Decisions in the Age of Abundant Data

A summary of bite-sized insights and views about decision making from fresh articles around the web.

Over the last several months at Cause Analytics we have been scouting the web to bring you some helpful articles from experts in business intelligence. Today, our team will summarise several insights we have collected in our Google+ community in order to save you time and help your organisation with decision making.

You probably know of the debate between speedy decisions based on intuition versus slower decisions that draw upon evidence and rational deliberation. It's a complex debate because both sides have a role to play in various contexts, but in the 'Age of Abundant Data', business intelligence can aid you on your quest for faster results. As analytics and algorithms increasingly offer practical predictions, executives and managers will be able to act upon these predictions to communicate decisions reflexively, and perhaps, intuitively.

Data-driven prediction, rather than intuition

In an article we shared from Fleet Owner - a magazine for commercial-trucking fleets, the the predictive capabilities of analytics is discussed as central to the return on investment in big data and business intelligence applications. To be able to make decisions about problems that have yet to arise is the power of what Richard Holada, VP-BI/AA for IBM's Software Group calls, "predictive real-time decisions". He explains that the "unknowns" in a business environment are traditionally addressed through an old model that requires sensing and responding based on business instinct and intuition which are then supported back-office analysis, much later on. Predictive analytics is the future because, as he notes, we can act in near real-time (on-demand) based on facts, rather than intuition alone. This predictive capability is becoming more accessible, transforming complex business information through richer imagery, charts and graphs. With data visualisation, better decision making is made possible for many more people within the organisation.

Understand the what, why, and how of a problem

If you're aware of Procter & Gamble's ability to tap into Big Data, then you may well be excited by their data dashboards and shared models that enable social collaboration between senior managers throughout the world. The company's CIO, Filippo Passerini, is rapidly growing the number of staff with expertise in business analytics and Big Data - with P&G's 58,000 employees accessing business intelligence through “cockpits” (dashboard displays) that link to common data sources and analytic models which answer "the What, Why, and How of a problem". At Practical Analytics you can read about P&G's models:

• The "What" model, which is said to focus on data such as shipments, sales, and market share. Passerini talks about "What" as the problem itself — is market share stable or has it shrunk two points?
• "Why" looks to the cause of the problem — was it a bad TV advertisment, out-of-stock shelves, or a competitor’s new product or price? The "Why" model highlights sales data by country, territory, product line, and store levels, as well as drivers like advertising and consumer consumption, factoring in region- and country-specific economic data. 
• The “Actions” analyses look at 'levers P&G can pull', such as pricing, advertising, and product mix, and provide estimates on what they deliver.

Understand people to predict the future

Predicting the future, according to leadership advisor Mike Myatt, requires leaders to understand people and their motivations. At Forbes, he recommends that leaders engage people of influence and observe the decisions they make, because, as the old axiom says, “talk is cheap”.  He adds, "There is no reason to be surprised by people’s behaviour, unless you've failed to observe it." Myatt tells us that, "few things will help discern the future of a career, project, product, or company like taking a close look into the character and commitment of the people driving them." 

Predictive analytics may require a culture in which people already use business intelligence. John Lucker, the head of the advanced analytics and modelling practice at Deloitte Consulting LLP, tells TechTarget that a lack of executive support can stifle such efforts, especially, when leaders cannot recognise the impact that predictive analytics can have on the bottom line. He suggests that endorsement from the top ranks can kick-start a culture of creative thinking, fresh ideas and data-based decision making: "better efficiency and smarter decisions, will come to fruition if we simply insist on making decisions based on facts."

Data Stewards: contextualising and validating data

For organisations that have woven data into their culture, the emergence of Data Stewards follows naturally. Microsoft explains that, "data stewards are business users with expert knowledge of business processes and how data is used within those processes." It recommends this role as as the “go to” person for data related enquiries coming from co-workers and managers seeking to "validate the accuracy, completeness, or validity of data within a business context." 

Beware hidden bias in Big Data and your brain

It has been said that "Interpreting data is more of an art than a science." Kate Crawford, a principal researcher at Microsoft Research reminds us in her post at HBR that "data and data sets are not objective; they are creations of human design. We give numbers their voice, draw inferences from them and define their meaning through our interpretations. Hidden biases in both the collection and analysis stages present considerable risks, and are as important to the big data equation as the numbers themselves." 

Analysts might also take heed: "the brainier you are, the better you can twist facts to your own pre-existing convictions." These findings come from a recent study by Yale's Dan M. Kahan and colleagues, and was featured by NPR as revealing that, "the better you are at reasoning numerically, the more likely you are to let your political bias skew your quantitative reasoning." 

We may have a tendency to think we are good at making predictions because we conceive of the world as more deterministic than it is. This can result in overconfidence about knowing what the future holds for our organisation. An example of this cognitive bias toward assuming patterns is noted in the Financial Times, "If a coin is tossed four times in a row, for example, and each time comes up heads, there is a bias towards expecting heads on the fifth turn." Economist Dr Ben Broadbent of the Monetary Policy Committee was quoted in the FT article, saying, “Unless reminded with hard evidence, people seem genuinely to believe that their prior predictions were different to what they actually were. The tendency to absolve ourselves of past predictive error is deep seated and, unless consciously checked, automatic".

Taking action in the face of the unknown

At times, data-driven predictions may only reveal enough for us to trust in the next step, without providing details of the path further ahead. Nick Tasler, CEO of Decision Pulse writes for HBR that "decision-making will always be an exercise in coping with an unknowable future. No amount of deliberation can ever guarantee that you have identified the 'right' option." He urges us to act on the best option or find ourselves caught up in planning, strategising, and number-crunching every possible consideration. 

Observe, orient, decide, and act

Data Strategist at Cause Analytics, David Pidsley agrees that there is often greater value in 'doing' rather than 'planning' when it comes to interpreting Big Data. He goes on to say, "Sticking to one path is easier when your organisation acts on business intelligence, which is founded on trusted data." When it comes to winning the race, military strategist John Boyd has a model called the OODA loop that was derived from his experience as a fighter pilot and the rapid decision making required in combat situations. Alistair Croll explains, at O'Reilly, how this looping processes relates to Big Data: the model consists of "observing your circumstances, orienting yourself to your enemy’s way of thinking and your environment, deciding on a course of action, and then acting on it." Croll notes that Big Data is about feedback and acting upon results in ways that can be observed and collected as data to be analysed in a looping process of "continuous optimisation that affects every facet of a business."

How to win the race with business intelligence

Developing predictive capability must start with integrating business intelligence into your organisation's decision making. Expert analysts are as prone to bias as anyone, so seek to identify hidden convictions and beliefs based on prior predictions that may skew findings. Appointing a Data Steward may ensure that findings fit into the business context.

Decisions based on trusted data can be used to explain and justify actions to stakeholders across the globe, using data visualisation and dashboards like those used by P&G. Ultimately, actions must be taken using the business intelligence one trusts, in uncertain conditions. It's up to leaders to encourage data-based decisions and actions throughout their organisations. Doing so will create the kind of feedback loops of continuous optimisation that are vital to changing business environments. 

At Cause Analytics, we want you to anticipate change with us in a continuous loop of accelerating growth and adaptation. Our data suggests you will win with us, if you so desire.