Data analytics for telecoms in the IoT age

IoT network
Pete Koat, CTO of Incognito Software Systems, says that the number of connected devices per households is higher than ever before, so what does this mean for communication service providers (CSPs)? How big an impact will the Internet of Things (IoT) have on service provider networks, and how will this affect subscriber quality of experience (QoE)?

These questions are important, but it’s also necessary to consider new ways to close the gap between the cost of network upkeep and subscribers’ increasing expectations — are there any opportunities for monetizing the services or data increasingly flowing through the network?

Recent forecasts estimate that more than one billion new Internet users are expected to join the global Internet community, growing from 3 billion in 2015 to 4.1 billion by 2020, while global IP networks will grow by an additional 10 billion new devices and connections in that period.

For CSPs, this presents a conundrum — subscribers’ increasing appetite for IP-based applications is driving bandwidth usage and causing network congestion, which results in poor subscriber QoE and/or expensive network upgrades. However, with every challenge there are opportunities; these changing subscriber habits also offer service providers potential revenue streams.

Beyond the obvious implementation of bandwidth tiers and consumption monetization, the very devices and applications favored by subscribers today are a goldmine of data. Analysis of this data could be your company’s most valuable asset — but few providers are maximizing this resource for its full benefit. Focusing on data analysis and monetization while still respecting subscriber privacy is possible — and necessary — to stay afloat in an increasingly competitive market.

In Telco or converged environments, device usage statistics and subscriber data may be gathered through technology or protocols including the Broadband Forum’s TR-069, SNMP, Deep Packet Inspection, through IPDR or AAA records, or even technologies like virtualized customer premises equipment (vCPE).

Whatever the method of data extraction, the challenge comes in extrapolating trends and utilizing intelligence at the edge to generate new streams of revenue. Make no mistake — Big Data is extremely valuable. This information has potential for monetization at every department of a modern communication service provider, from engineering, operations and planning, to marketing and sales.

For example, the data contained within subscriber and network statistics enable:

# Identification of heavy users and premium subscribers for network improvements and new revenue opportunities

# Implementation of network-wide or targeted revenue-generating policies and promotions

# Identification of areas or subscribers with lower QoE due to connectivity issues such as network congestion or WiFi topology

# Proactive care scenarios and personalized services to increase stickiness and reduce customer churn

# Deep informatics for visibility within subscriber networks to provide granular control and new service offerings such as unified IoT management and parental controls

Analytics for Policy Enforcement and Improved Network Performance

Bandwidth is a shared resource and the consumption practices of heavy users can impact service quality for other subscribers, not to mention increasing the cost of network investments. Use data to enforce fair usage policies such as bandwidth caps to ensure subscribers are not costing more than the revenue provided. Identifying premium or heavy users also offers new revenue opportunities. For example, target subscribers that may require customized top-ups by alerting those customers that are nearing their bandwidth cap so they can easily make a purchase for additional data.

CSPs should be able to offer differentiated or premium offerings to address the needs of those customers to maximize revenue opportunities. The data gathered provides the ability to identify premium subscribers and correlate information about network congestion to see whether upgrades should be prioritized in a particular area to meet service level agreements (SLAs).

Launch New Services

Data analysis provides unique insights into how subscribers use your infrastructure. As the options and applications available on the myriad of connected devices grows, the average usage profile is becoming as unique as the individual. The old world of three tiers of services is no longer sufficient for most service providers. But every challenge presents opportunities — the growing world of over-the-top (OTT) services, converged access, mobility, and IoT can be monetized.

Analytics of subscriber behaviour combined with visibility into customer premises networks enables granular control of new service offerings. This is essential to launch offerings that are relevant to your customer base quickly and easily. For instance, parental controls, content filtering, IoT management, and even personalized services such as speed boosts during peak hours are only possible with visibility into subscriber networks and an understanding of usage habits.

Existing protocols like TR-069, as well as emerging technology like network function virtualization (NFV) and vCPE enable the fast rollout of these services, but also remote management, granular control, and easier diagnostic capabilities.

Boost the Customer Experience

As the number of devices, applications, and access technologies continue to grow, the IoT and WiFi topology becomes increasingly complicated to manage, troubleshoot, and optimize.

The reality is that telecom network operators often have SLAs defined for commercial or premium subscribers. However, the increasing number of devices using data with those networks can make it challenging to ensure a high service level without visibility, control, and an automated means to ensure those agreements are met.

Insight into customer premises networks and the WiFi landscape is therefore essential to ensure subscriber QoE expectations are met. A robust device management solution enables you to see devices on the network, as well as identify issues that could impact the customer experience such as WiFi signal strength, outages, unknown devices, or failing equipment. By implementing proactive, real-time monitoring and management of devices beyond the gateway, these solutions can detect and stop issues before the customer is even aware of them.

Collect and analyze subscriber statistics on an ongoing basis, including channel selection, frequency band, and output power. Using these valuable statistics, you can continually optimize WiFi in high-density areas such as multi-dwelling units (MDUs) to ensure consistent QoE and pre-emptively avoid quality issues.

Data collection and analysis are essential to reduce support costs. With a full device history at their fingertips and insight into issues affecting the subscriber network, customer service representatives (CSRs) can quickly understand common problems — such as dropped signal, interference, range issues, or network/device outages — reducing customer frustrations without escalating support costs.

Offer Tailored Promotions and Improve Return on Investment

Visibility into usage habits and trends makes it easier to understand what new products should be offered, and to whom. Identify when a customer should be offered a different package or media access, or a trial upgrade to better suit their bandwidth utilization history, traffic hours, traffic type, and subscriber type (business or residential). A small or medium-sized business using too much bandwidth may have different needs to a residential subscriber — for example, a hotel may be interested in a new fiber service to improve quality of service rather than a speed increase. Understanding usage data allows for smarter marketing and sales strategies and the ability to successfully plan future revenue-generating opportunities.

The hardest part for most service providers is finding, tracking, and making sense of the increasing amount of data available. Today, it is more vital than ever to filter and analyze raw data to make sense of current network status, forecasts, and subscriber behaviour to uncover under-utilized resources, plan for the future, and strategize for new revenue opportunities.

Without the ability to find and analyze the information hidden on your network, you risk increased bandwidth congestion, lower service quality, greater subscriber churn, and worst of all, a loss of profitability.

Increasing average revenue per user (ARPU) in a competitive market is always a challenge; however, with the increased volume of data and improved granularity CSPs can generate new revenues from hyper-customized offerings for today’s increasingly unique customer profiles. At the end of the day, data analysis provides service providers the ability to deliver increasingly customized services, improving customer experience and increasing the profitability of the deployed infrastructure.

Pete Koat, CTO of Incognito Software Systems