2016 has been a year of transition for many CSPs. We have seen numerous NFV projects initiated and new 5G developments announced on a monthly basis. This transition looks set to continue into 2017 as data monetisation becomes a reality, prediction technology advances and the concept of ‘data for social good’ arrives. Read our 17 predictions for big data analytics in telecoms over the next 12 months.
1. Operational silos will finally be eliminated: Over the past 5 years CSPs have progressed toward centralised operational analytics. As the benefits have begun to materialise, traditional network and IT functions have begun to work more closely, and 2017 looks set to be the year when these operational silos can be eliminated.
2. Automated NOC operations will increase: Data automation has made its way into the Network Operations Centre (NOC) via the use of consolidated operational intelligence software, with use cases such as automated KPI breach alerting and automatic ticket creation. As more complex use cases are identified, automation looks set to increase in 2017.
3. Predictive network planning will emerge: Network planning teams are now looking to invest in network infrastructure based on business value. This means that they will use more and more data analytics to forecast future areas with high network congestion or where VIP customers’ quality of service will be impacted.
4. Analytics will become a truly horizontal function: Since network performance has such a dramatic impact on customer satisfaction, it’s no surprise that commercial and operational analytics are beginning to merge. In 2017 centralised data lakes and analytics platforms will serve both network operations and commercial roles such as customer care and marketing.
5. We will see a rise in OSS as a service: With budgets stretched, CSPs are looking to move away from high CAPEX spends and transition to service led models. OSS as a service is likely to rise in popularity into 2017, with the potential of reduced upgrade cycles, elastic scaling and predictable investment cycles.
6. There will be a revived interest in 2G: 2G networks were set to be retired over the next couple of years, but this is by no means set in stone. To support the ever increasing amount of IoT devices, a ‘GSM Evolution’ could be on the cards, reviving interests in the 2G network in 2017 and beyond.
7. IoT analytics use cases will emerge: IoT chips are advancing, and can now do much more than simply report small bits of data. However, for CSPs, the actual use cases of IoT analytics still remain somewhat unclear. Could 2017 be the year of clarity in IoT analytics? One example being evaluated is the ability for CSPs to recommend add-on services or products based on a consistently demonstrated need via IoT device analytics.
8. NFV analytics will rise up the agenda: Many CSPs have made good progress in 2016 towards the dream of virtualisation. With the infrastructure and virtualisation layer now in place, 2017 will see an increased focus on actually monitoring and analysing SLAs and customer experience to ensure a continued quality of service as different services transition to NFV.
9. Cell-less RAN will bring new questions: As just one of the 5G technologies being researched by CSPs, cell-less RAN has the largest potential to impact big data analytics. But as a radical change in the way mobility is handled, it won’t be till the end of next year that we will get a better understanding of its true impact, and indeed viability.
10. EAN analytics requirements will become clear: Expected to launch mid-2017, the European Aviation Network (EAN) will combine land-based LTE networks with satellite infrastructure to provide in-flight connectivity. This new combination of technologies will of course require data analytics to understand usage and optimise the quality of service.
11. LTE-M will join the market: LTE-M looks set to join the already complex communications market in 2017. Designed to support IoT and M2M communications, LTE-M will have a peak download and upload rate of 1 Mbps. Of course as with any other technology, LTE-M will need continual management and optimisation to ensure coverage and connectivity for M2M devices.
12. Data monetisation revenue will finally materialise: Data monetisation has been on the telecoms agenda for many years, but actual revenue has been sparse to say the least. But this looks set to change in 2017 as CSPs begin to put real values on their data insights and begin to pitch it to industries such as advertising, retail and tourism.
13. Smart city insights will deliver savings: Collaborations between CSPs and smart city projects are identifying huge savings. In just one use case, BT identified a £104 million saving for a local council through the analytics of smart parking sensors. This trend looks sent to continue as IoT and M2M connections increase.
14. Machine learning will unlock new revenue opportunities: Machine learning is slowly finding its place within the telecoms industry, as use cases begin to emerge. By combining machine learning with personalisation, CSPs can unlock new revenue potential via rapid marketing campaigns and retention programmes.
15. Big data analytics will restore the work-life balance: For a long time CSPs have been stretched with increased data loads and limited resources, which has no doubt had an effect on the work-life balance. But with automation and prediction technology advancing, alongside easy data access through tablets and mobiles, CSPs are now restoring the work-life balance as they head into 2017.
16. Data for social good will see a surge: Although a relatively new concept, many large CSPs are now teaming up with governments, aid agencies and environmental groups to use their data insights for social good. This is helping the understanding of human behaviour in emergency situations and natural disasters.
17. Recruitment will remain a challenge: In 2017 CSPs will face similar challenges they see today with recruiting new talent. New technologies such as 5G and NFV require a unique combination of skills; software engineering, telecoms knowledge and data analytics – this is proving difficult to find!