How zero-touch operations can save costs and delight customers

In world where the need to efficiently handle increasing data demands is greater than ever, zero-touch operations (ZTO) is becoming an essential facilitator for securing the future of reliable and profitable network operations. Many CSPs are already leveraging the benefits of implementing ZTO to facilitate service innovations, reduce costs and ultimately deliver an improved customer experience. In this blog post we’ll explore some key insights on the ways it can deliver ROI, as well as the solutions being adopted to overcome any potential barriers on the zero-touch journey.

ZTO is key in facilitating continued innovation by CSPs

5G is being introduced at a faster rate than any other previous radio technology, and it enables a whole range of new services to be launched. Better still, it enables several new potential sources of revenue and opportunities for CSPs to take advantage of.

But that also means a lot more devices and data to manage – the latest forecast from Ericsson’s Mobility Report June 2022 shows that 5G mobile subscriptions will surpass one billion in 2022, and by Q1 of this year mobile network data traffic had already grown 40 percent YoY.

All these new services and innovations produce data, so it is becoming increasingly complex to manage network operations and assure a good customer experience for each service. The only way for CSPs to be able to cope with all this increasing data is to move towards ZTO, and by leveraging new key technology enablers like the cloud, more automation, and increased harnessing of AI. Quite simply, ZTO is essential to cope with increasing complexity and efficiency demands while securing a good experience for end users.

There’s also the all-important cost factor. Research we’ve carried out at Ericsson shows that by leveraging AI and zero-touch operations CSPs can avoid an almost 130% increase in operating costs that would otherwise be incurred if those technologies are not harnessed to cope with the explosion in data demands and network complexity that is now ongoing.

Download Increasing operational complexity requires AI infographic (pdf)

Benefits for customers include quicker support and service activation

As the recent Lessons in zero-touch operations report showed, one of the top drivers for CSPs pursuing ZTO is to improve customer satisfaction, with network executives reporting that automation has already improved the quality of user experiences and boosted customer satisfaction as a result.

There are some clear examples of how this can occur thanks to automation and a more AI-powered backend. From a customer support perspective, it could mean that by the time someone (who has opted in to the service) calls the helpdesk of their provider, the technical support staff already have a breakdown of the error their call regards, and how it can be fixed.  The help desk will be able to almost instantly know why there’s a problem for the user, be it due to radio coverage, or a bad phone or device.

Also consider the possibility that when someone purchases a new service, it is up and running within minutes, rather than a waiting period of days. Details that can ultimately have a big impact on customer loyalty.

ZTO can even boost the creation of new services, providing a clearer picture of what services customers will actually need going forward, delivering better recommendations, and on top of that, a snappier interface when the service is eventually created.

Lessons in zero-touch operations

Read more

Energy efficiency gains mean cost-saving and sustainability benefits

In the past ten years data traffic has increased by 300 times the amount it was at the beginning of that period, and it’s no secret that energy bills for operating the growing IT and network environments are a major cost for CSPs today. At present Ericsson estimates the annual global energy cost for running mobile networks to be about USD 25 billion, and it is only growing. As a response Ericsson recently introduced an energy-efficient RAN portfolio with next-generation radios, antennas and software that makes a big difference. However, in complex multi-technology and multi-vendor environments there’s even more efficiency to be gained.

Through ZTO-aided efficiency gains, CSPs can reduce energy usage, reduce the costs incurred, and ultimately reduce CO2 emissions. The key is understanding and dynamically acting on the behavior of how people use services. A simple example is that during the night where data usage is generally lower, a number of network components can be powered off, then powered back up within mere milliseconds when required to ensure a consistent customer experience.

Quickly redirecting and optimizing radio antennas for major live events is another case that can fine-tuned thanks to ZTO. If there’s a major football match occurring, the network can be redirected to have higher efficiency during the game, then dynamically power down. The concept of powering up and down isn’t new of course, but ZTO allows it to be more granular, dynamic and ultimately much more effective. You no longer need people to manually input information that there is a big game on a specific date, antennas need to be tilted, and energy needs to increased.

Instead, it’s automatically done and optimized by a digital twin or AI behind the scenes. The energy savings gains there mean the network and service is running in a more cost-effective as well as more sustainable way, so the benefits are twofold.

Start small and scale-up so staff become ZTO’s biggest champions

New technology always brings a certain degree of resistance, and of course for those who have been working with optimizing or operating a network for 15 years or so, there is a logical trust barrier to overcome when AI-based software is introduced that the expert initially sees as doing the same thing as their job.

