by Isaac Sacolick

5 primary activities for CIOs transforming to a data-driven organization

Oct 21, 2019
CIODigital TransformationIT Leadership

No one said transforming to a data-driven organization was going to be easy...

It’s become popular and almost cliché for CIOs, chief data officers, and IT leaders to proclaim the need to transform into a data-driven organization. We want leaders to leverage data and analytics in their decision making instead of the loudest or most experienced executives asserting decisions based on intuition and bias. We recognize that data is the new oil and that our businesses need to compete and differentiate with customer-facing, operational, and predictive analytics. Lastly, we know that data is a critical asset that must be managed, protected, and leveraged appropriately while still democratizing data so that there is increased utilization across the organization.

CIOs, CDOs, and IT leaders need a data strategy. Given the tall objectives and organizational needs around data, having a strategy and execution plan is critically important to empower people and drive change. But data strategies cut across many organizational needs, skillsets, best practices, partnerships, regulations, and technologies. At events such as SINC IT leader forums where I speak on a mix of digital transformation topics, I hear and answer a variety of questions on what goes into a data strategy and how to develop one.

Here are five focus areas for CIOs and IT leaders to establish a data strategy and execution plan.

1. Identify underserved departments with significant data opportunities

When I was a business unit CIO at McGraw Hill, one of our more significant opportunities was working with the marketing departments who were experimenting with a mix of digital marketing tools to reach prospects and nurture leads. To make sense of what experiments were yielding the most promising results, they were painstakingly integrating data into a multitude of spreadsheets and analyzing the results.

In IT, we viewed this as a messy, inefficient ways to integrate and analyze the data on a regular basis. We replaced a manual, error-prone process by deploying Tableau to a few data-savvy people in the marketing group and asked IT to help them automate the data integration and provide access to internal data sources.

Transformation programs are both top-down and bottom-up change programs, and it’s essential for CIO to establish a bottom-up analytics capability. IT can help with activities like data management and dataops, but the subject matter experts in the line of business are best equipped to ask questions and self-service their answers using citizen data science and data prep tools.

2. Enable experimentation with agile analytics practices

Applying agile practices in analytics and machine learning initiatives may be more critical for driving results even compared to agile software development.

That’s because there’s an inherent discovery process for analyzing data. Someone starts by asking a question that can help prioritize efforts, realign resources, rank sales opportunities or drive other business-impacting decisions. When analyzing the underlying data, it often doesn’t yield a straightforward answer.

  • The discovery process usually identifies missing data, data quality issues, new ways to segment the analysis, and additional discovery questions.
  • The data analyst may elect to implement a data visualization dashboard only to realize that alternative and complementary views are needed to tell the story.
  • A data scientist may choose to implement a machine learning algorithm only to find that it underperforms, and another algorithm may work better.

Applying agile data practices enables the feedback loops required for experimentation. After implementing one approach, look to capture learnings in the backlog and decide which discovery and implementation options are worth prioritizing.

3. Train decision-makers on leveraging analytics

If you build it, will they come? You already know the answer to that question is, “No.” Now ask, if you build the data visualization and provide access to it, will decision-makers know how to use it to drive decisions?

The answer again is, “No”, even as data visualization platforms such as Tableau and Microsoft PowerBI have enabled organizations to develop intuitive dashboards. Sometimes, the dashboards implemented miss on many data visualization best practices making it difficult for end-users to use them intuitively.

But even when dashboards are easy to use, developing business processes to leverage the analytics, training programs to educate decision-makers on the underlying data, and establishing a change management program to gain adoption are all still required.

4. Introduce proactive data governance

A CIO once told me that they were struggling to introduce basic data governance practices before introducing data, analytics, and self-serve capabilities. He believed that introducing analytics before governance is like leading with the cart before the horse.

In my experience, it’s nearly impossible to gain traction on data governance without leaders and executives seeing the data and demanding more analytics capabilities.

That doesn’t mean sidelining the necessary data governance including establishing usage policies, implementing data security, and publishing data catalogs. What it does mean is that CIOs have to drive data governance programs in parallel to analytics enablement initiatives. There isn’t a cart or a horse, and the demand for analytics should help expose the need and value to address dependent areas of data governance.

5. Challenge executives to use dashboards at strategic meetings

The two most abused technologies used for executive decision making are Microsoft PowerPoint and spreadsheets. Both these tools require many manual steps to go from the data source to the presentation and are thus inefficient, error-prone, and are easy avenues to introduce bias.

I speak at conferences on the CIO’s role in challenging the C-Suite. Changing the tools and methodologies used for making executive decisions is tremendously challenging but critically important for CIOs to challenge the status quo. Answering on-the-fly complex questions about the business and reviewing analytics in real-time are competitive differentiators, and both can’t be achieved using tools that largely present static data views.

That means helping executives understand and leverage real-time dashboards during executive and other meetings where strategic decisions are discussed.

No one said that transforming to a data-driven organization was going to be easy.