From the monthly archives: March 2012
Today’s post comes from Warwick Graco, a founder of the IAPA and one of Australia’s leaders in Government Analytics:
Analytics like all disciplines is not immune from the effects of the Digital Revolution. Indeed its very existence is due to the rise of Big Data where many organizations are now drowning in data as a result of the rise of social media and the use of digital mobile devices. Furthermore, very little of the data collected by organizations has been converted to knowledge and even less to intelligence in terms of identifying the important insights that should underpin the decisions taken by managers.
One of the other noticeable effects of the Digital Revolution has been the compression in the timeframes between the collection and processing of data and the dissemination of results to decision makers. One has only to go back a few years where those doing analytics had the luxury of months to develop and deploy models. Now that time frame is rapidly decreasing to a few days and in some cases to one day or less. In addition to very short time-to-market requirements, those doing analytics have to still produce quality solutions that provide value for money. Return on Investment or ROI and the like are still critical requirements for all those who analyse data and report results.
The above all adds up to a demand that has become a cliché ie the need for agility. All who do analysis and reporting these days have to be highly skilled as well as be flexible, adaptable and available. We live in a world that is shrinking in space and time as a result of the Digital Revolution. It no longer matters where people live, work and play as long as they are connected and can communicate in cyberspace.
This begs the question of how one becomes agile in a digital world. There are many spokes to the wheel with this issue. It is suggested that one important spoke is the need for multidisciplinary, integrated teams that can take data and convert it into products that meet the needs and expectations of users.
The type of specialists required in these teams will vary from issue to issue but typical skills include:
• Intelligence Specialists who collect and analyse information on issues to determine opportunities and threats. The intelligence generated informs risk analysis
• Risk Analysts who identify the risks with particular opportunities and threats and the mitigation measures required to reduce their negative consequences. For example, a cyclone is a threat that has many risks
• Profilers who identify the modus operandi of those who are either threats or opportunities and their defining attributes. An example is profiling people who steal the identities of other citizens to identify their signatures
• Miners and Modellers who explore data for new insights and who develop classification and prediction models such as those that identify customers who will churn in specified timeframes
• Data Analysts who extract, clean and manipulate data required by those who do analysis and reporting
• Business Intelligence staff who takes the results of analysis and present it in suitable forms for the consumption of managers. This function is critical because managers judge the value and relevance of analysis by what they see and hear. Therefore, it is very important to get this requirement right
Traditionally these specialists have tended to work in silos and therefore in isolation from each other. This has led to ignorance and misunderstandings of what each class of specialist does and to less than optimal results being achieved because the full capabilities of these highly skilled people are not brought to bear on the problem at hand.
One solution to this challenge is form integrated teams made up of these specialists so that there is both unity and economy effort, that all the important skills are focused on the problems requiring resolution, that there is teamwork and team learning so that all profit from the work done and that learning can be applied to future problems and issues.
High performing teams such as these are agile and proficient. They can turn around tasks in very short timeframes. This is an imperative with issues such as customer churn, credit-card fraud, detection of improvised explosive devices in a theatre of war and the effects that a flood can have on a rural community.
There is much to commend in having integrated, multidisciplinary analytical teams. They are often worth many multiples of what they cost.
Warwick Graco has worked in analytics for nearly 20 years starting in this occupation when the term ‘analytics’ did not exist. He has seen the profession grow from its very small beginnings to what it is today.He has worked in defence, health and lately revenue collection where he heads a small team responsible for operational analytics. His academic interests include analytics, organizational change and organizational decision making
The buzzword of the year seems to be “Big Data”. There is a massive wave of promoters of the term, and there are inevitable detractors. There is also the issue of exactly how to define it. What follows is the A1 view on Big Data.
It is real, it is a game changer, and it is here to stay. It is no one thing, and its definition, both quantitiative and qualitative, is rather fluid. Nevertheless, some basic truths apply: Big Data is not a brand name. Neither is Big Data a tool, a business process or a solution. It isn’t even an idea as such. In fact, Big Data is best understood as a problem. Not a problem as in “trouble”, but a problem in the sense of a challenge or puzzle, or more precisely a growing family of problems that we are increasingly forced to grapple with. It’s a problem that does not come with an automatic solution, although there are a growing number of tools to help roll it around.
The A1 angle on this is: you cannot outsource your investment in Big Data any more than you can outsource your own education, or exercise, or being a patient in a surgery theatre. In this sense, what is true of Big Data is also true of Analytics.
Getting Big Data right means getting Small Data even righter. The sort of business that can get value out of Big Data will be one already getting value out of Small Data. Without the business fundamentals in place, Big Data will produce only Big Nonsense. Alternatively, if the logic is there, then Big Data will enhance an existing value-adding framework.
So: small data first, then big data. And before small data, tacit data, which you can always get your hands on, even if you have trouble wrangling the electronic stuff. And before all of those: logic, and human infrastructure. A well understood, well defined business model with well defined intelligence objectives. And incentives, with staff capable of navigating such an environment, managed by a sponsor possessed of the A1 “holy trinity” of adequate influence, appropriate motivation and sufficient understanding of the value, role, and needs of Analytics under their command. Is this too much to ask for?
I should also probably mention tools. Maybe. Last. Do they matter? Of course. So does oxygen. But it is ubiquitous, effectively free, and we take it for granted…
A1 is a proud supporter of the AIPIO Collective Forecasting Competition, hosted on Presciient’s new collective forecasting platform System II.
A beginner’s guide may be found at the top of the page.
Collective forecasting and related methods such as prediction markets represent the area of analytics that we call Tacit Data Mining, and allow the extraction, deployment and analysis of the most vital data in the organisation, which lives in people’s heads. It also provides the ultimate data fusion platform, fusing all available data through human filters to provide powerful strategic decision support.
Collective forecasting allows accurate forecasting of future events, and also can condition those events on possible actions, thus providing a powerful decision support. It identifies the consistently most effective forecasters, acting as a filter for the most insightful and prescient members of staff or the public.
It has application in any strategic decision support domain.
The competition at hand has 3 expiry dates for predicted events: in April, July and October, each has prizes for 1 month ahead, 1 week ahead and 1 day ahead. The July and October expiries also have 3 months ahead prizes, and a six month ahead prize for the October expiry.
The one-month ahead April expiry deadline is tomorrow, so don’t delay, register and put in your predictions.
Follow Us!