That’s why Ericsson puts a lot of focus on what we call trustworthy AI capabilities, with ZTO no exception. We strive to have explainable AI, where we can clearly demonstrate that given ‘x’ input, these are the different kinds of output expected. We let the AI recommend the best scenario, but also build in capabilities that detail how the models are performing and why they’re recommending the things they’re suggesting. It should never be a mystery.

An approach that can be beneficial is to start small and gradually scale-up: there are many options for how extensive the automation can be. Should the AI capability be configured between +/- 1% or +/- 10% over one hour or one day? Different levels of boundaries can be set so you can first make sure the staff are gaining understanding of the process while also maintaining some control, until they eventually see that it’s outperforming what they were doing manually, and that the AI can help them by testing millions of scenarios in a minute.

Explainable, trustworthy technologies ensure the experts can safely and verify, then ultimately, trust these technologies. Then over time they can be rescaled, as the experts turn into the owners of the AIs, steer them and build more policies on their reach. In the end they will likely let go of more of those very cumbersome and less rewarding manual activities, meaning the technology benefits their quality of work. In that case they usually become the biggest advocates for the new technology.

ZTO in Network Healing is complex, but with a huge upside

The Lessons in zero-touch operations report showed that network healing is an area where CSPs expect some of the biggest benefits of ZTO, and that’s understandable when healing and service assurances’ impact on the end customer is easily measurable.

If a customer has just bought a service, is watching a live event on their phone, and then it suddenly drops out or breaks, the impact on customer trust is clearly enormous. It incurs a direct cost, but also an indirect cost where it could ultimately lead to customers switching to another competitor or service provider. Not ensuring a stable, positive service and customer experience has a huge negative cost impact for CSPs, and ZTO can mitigate that risk.

In a pre-ZTO world, the largest-scale problems are the easiest ones to pinpoint. For example if a major Atlantic cable breaks, it impacts internet traffic, which then impacts a range of services, and leads to a major troubleshooting operation. But it’s relatively easy to find the root-cause given the sheer amount of data points to help in that troubleshooting process.

The smaller scenarios are much more challenging. If a CSP launches a new VIP service that’s only available to customers with a certain premium subscription, then trying to understand why that service is misbehaving can be like looking for the needle in the haystack. That’s where AI can really show its worth, rapidly finding hidden service anomalies even when the data sets are small.

For network healing ZTO also facilitates more proactive healing capabilities, such as simultaneously rerouting traffic while sending messages to customers to apologize and explain when their service will be fully working again. That kind of proactivity is very valuable, and has some of the clearest potential for ROI.

Operators have been looking at alarms and performance management for decades but advancements in areas like the Cloud and AI in recent decades make it possible to process significantly more data in aid of root cause analysis and recommendations: ZTO which leverages more advanced technologies can put CSPs ahead of the curve. Exactly what is wrong and why it is wrong is identified quickly, and can be proactively addressed before it becomes a significant problem.

These processes are now becoming more and more sophisticated: rather than just predicting an issue 30 minutes in advance, it can come several hours or days ahead, raising a detailed red flag that if a service’s allocation at data centers isn’t scaled up, then it will start to slow for the majority of users, leading to unhappy customers if unaddressed. Until only five years ago the process was reactive: you scaled up once people called in to complain. Now, thanks to AI and more extensive cheap data processing CSPs can be predictive, more data-driven, and reap the rewards, with even more potential to be leveraged further in the future.

The business cases for ZTO are already clear  

The business cases for ZTO are ultimately already clear. A simple example is how it reduces total cost of ownership. That can come by reducing capex – leveraging guidance on where to build your network and data centers in the smartest way, reducing the number of sites you have to operate through network efficiency.  It can also come in OPEX savings, how your networks and IT environments are run, and as mentioned above, energy savings that can be made.  

There’s also the matter of how many people are needed to build new services and operate them, and helping CSPs to reduce leakages. Networks have thousands of new activations on their network every day, new subscribers being added as they upgrade or buy services, and every one of those networks in the world has leakage to some level. So if the service didn’t get activated in the proper way, that means a loss in money while the service isn’t active. ZTO can reduce that revenue leakage and help CSPs get new revenue into their business more efficiently. 

But as we already noted, in the big picture and long term, customer experience is the biggest, clearest case for investing in ZTO. Ensuring you have the best network quality of service possible and the best service from your helpdesk will improve customer lifetime value and make them advocates for your products. Customers will stay with you, buy more and more services, and promote your service to their friends who also join, and the cycle continues. The payoff there is significant.   

Learn more

Explore Ericsson’s AI solutions customer success stories

Read more Insights on how  zero-touch operations is being leveraged by CSPs