – means there are barriers to the development of a comprehensive middleware layer that can bring compatibility and interoperability to the network API ecosystem.
What’s so great about APIs anyway and why should enterprises get excited?
When it comes to marketing API solutions, some telco voices are keen to distance themselves from what they suspect might be interpreted as ‘old’ 20th century attitudes.
“Network APIs allow us to talk to our network infrastructure,” stated Peter Arbitter, SVP of product management at Deutsche Telekom’s Magenta API Capability Exposure (Mace) unit, during this panel discussion. “But I don’t like [the term] network APIs too much because it’s pretty telco centric and pretty inside out. I’d rather describe them as business APIs.”
Arbitter is onto something here. All the marketing nomenclature, including “telco platform” itself, describes the concept from a telco point of view. More customer-centric thinking won’t hurt the telco platform effort – words matter.
Arbitter defines two broad API categories: The first enables the user to get that ‘inside information’ from the network; the second offers the ability to configure the functionality of the network.
Amongst other things, information-gathering APIs can allow enterprise end users to correlate information from the telco network with their own application data to really understand what’s driving application performance and what could yield cost savings, meet sustainability goals or enable new service creation, says Verizon’s Beth Cohen.“I’ve been saying for years that we really need a better integration of applications with the network. I think these APIs can allow us to get to that next level. We have a portal from which we share them, but we really don’t expose our internal APIs and we need a middle layer to simplify [the rules around what can be shared and what can’t].”
“It actually gets more complicated than that,” noted Petar Torre, principal engineer at Intel. “Once all the information is exposed we have so many countries and then operators per country and, even if you make it technically very easy, you still have all the commercial and legal terms to sort out and send to the developers to clear their contracts.” Clearly the word ‘exponential’ applies here.
Despite all this apparent complexity, DT’s Arbitter says he remains “bullish” that these problems can be solved. Perhaps the answer lies a few layers higher up with intergovernmental agreements on harmonising privacy and data regulations? One thing’s for sure, it won’t be solved quickly.
Business in focus
Regulation aside, there’s still a lot of marketing and convincing to do if the telco-as-a-platform concept is to get real momentum. Several commenters opined that the only way enterprises were going to warm to APIs was if they were aimed at a “business purpose”.
On this point, Paul Miller, CTO at Wind River, noted it is important that all the companies working on telco APIs extend the northbound interfaces of their products. “If you don’t extend visibility [into your systems] with rich APIs, then the conversations about API gateways and brokers and the way a service provider can aggregate those API functions into something that’s consumable for a business purpose becomes very difficult.”
Miller might be being too gentle here: The idea that an API or two might form the basis for a ‘use case’ – just add your own applications and integration and a business model will emerge – just won’t cut it. According to Miller and others, there has to be a compelling business purpose identified before enterprises will bite. API adoption, they say, must enable service enhancement, a material cost saving, a major goal alignment (sustainability, net zero etc), or all three to get traction.
And several panellists stressed that it was no use telcos waiting for a platform environment with comprehensive standards and extensive API interoperability to be formed before making a start: Telcos should start developing integration capabilities now, even if they only involve one API and an enterprise application, to get the ball rolling.
Developer awareness
The software developers must be cultivated and convinced too. According to Wind River’s Miller, there needs to be an understanding of the value of a “developer tool chain”: In addition to selling its software and technology into service providers, he claims Wind River saw the value in developing “a completely cloud-native, DevOps tool chain” to help the software developer manage everything from requirements definitions to building and testing the final software product. “This is the modern software development environment that a developer lives in,” he says, “and you need the modularity and the libraries to be able to easily incorporate APIs published by the service provider.”
Another requirement, says Miller, is good API revision management. A service provider is very attractive to a developer if it has proved itself to be skilled at maintaining API backwards compatibility. Conversely, breaking backwards compatibility makes it very painful and costly for a developer to deal with.
Also, “if you’re a developer and you build an application for one network and then try to move it to another [network operator], but can’t, you’ll have to rebuild the entire application,” so a third requirement is commonality for similar network services and similar APIs across the service providers, using things like Linux Foundation’s Camara project, which addresses challenges in porting and reproducing API services across heterogenous operator and cloud architectures and develops the specifications for APIs promoted by the GSMA’s Open Gateway initiative.
DT’s Arbitter wholeheartedly agrees, pointing out that in the past telcos had failed to “bring these capabilities to the market because at the end of the day we did not understand the developer play – let’s be honest.” This time, he says, “we must ensure that [API adoption] is absolutely easy.”
A crucial question remains. With all this potential openness and the looming radical mash-up of today’s network and cloud technologies – disaggregated, distributed and reassembled to suit a range of business models – will the dominant platforms turn out to be recognised as telco platforms? Or will some be built on different principles and infrastructure and be orchestrated by an alternative set of players, such as hyperscalers or aggregators? Put another way, will all that openness and cloud-native operation mean advanced network capabilities can and will be built in the connected cloud by anyone?
Alternative approaches
No doubt the telcos that are enthusiastic about the telco platform approach have factored in the inevitable commoditisation of highly prized network functions and features on the assumption that if they don’t make them available on multi-cloud, multi-infrastructure networks for enterprises – either by themselves or through partners – other players will pick up the baton and get there before them.
Indeed, alternative platform approaches are already being developed by several players using the cloud backbone. Back in 2021, for instance, Microsoft teamed up with US multi-cloud backbone provider Alkira with a view to incorporating Alkira’s cloud network smarts with Azure, thus easing the multi-cloud ‘journey’ for corporates wishing to take distributed applications, such as Office 365 and Teams, with them for the trip. Other similar approaches are also in development.
Competition or symbiotic partnership?
So what strategies should guide telcos as they look to bring their advanced technologies to bear. PWC analysts shared a blog post in mid-2022 that’s worth summarising.
Telcos’ multibillion-dollar investment in 5G and multi-access edge computing (MEC) infrastructure have, for the most part, only served to retain already established revenue, claimed the analysts. Any incremental growth has come through market share gain at the cost of more competitive pricing and costly promotions. The telcos strategy of banking on in-house innovation has missed its target, while mergers and acquisitions seeking expansion into adjacent sectors have tended to centre on the wrong targets at the wrong time.
But, they claim, instead of worrying about rectifying weaknesses, “telecom operators can more profitably concentrate on what their substantial strengths can bring to the table.” In other words, perhaps they should keep building out, speeding up, and adding functions and features to their networks to capture profitable connectivity growth, however it arrives and whoever has stimulated it.
- Ian Scales, Managing Editor, TelecomTV
Email Newsletters
Sign up to receive TelecomTV's top news and videos, plus exclusive subscriber-only content direct to your inbox.