Big Data Summit: Big data needs a boardroom champion

Research unveiled at today's Big Data Summit suggests a clearer case needs to be made for big data solutions

Big data needs a champion on the board to push for its adoption in business. That's according to Computing research unveiled at today's Computing Big Data Summit.

Computing surveyed and interviewed hundreds of IT professionals, with almost two-thirds - 64 per cent - saying that a clear business need would be the main driving force behind any big data intervention, rather than a response to ever-increasing storage demands or the need to keep up with technological innovations.

Email is the key area in which data is growing, according to the respondents, followed by soft copy documents - such as Word or Excel documents - then web traffic and social media streams.

Part of the reason for this growth is simply the ability to store more, as a CIO in banking, argued: "As your ability to store the data increases, the amount of data you retain increases." It is the aim of big data technologies to exploit all of this extra data to the benefit of the business.

However, increased capacity is not the only reason we are storing more data. An IT director in local government told Computin g that CCTV camera footage is "kept indefinitely in case it's required for a criminal prosecution".

Making this type of unstructured data searchable and available for analytic processes is where big data technologies come in. Big data is about structuring the unstructured. Given that the proportion of stored data that is unstructured is rising all the time the need for such tools is increasing. According to the survey, in 15 per cent of organisations unstructured data now makes up more than 70 per cent of all data they store. Within five years respondents estimate this number to double.

Currently, however, the vast majority of analysis is still done on structured data using standard business intelligence tools. The integration of unstructured data proves to be a challenge, as does interpreting the true meaning of the data.

As one IT professional in the NHS said:

"The data might be a set of notes that say something about an individual who might have Parkinson's Disease, or it might be about a Mr Parkinson. The problem we've got is looking at that data in a way which is governance-wise and carefully done. Secondly, we need to draw out what the meaning is. We're spending a lot of time on those techniques at the moment because a lot of the rich content data tends to be in free text notes."

A steady stream of people are doing likewise. Among those interviewed for Computing's research, those who've already implemented big data solutions would generally recommend other IT professionals and organisations to do the same.

"My advice to anyone doing a big data exercise would be to go for it because the benefits are quite enormous," said a local government IT director.

There is an important caveat, however: it's important to employ those with knowledge of the subject.

"Get somebody who's got some experience and knows what they're doing because these projects can go horribly wrong... " agreed the CIO of a bank.

"Ensure you have got key stakeholder alignment around what the outcome is going to be because the allure of data to an untrained person means that you open a whole rabbit warren and the insights that you want just don't exist... " warned another in media, indicating that the approach you choose will have a profound effect on what Big Data will achieve.

As the opening session of Computing's Big Data Summit concluded, there needs to be a balance between experimenting with big data and achieving buy-in.