Telcos are collecting growing volumes of subscriber data and face pressure to analyze it and act on it more quickly than ever. Over time, telcos’ data warehouses have become complex and expensive, and it has become evident that traditional architectures will not be able to sustain the explosion in demands placed on them.
One potential solution is “analytics as a service” (AaaS), a hosted, general-purpose analytical platform that can be shared across an entire company. Enterprises such as Amazon, Opera Solutions, and eBay have already deployed AaaS to meet increasing analytics demands within time and capital constraints. Taking the concept further, Amazon Web Services recently introduced a cloud-based AaaS model – in other words, on-demand and pay-as-you-go.
Ovum recommends that telcos and IT suppliers serving the telecoms industry actively explore AaaS, because it may be less expensive than traditional, on-premise solutions and it enables the more rapid deployment of advanced analytics capabilities than traditional solutions.
Telcos struggling with departmental analytics
Today, telcos typically have a departmental business intelligence (BI) and analytics structure. The increase in front-end analytics tools, back-end BI tools, platforms, and data marts has raised the cost of managing, maintaining, and developing the “data-to-analysis” value chain. Telcos are increasingly finding it necessary to restructure disparate platforms and tools into centralized, yet flexible, analytical infrastructures.
After years of cost-cutting, telcos are now pursuing top-line growth and see analytics as a way to drive it. Operators are concerned that traditional BI solutions will be unable to cope with the growth in data volumes and the need for flexibility in data structures. But they also are constrained by huge investment costs, and face the challenge of predicting infrastructure needs for emerging analytics scenarios. These factors, together with a general global shortage of trained business analysts, are creating new opportunities for IT vendors to deliver AaaS to telcos.
The AaaS operating model should not be confused with the software-as-a-service model. AaaS solutions may be deployed on-premise or based in the cloud, whereas SaaS solutions are primarily cloud-based. Also, SaaS implies the use of a single application, while AaaS provides a general-purpose analytical platform as a shared utility that is not limited to a single database or application.
An AaaS solution, regardless of deployment model, is available to various departments as a service. Any business unit can bring its own data to the centralized infrastructure, perform its own analysis, and combine it with any other data already present in the infrastructure. AaaS can enable telcos to carry out petascale analytics while maintaining consistency, availability, and security in a cost-efficient manner.
Cloud analytics - a key battleground
Business analytics and the cloud are two technologies that are changing the face of the communications industry. Vendors are promoting AaaS as a platform-as-a-service (PaaS) model. IBM acquired cloud-based analytics vendors DemandTec and Coremetrics to extend its Smarter Commerce initiative. In 2009, Adobe acquired a cloud-based web analytics vendor, Omniture. Amazon is taking the first steps to promote the concept of fully cloud-based AaaS, rather than a cloud-based extension of on-premise BI solutions. This is a great example of value migration from on-premise models to new architectures that can better satisfy businesses’ changing priorities. IT vendors are targeting firms that deal with large data volumes and need flexible infrastructures.
AaaS will be of utmost importance to the telco industry as the M2M market evolves and billions of appliances are connected to its networks. It will help overcome the challenge of building complex infrastructure to harness data in motion and derive insights quickly, without waiting weeks or months. The ultimate goal for AaaS is to provide unconstrained tools on a pay-as-you-go basis.
IT vendors are witnessing an uptake of subscriber data management (SDM) solutions that create a centralized repository for all subscriber data. Leading suppliers such as Huawei, ZTE, and Amdocs have clearly defined plans to introduce SDM in the cloud in the coming 12 months to 18 months.
As vendors introduce SDM in the cloud, IT vendors can also explore the opportunity to provide AaaS in the cloud. With the AaaS model, IT vendors can provide telcos with a pay-as-you-go model, helping them in their evolution path without having to upgrade their own BI systems.
Although the AaaS concept is too new for telcos to have developed any case studies, the shared utility model makes business sense. But given operators’ huge investments in multiple BI projects that are already in progress, “rip-and-replace” is not a good option in the short term.
Soon, telecoms operators will look towards IT vendors to deliver what innovators like Amazon are providing. Amazon’s AaaS strategy shows the direction in which the enterprise analytics marketplace is most likely to evolve. In addition, BI and analytics providers can consider AaaS as the basis of a potential new outsourcing offering for telcos.
Shagun Bali is an analyst with Ovum, covering IT applications and vendors serving the telecommunications sector.