Why culture and data velocity are hindering your big data success
- 11 October, 2013 18:38
Cultural acceptance of data analytics insights is one of the biggest barriers for organisations looking to utilise big data to drive competitive advantage, Accenture Australia’s analytics chief claims.
CMO caught up with the consulting group’s analytics practice lead, Michael Pain, to discuss the rise and application of big data and data analytics within organisations today, as well as its transformative impact on all functions within the enterprise.
According to Pain, more and more organisations are investing in their ability to deliver analytics capabilities across all areas of their businesses. The most active area to date is sales and B2C transactions and marketing, which he put down to the advent of the Internet and rise of new consumer interaction and business delivery models. Key industry sectors include telco, retail and banks.
“That’s why the CMO’s role has evolved; they sit at the centre of what is a major tectonic shift in the consumer interface,” Pain claimed. “When you talk about sales and marketing, it’s generally an easily measurable impact area, and in working on analytics project to understand customers, generate a more detailed go-to-market model, develop leads and run campaigns, we have seen large ROIs of 300 per cent, depending on industry and sales.”
He pointed out, however, that big data analytics can apply to virtually all aspects of business, with other significant applications including fraud detection, asset management and engineering maintenance.
For Pain, the key to successfully conducting data analytics projects comes down to “the three flavours of data velocity”. The first, speed to insight, is the speed with which a business can use data to develop insights and models. This has significantly compressed thanks to the Internet as well as the availability of a range of technologies for crunching big data, he said.
The second, speed to application, is how that insight is implemented and how organisations apply it to their business and processes. The third flavour of data velocity is speed to iteration, and how quickly your processes and insights can be improved, measured, augmented and restructured to drive competitive advantage.
While most organisations investing in data analytics projects have a handle on the need to develop models and insights, many struggle with the application or “industrialisation” of these insights, Pain said.
“This often involves taking an insights or an action and building it into what might be legacy platforms or processes in one form or another,” he explained. “Also, measuring your impact, or that speed to iteration, is something organisations still struggle with.”
For Pain, the roadblocks to getting data analytics projects right vary depending on organisation. One reason can be that organisations haven’t started with a good enough hypotheses or question to ask.
Compiling data sets in a way that it can be utilised is another huge step. “We find that when we undertake the proof of concept with clients, at least 60-70 per cent of the time we spend will be on getting legacy data organised, available and scrubbed is frankly a huge barrier,” he said.
Arguably the bigger barrier organisations have to work harder on addressing is the application of analytics insights within the organisation, Pain said. This can be particularly hard when you have staff who haven’t seen analytics as the way they’re going to make decisions, he pointed out, raising significant cultural challenges.
“Those people may have been making decisions on gut feel for many years, and been on the front-line selling to customers that way,” he said. “When you say you’re going to give them an analytical driver for your business and expect them to follow that, it’s a non-trivial issue in many cases.
“Organisations almost have to look at changing their operating model so that the analytics function is not a standalone separate function, but an integrated part of the processes they’re running.”
According to the 2013 State of the CIO survey compiled by CMO’s sister publication, CIO, half of CIOs are not currently engaged in big data analytics projects for their organisation.
Pain agreed the ratio was about the same across Australian businesses generally, but said most of his clients are at least exploring use cases of big data.
That said, big data analytics projects are only in their first stage of iteration, and Pain saw advancements in using new and unstructured data sets as driving further innovation.
“Big data technologies are here to stay, and people are in the first wave of making those kinds of platform and asset decisions,” he said. “In reality they are still using traditional data sources to a large extent, whether it be historical data internally or easily available digitised and structured data from external sources.
“We still haven’t gone to next-generation data analytics, where people will start to leverage video data sources, audio – text is becoming a more common data source too. There is definitely going to be another generation where people are accessing much less structured data sources such as geo-location, spatial data and try to drive insights from that. So there’s more to come.”