By Paul Colgan
The commercial outlook for Telcos will not be significantly different from last year. Margins are still being squeezed with flatlining revenues driven by a reducing euro per gigabit ratio as data volumes rocket even further. Telcos are concentrating their efforts on finding new sources of revenue and streamlining how they run their traditional operations. This, in turn, highlights several industry themes:
• B2B opportunities
• Open integration
• Standardization and automation
• AI/ML with real-time analytics
You may be surprised that I haven’t specifically called out 5G or autonomous driving which are certainly themes de rigueur. These are worthy of a blog in themselves along with one about the growing appetite for Private LTE. But to paraphrase a former UK PM, “it’s time we got back to basics”! Watch this space.
New Revenue Streams – Grow B2B
As basic connectivity continues to be commoditized the most obvious place for Telcos to look to gain more revenue from are their enterprise clients. New opportunities will continue to open with Industry 4.0, network slicing and every company globally undergoing some form of digital transformation.
One initiative recently that caught my eye is the TM Forum’s Catalyst involving Vodafone, Orange, BT, China Mobile, and others. As an enterprise-focused shared marketplace app store hosted and delivered by the Telcos, it enables Telcos to share services they profitably created so that they can be replicated by another Telco in a non-competing geography. It requires the use of an open framework with API management, service and resource orchestration, order, catalog and charging management. The reality is that most Telcos don’t have this capability in place today, but all have programs working towards such a setup.
To be successful these services need to be easy to consume for the customers and to use with the least interaction between the Telco and the end enterprise. Selling digital services to a new demanding customer type is not in a traditional Telco’s DNA so it needs to be simple to deliver.
Reaching a Critical Mass for Auto-Scaling Infrastructure
Overtime as a critical mass of Telcos develops the ability to spin up and auto-scale agile (per network slice) infrastructure alongside their current capacity, more complex services will emerge. To date, all operators have implemented some form of simple NFV deployments. A few have added full automated lifecycle management of the infrastructure and service. Once this becomes the norm we will see a wider range of profitable services launched cost-effectively at scale. Watch out for more examples of the transformation of enabling functions within Telcos in 2020. It’ll be interesting to see how this type of marketplace moves from concept to mass delivery.
This is not the only route, of course, other non-traditional revenue sources are being pursued by some Telcos. Mobile Payments (mPESA in Vodafone), Fleet Management (Verizon) and Healthcare (TELUS) are a few examples but this is not the norm across the industry yet and they don’t generally need on-demand infrastructure growth of the Telco network for delivery. They are separate entities altogether.
The challenge of how to put in place the integrated backend orchestration, automation and APIs for these new network focused services is driving Telcos to review their analytics stack – especially their real-time analytics capability. Without auto-scaling being driven by real-time per subscriber per network slice visibility these services will incur more costs and complexity and delay the transition of the Telcos to the much-vaunted DSP model and its promise of profitable growth.
System Integrators (SIs) will play an important role as they are more likely to combine and package compelling B2B service offerings that Telcos can sell on their marketplace to enterprises. It is the SIs and not the Telcos who have the industry and domain knowledge to create these offerings. Telcos can create digital marketplaces and IoT platforms but they will need domain knowledge to package enterprise offerings to offer in these marketplaces.
On the consumer side of the business, new technologies like 5G and network slicing are also opening up new revenue opportunities. Gamers have always been demanding niche consumers, willing to spend large sums of money on software and hardware. Samsung is offering a month’s subscription to Hatch (playhatch.com) with its new S10 5G device. Hatch is the world’s first mobile cloud gaming service created by Rovio (of Angry Birds fame). Telcos are following suit with Vodafone in the UK, NTT Docomo in Japan, Sprint in the US and Deutsche Telekom in Germany all following suit. With no downloads or updates, this all-streaming service allows games to compete against each other on a live scoreboard while on the go. As it is a streaming service it requires low-latency and high bandwidth to provide a good customer experience. Perhaps a perfect marriage to 5G’s key features?
