Toward Data-Driven Decision-Making

In this writeup, we explore the increasingly prominent role of data in effective decision-making.

Author:  Data Strategy Professionals team  |  Post Date:  Mar 21, 2024  |  Last Update:  Mar 30, 2024  |  Related Posts

While some organizations possess a high level of data literacy that supports data-driven decision-making, many organizations struggle with logistical challenges and fail to reach their potential in terms of using data to generate value.

Contents

• Challenges of Becoming Data-Driven
• Examples of Data-Driven Organizations
• Cultivating a Data-Driven Culture
• Conclusion
• Related Posts

Challenges of Becoming Data-Driven

Data-driven decision-making has increasingly defined organizational success over the last decades:

• MIT Sloan's 2010 survey of over 3,000 executives found that "top-performing organizations use analytics five times more than lower performers"
• A 2011 study surveyed 330 firms about their business practices and found that data-driven decision-making was correlated with 5-6% increased output and productivity relative to firms not using data-driven decision-making
• The World Economic Forum (WEF) highlighted the newfound dominance of data-driven tech companies amongst the top five publicly traded companies between 2013-2018
• Management consulting firm McKinsey & Company observed in 2018 that their less data-driven clients had a weaker understanding of their business opportunities and generated less "insight-based value" from their data than their more data-driven clients
• A 2020 study exploring the relationship between data use and innovation proposed that "the secret of a firm's success lies in how it can scientifically analyze customer data"

Of course, becoming data-driven is easier said than done. The 2023 Data and Analytics Leadership Executive Survey from consulting firm Wavestone asked executives from 116 Fortune 1000 companies about their business practices. The results highlight how "becoming data-driven is a long and difficult journey" primarily due to people and process issues such as: leadership receptivity to new ideas, the alignment of team skills with business objectives, and communications.

Examples of Data-Driven Organizations

Two examples of high performing organizations that excel at making data-driven decisions are Netflix and Google. Data Analysts at Netflix approach data as a tool "to inform a wide range of questions" related to the priorities of their department. For each of Netflix's six verticals, the Data Analytics team is "given freedom to choose their projects and [is] responsible for prioritizing the ones that will have the most business impact" in an effort to "leverage their unique skills to make Netflix better."

Meanwhile, Google annually collects data from all employees to understand "what a great culture would include: innovation and autonomy, forward thinking, teamwork." Programs are then re-designed later in the year to address all areas in which feedback was poor. Claims are also expected to have "data to back [them] up," epitomized by Google's Project Oxygen which involved research into the efficacy of managers at Google after engineers shared skepticism about their managers adding value to the company.

Netflix and Google can prioritize empowering their staff with high levels of autonomy to maximize data-driven decision-making because they've largely solved the basic people and process barriers highlighted in Wavestone's report. For the majority of organizations, one activity that could be beneficial is to undertake an organization-wide program to enhance data literacy.

Gartner's 2022 Chief Data Officer Survey highlighted poor data literacy as being amongst the top challenges for CDOs, prompting Gartner to create a simple and practical guide to data literacy:

1. Baseline assessment — establish where the gaps are in data literacy such as an understanding of basic statistics and what outputs key systems produce
2. Training program — find staff already confident with data, identify communication barriers stopping data-driven decision-making, and trial a proof-of-concept workshop; online courses and other existing guidance can also be referenced
3. Leadership role modeling — data-driven decision-making should permeate through business meetings and be visible to all staff
4. Measure initiative effectiveness — measure program outcomes against quantitative success criteria such as the proportion of staff who've passed a certification, collect staff feedback, and support staff to immediately start applying what they've learned

Fostering data affinity will likely be on the list of projects the Data Management function should undertake to encourage data-driven decision-making. Organizations must also surmount the common people and process issues that hinder data-driven decision-making. To start on this journey, Harvard Business Review suggest that data practitioners "start small, benchmarking your performance, documenting everything, and adjusting as you go." You may be interested in reading Asana's step-by-step advice on how to make more data-driven decisions.

Cultivating a Data-Driven Culture

As we started discussing in our mid-March Newsletter, a data-driven culture is essential to enabling data-driven decision-making. While there are different definitions for a data-driven culture (as discussed in these sources: 1, 2, 3), we broadly define it as an active organization-wide coordination of data practices to support data-driven decision-making.

Google has invested heavily into developing a data-driven culture. Their annual Googlegeist survey collects data from all employees on "what a great culture would include: innovation and autonomy, forward thinking, teamwork". Programs are then re-designed to address all areas in which feedback was poor. Claims made by Google employees are also expected to have "data to back [them] up". Google's Project Oxygen empirically researched the impact of managers at Google after engineers questioned the value of managers. Results showed that more highly rated managers increased happiness and reduced turnover amongst their teams, and helped to identify the characteristics of high performing managers to inform training for lower performing managers.

But not all organizations are in a position yet to build data into the "DNA of [their] culture" like Google has done. Only 24% of consulting firm Wavestone's 2023 Data and Analytics Leadership Executive Survey respondents saw themselves as data-driven while only 21% believe they've developed a data culture. Amongst the key conclusions drawn was that "data transformation and becoming data-driven are about helping everyone use data." One of the key steps towards everyone using data is high organization-wide data literacy.

Gartner's 2022 Chief Data Officer Survey highlighted poor data literacy as being amongst the top challenges for CDOs, prompting Gartner to create a practical guide to data literacy:

1. Baseline assessment — establish where the gaps are in data literacy such as an understanding of basic statistics and what outputs key systems produce
2. Training program — find staff already confident with data, identify communication barriers stopping data-driven decision-making, and trial a proof-of-concept workshop; online courses and other existing guidance can also be used
3. Leadership role modeling — data-driven decision-making should permeate through business meetings and be visible to all staff
4. Measure initiative effectiveness — measure program outcomes against quantitative success criteria such as the proportion of staff who've passed a certification, collect staff feedback, and support staff to immediately start applying what they've learned

McKinsey & Company put it well:

"Culture can be a compounding problem or a compounding solution... The technology [of data analytics], after all, is amazing. Imagine how far it can go with a culture to match."

You may be interested in McKinsey & Company's Why Data Culture Matters for insights from leading cross-industry executives.

Conclusion

An organization that is more committed to data-driven decision-making will generate significantly more value than a comparable organization that uses data less effectively. Learning from high performing data-driven organizations can be useful for finding evidence-based practices to adopt or improve within one's own organization.

Becoming data-driven first requires an organization to assess the current state of its data practices through a Data Management Maturity Assessment (DMMA). An assessment will allow your organization to identify technical, people, and process issues that can be addressed.

For more help with Data Strategy best practices, we recommend checking out our article on this topic, as well as our Data Strategy Workbook and Document Checklist.