Data Driven Decision Making in the age of AI

"Executives want greater speed and sophistication in their decision-making, but most say their ambition is greater than what their organisations are ready for."

In the first of four articles we have been writing with Forbes, we explore what it means to be a truly data-driven organization, and how the power of today's analytics techniques and technologies are impacting decision making at all levels of the enterprise.

There can be little doubt that the best laid plans of the c-suite are constantly being challenged by the complex, disruptive realities of competition and the socio-economic environments they operate within. As my colleague Paul Blase has noted, 

[while] major advancements in data availability and analytic techniques are now available to executives to help them weigh various options… many companies fail to use data and analytics to connect decision making, planning, and actions—leaving good decisions to fizzle before they can be acted upon, or plans too brittle to adjust as the environment changes"

PwC’s Global Data and Analytics Survey 2016: Big Decisions?  reveals that only two thirds (61%) of global executives feel that their own companies’ decision-making is only somewhat or rarely data-driven, and that only 29% of companies regularly use predictive analytics to inform decision making.

We suggest three drivers should be acknowledged when looking to improve the effectives of your organization’s decision making:

1. Data-driven companies take a more holistic approach to decision making, looking backwards when needed, but also using predictive and prescriptive analytics to model the future.
2. Executives want decision-making to be faster and more sophisticated but routinely acknowledge significant gaps across both dimensions. Improving both speed and sophistication helps maximise the return on investment for data and analytics, but it is not inherently straightforward nor easy.
3. Best in class organizations foster a balanced mix of mind and machine - augmenting human judgment with machine algorithms and artificial intelligence to create better outcomes.

As my colleague, Dr Anand Rao has written, the term “artificial intelligence” is often misused. As a branch of computer science, AI includes a variety of disciplines including machine learning, natural language processing, simulation modeling and robotics. Each of these techniques can be used to solve particular kinds of problems and support decisions.

It may be useful to consider AI as supporting decision making in three distinct ways:

1. Assisted Intelligence, such as robo-advisors, that use data enrichment, agent-based simulation and behavioral economic techniques to offer cost-effective, sophisticated, personalized advice upon which an end user (ie a human) can make a decision;
2. Augmented Intelligence, such as we might find in precision agriculture, that augments sensor, satellite, weather station, and visual data to support decision making for crop yield management; and
3. Autonomous Intelligence, such as driverless cars, in which AI takes over the performance of certain tasks traditionally done by humans – making autonomous decisions based on sensor, satellite, GPS and visual data.

While data and analytics can make decision-making faster and more informed than ever, the process will never be completely hassle-free. A strategy through execution approach, combined with early stakeholder alignment and the use of pilots and stage gates, can help translate decisions into competitive advantage and avoid costly delays. What challenges does your company face in decision-making, and how are you addressing them?

Check out the full Big Decisions Survey at www.pwc.com/bigdecisions.