This focus on low-latency will continue in earnest in 2020 as Telcos look to compete with Amazon, Microsoft et al on new services that demand low-latency performance. By deploying MEC solutions throughout their geographies they will have resources much closer to the end-client than the hyperscalers. Are there business problems today that require single-digit millisecond response times rather than the circa 15ms that the hyperscalers can provide from further afield? In the short term, I am not convinced, especially for the consumer side of the business. Ask me again when autonomous cars and smart cities are the norms! Then we will need to have extremely low-latency compute to ensure all the C-V2X (Cellular vehicle-to-everything) communications can be delivered within the strict time limits. On the enterprise side, I think MEC solutions will have success in the private LTE/5G marketplace.
In this context how AWS Wavelength plays out in 2020 will be an interesting development to follow and raises some awkward questions. Have the Telcos made a mistake and let the fox into the hen coop? As developers can now realize low-latency capabilities on the AWS cloud, Telcos no longer have a unique differentiator. It’s worth considering if Telcos would have been better off offering open-source services on their MEC compute rather than enabling the hyperscalers.
There is an argument to say that today people are more likely to use this low-latency capability as it is linked to the AWS cloud and its myriad of popular services, but I believe the Telcos have made a strategic mistake. This is something that will become clearer in 2020 and beyond and I for one will watch with interest.
Streamline Operations – OPEX and CAPEX Plays
In many ways, streamlining current operations would seem to be a more attainable goal. Without a doubt, this is a difficult process but there is a clear path to achieving OPEX improvements using maturing technologies. On the CAPEX side, Telcos seem to be following a couple of approaches. To raise the capital required for their 5G network investments Telcos are offloading their network towers to dedicated entities – either selling to a 3rd Party immediately or splitting them out into entities ready for sale. This removes the ongoing costs and frees up cash for their build-out programs.
Network sharing among operators has re-emerged as another method of reducing additional CAPEX. In the UK and Italy, Vodafone is sharing its 5G network with O2 and Telecom Italia. With the physics of a higher band spectrum in 5G requiring more cells per fixed area, we will see more of these agreements that will not only deliver less capital expenditure but may also mitigate public opposition to lots of cell towers from numerous Telcos across our cities. On the OPEX side, numerous approaches are being pursued. Illiad Free has shown the benefits of using online-only distribution and lean IT systems, to the point they operate at half the cost of competitor incumbents. From a network perspective, in 2020 there will be a significant increase in the shift to lower-cost OTS hardware with real-time process management of virtualized NFV/SDN infrastructure. The use of AI/ML-driven automation and orchestration will reduce the OPEX required to run the network.
NFV/SDN/Containers – Delivering Promised Business Benefits at Last
To achieve the agile infrastructure, Telcos started to add VNFs and SDN capabilities to their networks. Most have done the easy stuff by now with their BSS and IMS systems now largely rationalized from four or five legacy vendors to one or two running as VNFs. They have gone from bare metal to VMs to containers but haven’t gone all the way to full lifecycle management, API exposure, real-time analytics and the automated action needed to bring the promised operational savings and service agility.
As vendors stop putting the same legacy software in a Docker container and calling it cloud-native and Telcos continue to rationalize and integrate their data silos and systems we will start to see the promised business benefits. Without these building blocks, integration and automation can’t happen. We are in the “Trough of Disillusionment” with NFV but in 2020 we will start to emerge from it as more systems interconnect and more real-time analytical intelligence embeds in the infrastructure to enable the right agility needed to bring the promised business benefits.
Open-Source or Open-Approach: Which Road for Success?
For the last number of years AT&T has tried to do everything open-source but with billion-dollar contracts with companies like Oracle and IBM, it seems to be balancing the approach somewhat. Successful open-source projects usually have one or two benefactor commercial companies behind them contributing to code improvements and direction. A critical mass of Telcos have not followed AT&T’s lead and are going for an Open-Approach rather than an Open-Source.
That is why the approach of TIP, being driven principally by VF and Telefonica, I believe has a much better chance of being successful. Predominantly small, lean commercial enterprises are interested in building these open interface products with new business models detached from the old appliance/hardware-based one favored by the large incumbent vendors like Huawei, Ericsson, and Nokia.
