Why big data is a huge challenge for telcos / Dominic O’Connor, Client Director, Acxiom

By Dominic O’Connor, Client Director, Acxiom

The proliferation of media channels and rapid increase in devices which can access them have changed consumer behaviour radically.

A potentially valuable by-product of this is the huge volume of data which is now available to marketers. Against this backdrop, telcos could justifiably claim to be at the leading edge of the big data gold rush – if only they knew how to handle all the information.

As providers of both communication channels and content, telcos – perhaps more than any other type of business – are having to cope with the enormous challenges and opportunities posed by the always online, fully interactive customer. Smartphones, tablets, web-enabled consoles and smart TVs are all generating vast amounts of data from customers’ emails, texts, tweets, likes and purchases. In the UK alone a single telco could process over 1 billion interactions per day from its customer base.

In fact the sheer volume of data poses a significant challenge – how to handle it, let alone analyse it.

Telco providers

Adding a further layer of pain for telcos is the fact that customers expect the companies which manage their communications to know who they are every time they get in touch. If I ring a call centre, walk into a shop, fill in a web form or send an email I expect my telco providers to know who I am – after all my mobile network provider knows where I am at all times (do you ever switch your mobile off or leave it at home?); my TV provider knows what I watch and when. So how come when I ring them up they don’t know me? Why do I get new customer offers through the post when I’ve been a customer for a decade?

Customer recognition is getting more difficult .Over the last decade most of the big operators implemented customer relationship management programmes and the technology to support them. However, these systems were designed to handle fairly rigid interactions: response to mailings, contacts to call centres and so on. The problem now is that the connections have multiplied and are quicker and in unfamiliar structures – text messages, tweets, emails, web logs – and the old systems just can’t cope.

In the past, traditional mainframe databases and analytics solutions like SAS and SPSS were fine for dealing with data, but that is no longer the case.

The positive news is that frameworks like Hadoop (open source software) can help analyse the vast quantities of unstructured data. Hadoop is primarily used in order to capture and analyse unstructured data, which is currently proving the most challenging aspect of the data deluge. However, Hadoop alone can’t manage everything, which is where third-party data companies come in.

Customer behaviours and interaction

The key to mining the data effectively, analysing it and making the insight actionable is customer recognition; get this right and everything else falls into place. At Acxiom we are building data and knowledge-based solutions which recognise consumers across all channels and maintain persistent keys. This allows telco clients to not only recognise and act upon customer behaviours and interaction, but also to maximise these new channels without cannibalising their business or annoying their customers.

Currently many telcos’ only solution is to aggregate the data in order to make some sense of it – but this is throwing away individual level customer data which could be gold dust.

Industries like banking don’t have this issue, as unstructured data is not a big concern for them and customer recognition is naturally at the forefront. The onus is on the customer to identify themselves and, despite the demise of the personal bank manager, data still allows banks to maintain the emphasis on creating a good, lasting customer relationship.

In terms of which technology solutions are flexible enough to handle the kind of data that telcos produce it depends entirely on what you want to do with the data. For example, if the aim is to use the data to manage customer interaction, it is essential to have the technology to analyse and act upon real-time interactions, or as close to real-time as possible. If you’re leveraging insight, there is less need and data can be handled in batches which are more manageable.

In all cases, however, remember that customer recognition is key. The telcos which act upon that will be the ones to profit from big data, rather than being overwhelmed by it.

By Dominic O’Connor, Client Director, Acxiom