Why and how CFOs can leverage big data
At the CFO Summit on the Gold Coast earlier this month there was a big focus on data, and data analytics, and business intelligence, and big data, and digital disruption, and developing a digital mindset. Perhaps so much as to make some people’s heads spin!
A very astute point by the CFO of Exact Mining raised a painfully obvious issue for almost everyone else in the room. The finance team is typically not the place to find people with the necessary data analysis skills to be able to mash up and reinterpret ‘big data’. Some might have felt these discussions were best placed in a forum for CIOs or niche specialists like data scientists. But surely interpreting data, concepts like forecasting, predictive, prescriptive analytics, indeed identifying commercial opportunity from data, is absolutely the domain of the modern CFO?
It’s not really about big data, it’s about leveraging big data, indeed any digital asset, precisely to aid decision making and for commercial gain and insight. So how can the modern CFO access or build skills within their team that make it feasible?
Dr Michael Rosemann was quick to point out that PhD students at universities such as his own at QUT were only too willing to engage with corporates to engage on these very services. And indeed this plays directly to the summit theme about digital disruption and a digital mindset. Services such as data sciences/data analytics (perhaps what we used to call actuarial skills) can legitimately be considered a commodity in today’s global networked economy. Just google “data science as a service”!
For instance check out Kaggle. This is an online community of data scientists who actively compete to out-clever each other. If you have a load of data and a question you want answered you sign up to Kaggle and create a competition. It’s not free of course, every competition needs a prize, but you can decide how big that prize is. I think it will be no surprise to anyone that the combination of an interesting puzzle and a juicy prize generates a greater response to the competition. At time of writing there are 189 teams competing for a $500 prize, 1218 teams competing for $10,000, 606 teams for $15,000-$16,000 prizes and 163 teams competing for $100,000. So the spread is huge but also note that the interest/opportunity sweet spot seems to be at that $10-15k prize point. That’s not a lot of money for a transformative answer to a cost or revenue quandary.
So what are some examples of other people’s competitions?
- Forecasting the use of a city bike sharing system. (For the fun and knowledge!)
- Probabilistic distribution of hourly rain given polarimetric radar measurements ($500)
- Predict annual restaurant sales based on objective measurements ($30,000)
- Identify signs of diabetic retinopathy in eye images ($100,000)
One of the conversations my colleague Jonathan Marcer had at the CFO Summit last week was with a state government agency handling vehicle incidents and providing support and preventative measures from operational to policy advice. Data that adds context and meaning to an incident, including mitigating risk and preventative operations, is increasingly digital and accessible, whether streamed from onboard technology, infrastructure, other agencies, collected through applications, gathered through case management activities, or available from context brokers. All this can come together to provide operational, management decision makers with “situational awareness”. Such organisations also leverage research on causal factors, likelihood, etc to help guide policy and procedural best practise. That work is often undertaken by specialists.
The question debated at the CFO forum was how can the data and the research be leveraged to develop predictive algorithms supporting context aware case management? Maybe a combination of Kaggle or the research team provide the algorithm; Microsoft Azure Machine Learning, a powerful cloud-based predictive analytics service, provides big data mining; and technologies such as XMPro delivers the operational intelligence and actionable delivery.
So what question have you got that you have all the data for but don’t know how to answer? And how much is that answer worth to your business? Take it to Kaggle and get that answer. When you have the data and the algorithms you need a tool to bring them together to crunch the numbers and visualise the results. That’s where we can help.
[cd-form type=”contact-2columns” title=”Need an answer?” action=”http://analytics.clickdimensions.com/forms/h/aQFTAdPgQOEOW6iXUblDtg” button=”Make an enquiry” thankyou=”Thank you for your enquiry. We’ll be in touch shortly.”]