With Vodafone now committing to replacing their RAN with O-RAN equipment there is a large commercial prize up for grabs. This will shake up the RAN market. As soon as it delivers operational and commercial success, the second wave of more risk-averse Telcos will look at these new vendors for their RAN too. They can’t do the latest and greatest version of the specifications but there are lots of operators who still need to expand their 4G network.
The US government’s concerns about the influence of Chinese vendors like Huawei and ZTE means there will be significant interest behind O-RAN in the US too as an obvious way to create an alternative US-centric market player. Mavenir, Parallel Wireless and Altiostar are 3 companies well placed in this new market, based on the TIP RFP results – all 3 have their headquarters in America.
Robotic Process Automation – a Key Building Block
As an extension to layering real-time analytics and orchestration on top of dynamic network infrastructure, Telcos need to digitize and standardize their business processes. Only when this robotic process automation (RPA) is in place can the frequent, repetitive tasks and processes be executed with the error-free, reduced costs that Telcos need. This robotic process automation is not only for the regular network operation processes but also for all backend processes such as order processing and fulfillment. RPA is the key building block of the API driven Digital Services strategy Telcos will follow as they try to attract ecosystem partners and create those multi-faceted marketplaces I mentioned before. Therefore, most Telcos will start to roll out RPA across a number of their large interconnected business processes this year.
AI – a Little More Conversation!
The importance of AI and real-time analytics can’t be overstated. It is vital for the automated and granular optimization and predictive maintenance of the Telcos networks, along with the digitization and automation of the end to end business processes around them. Most Telcos have dabbled in several AI/ML PoCs over the last couple of years. In 2020 I expect to see those to go into production. Well trained AI can analyze large volumes of time-series network data in real time, identifying anomalies and predicting events that require action, much better than traditional methods. Using these AI/ML algorithms Telcos can optimize and auto-scale up/down their networks. Arista’s recent purchase of Big Switch to layer analytics and automation on top of their portfolio of products confirms again the direction the industry is taking.
Outside of the network, nearly every Tier 1/Tier 2 Telco globally uses some form of virtual assistant or chatbot to manage their customer service engagements. With the massive number of requests per day, it was an obvious area to focus on first. When Vodafone state they’ve delivered a 68% improvement in customer satisfaction after introducing its chatbot TOBi, it highlights the concrete business benefit of introducing automated technology.
The significant proliferation of voice-enabled smart speakers in our homes (>20% in the UK) has made conversational AI another customer channel that a Telco must manage. Currently, most of the “skills” are very simple tasks rather than “conversational” but in 2020 we will see the start of this change. With massive global interest in this area there are new startups every day entering the market – “skills” are the new apps. As companies like area22.com push to move the interaction towards a human-like engagement with emotion, empathy, and memory, we will start to see the foundations forming so our Alexa/Siri/Cortana/Bixby, etc., is more of a “conversational assistant” and not the current transactional robotic engagement.
Risks – Living at the Edge
With significant opportunities for Telcos in both new revenue generation and cost efficiencies, there are also new risks. Some large enterprises, especially in the manufacturing sector, are investigating the use of private cellular networks for their mission-critical communication and IoT needs. With the regulators assigning the shared spectrum for these purposes, there is a risk the Telcos (MNOs) could get bypassed.
Security at the edge is a very real risk for Telcos. Millions of new IoT devices and connections to their networks mean more opportunities for security and privacy breaches. With Telcos deploying MECs in their networks and enabling 3rd parties to install their applications on this network infrastructure, 2020 will see Zero Trust isolation technology being used by the Telcos in their edge to protect themselves from within. Without such technology, there is a very real risk of their network being compromised.
More of the Same!
In summary one could argue that the more things change the more they will stay the same in 2020. The underlying business requirement to use evolving technology and solutions to deliver better services to existing and new customers, to drive margin gain and to improve operational efficiency will remain as important this year as it was in 2019… “plus ça change, plus c’est la même chose.”
Paul Colgan is the SVP of Telco solutions at Kx. He has over 19 years of experience working in the Telecoms software industry in roles spanning engineering, product management and go to market strategy. During this time the companies ranged from first-day start-ups growing to be acquired to large publicly quoted global entities acquiring others. The common thread was a focus on high performance, high throughput network-centric solutions to help Operators solve their toughest business challenges.