Project Management Institute

Analyzing the Future

3 Ways Data Science Will Change Project Management

img

GETTY IMAGES

Luke Desmond, PMP, PgMP, PfMP

The wave of big data and other data-science-related knowledge is turning into a flood. The economy is rapidly transforming into a knowledge-based one. In fact, the four largest companies based on market capitalization—Apple, Google, Microsoft and Amazon—are all data-driven organizations. How will this data science storm affect project, program and portfolio management? I believe there will be three main ways:

1. NEW ROLES

Within five years, almost all project management offices will seek to have a data science capability either permanently assigned to them or available when needed. However, the world already is facing a global shortage of data science professionals. Although many people have IT skills transferable to the data science world, very few have the five required data skill sets: business acumen and domain knowledge, computer programming, an understanding of statistics, the ability to build and maintain databases, and storytelling. This latter skill—the ability to convert the results of big data analysis into a message that stakeholders understand—is the largest gap. However, program and portfolio managers with a broad set of domain knowledge, business acumen and communication skills are ideally placed to step in and fill this gap.

The ability to convert the results of big data analysis into a message that stakeholders understand is the largest skills gap.

On the flip side, automation, machine learning and data science will disrupt the role of project coordinators, and their opportunities might shrink. Since project coordinator roles are where many project managers gain their practical experience, upskilling or retraining for project managers and coordinators will be vital.

2. NEW TOOLS

New products will make it easier for project leaders to analyze big data for quantitative elements. This includes the ability to correlate data to identify root causes of problems, the ability to perform a “what if” analysis, and the ability to help project managers prioritize tasks and milestones so they deliver the most value.

These new tools will allow project leaders to conduct a wide range of data analyses that span from simple to very advanced:

New tools will allow project managers to conduct a wide range of data analyses.

■ Descriptive analytics involves taking advantage of data aggregation and data mining techniques to provide insight into the past. For instance, a software firm could use it to discover it sold 50 subscriptions last month. It is useful for understanding profitability but little else.

■ Diagnostic analytics tells us why something happened. In the software example, that could involve aligning sales with positive online reviews of the product to help the organization understand why the subscriptions were purchased.

■ Predictive analytics involves using statistical models and forecasting techniques to understand the future. For instance, the software firm could forecast customer adoption based on growth in its target industry, demographics and competition.

■ Prescriptive analytics is using optimization and simulation to advise on possible scenarios and answers. At this sophisticated level, think of the software company analyzing data in real time to resolve a customer's technical problem before the customer even reaches out.

3. NEW MEASURES OF PROGRESS

Portfolio managers might be asked to assess their organization's level of maturity in using data. For this emerging area of assessment, the following categories can be used:

New: An organization that has barely identified the need for a data science strategy. Executive buy-in is lacking.

Investigation: An organization that has recognized it needs to develop a data science strategy but is in the early stages. It might be running some experiments, but it is not making substantial headway.

Early adoption: An organization that has made great progress. Pockets of the organization might have a strong data science capability. We would expect to see governance beginning to mature. Breaking through this barrier is a challenge for most organizations.

Extensive adoption: An organization that likely has a chief knowledge officer and an enterprise-wide strategy. Any manager purchasing a new software application would first have a discussion about how it integrates into the organizational strategy.

Leader: An organization born into a digitized world. Think of Uber and Airbnb, which rely on data to link consumers with an inventory of transport or accommodation options. Amazon is another data science leader, using data to make investment decisions around new business models. Major issues holding organizations back from reaching this state include funding, resourcing and company culture.

You don't need to become an expert in coding, but you should be able to have a meaningful conversation with your stakeholders and make sure you are aligned with the organizational strategy around using data as a differentiator. PM

img Luke Desmond, PMP, PgMP, PfMP, is a digital transformation leader at Cisco, Overland Park, Kansas, USA.
This material has been reproduced with the permission of the copyright owner. Unauthorized reproduction of this material is strictly prohibited. For permission to reproduce this material, please contact PMI.

Advertisement

Advertisement

Related Content

Advertisement