Skip to main content

Select your location

Search
Women in tech having discussion

Becoming a data-driven enterprise: 6 Actions for Senior Leaders

  • 05 August 2021 / By Mark Ardito
  • Cloud Modernization Data and Analytics

Most companies are “data-rich, insight-poor.” Organizations, as well as individuals, are constantly consuming data records in the form of documents, images, videos, social messages, search queries, news and so on.

Every second, 1.7 MB of data is generated per person, and based on research, the total global amount of data will increase exponentially from 2020 to 2025, from 44 to 163 zettabytes. Which really puts into perspective the vast amounts of data that have been captured and stored over time, but only 14% of companies have this information widely accessible to employees, making it impossible to use for meaningful insights. So what can executive and senior technical leaders do today to flip the script and enact change?

Recently, I welcomed several new members to our CIO community and hosted a discussion in partnership with Google Cloud, for women leaders working in the area of data and analytics. While participants came from various industry sectors, they all work in large enterprises and are engaged in significant, transformational data initiatives. Together, we discussed common challenges, actions, and lessons learned as we shift from treating data as an afterthought to data as a product.

From establishing cloud centers of excellence, to managing biases, to upskilling talent; if you are a senior leader tackling data democratization, here are 6 key learnings with accompanying actions you should take today:

Support and direction from the top is a must

Turning a data-rich enterprise into one that is insight-rich takes years, not months. It is a focused step-by-step progression supported by leaders with a keen focus on the desired outcome. However, getting data to a place of optimum value takes time and money which means buy-in from key stakeholders is critical.

The value of each use case may need to be demonstrated before moving to the next so that stakeholders understand the importance of being insight-rich. Given the time needed and scope of the undertaking, there needs to be a sizable investment and a strategic roadmap to be successful, along with faith that the ROI will be seen over time rather than at the time a single use case is complete.

Action

Support and strategic direction from the top is required. When company leadership understands the scope of the undertaking and the payoff for pursuing it, the team is empowered to accomplish the goal. All forms of data are needed across all areas of the business, everyone needs visibility to the data to aid in their decision making.

This is no longer a single IT department run operation, these needs must be managed and run cross-functionally. Consider forming a steering committee with your Chief Data Officer and including cross departmental representation. This committee can bring people across the organization together to align on the roadmap and priority of data initiatives and report on results.

This way all parts of the business are accountable and participate in driving business decisions through the use of data. Define who should attend vs present and use this time to share updates, identify KPIs that are important and critical to the success of the business, go through cleanup, review dashboards and have quality checks.

A cloud center of excellence (CCOE) is also something that helps drive momentum across the organization, and develops a real accountability model. The benefits range from developing reusable frameworks for cloud governance, managing cloud knowledge and learning, and having a center of excellence for overseeing cloud usage and plans for scale. This can make a material difference in terms of ownership and scale and the ability to continue to be agile as organizations think about how to leverage all of this technology.

Making progress while waiting for the seismic shift is possible

Data and data transformation have only been accelerating, forcing organizations to look at ways that they can continue to not only keep the lights on, but rethink their revenue stream and how they are going to continue to differentiate through the use of technology, and in particular, cloud technologies.

The impact of COVID has not only reinforced organizations' desire to look at ways to embrace technology, but the necessity of it. Everyone is on a journey to try and understand how they can differentiate themselves by using their most incredible but underutilized asset, their data.

But the process of tapping into the full potential of your organization's data, as mentioned, is time intensive. Leaders need to find ways to take advantage of the strides and progress being made along the way.

Action

The first step of any analytics journey is the collection of data, identifying the sources of the data and collecting the raw data to be processed and interpreted later. Prioritizing data collection and understanding what data to collect is extremely important since collecting everything isn't useful and can be costly. This allows you to make use of your digital applications and tools for different functions and begins to break down silos removing barriers to access or understand information across the organization.

The second step is data transformation and cleansing, which can be tremendously challenging because it involves identifying the data heuristics, models, and then actually implementing the ETL to transform. Initially, this can streamline the use of the data, provide some data unification, and be able to share with other teams.

The next task to tackle is creating relationships between data to ultimately move from descriptive analytics towards predictive and prescriptive analytics. At this point data can be leveraged quite heavily for operational and customer experience purposes, which is when you can really add value for the company and your customers. Thinking about data as a product, allows you to plan, iterate and improve upon its use in order to improve operational efficiency, product features, ultimately improving the experience of your employees and customers through more personalization and understanding the next best action.

In our 12-week engagement, the Kin + Carta Data Labs team will work with you to take that first step in harnessing your data better. We will work shoulder-to-shoulder with your team, working with a single business unit, to harness their data in a digestible and visual way. This is the first step in experiencing the business benefit of data democratization.

Trusting your data goes both ways

Trust is one of the biggest issues about data democratization. Customers trusting what happens with their data and how it's handled is one thing. And employees trusting in the analytics that comes out is another.

For customers, businesses need to ensure that processes are put into place to protect customer data but also make their customers feel that sense of security. For employees, without understanding the data, small differences in the information at different times can create distrust that the information they see is accurate.

So how do organizations handle protecting customer data and access the truth and the literacy of what employees are accessing?

Action

In certain business sectors, there can be a risk that personal identifying information (PII) is inappropriately being accessed or transmitted. It is essential to know what data is being processed and where who has access to it. Based on that information you can better govern where you have confidential or sensitive information, like PII.

This will identify where there is a need to put gates in place to protect or stop the leaking of private data. Today there are IoT devices that can manage secure communications and devices that capture consumer data without including PII.

In terms of trusting the output of what you receive and the intelligence that goes around that, you have to believe that the process that you built works. Continuously question insights if they seem too good to be true by isolating the data, making more detailed measurements, scrubbing the data, and improving wash routines.

Though it has advantages, AI also comes with risks

As AI continues to mature and demonstrate its value to the business, it will become an integral part of the business across most, if not all, industries. There is a risk that companies will enter the “creepy zone” in their interactions with customers based on buying behaviors.

At some point, the communications stop being helpful and start being intrusive. As one participant observed, “We love getting the alerts about our oil needing to be changed, but at some point, those alerts start to have a creep factor.”

Action

For customers, sometimes being so accurate can feel intrusive even though it's respectful and it's private, you're using their data that they've consented to, to produce insights into their behavior that only they can see. As helpful as it may be, if they lack the understanding of how this data is protected and produced, it can still feel invasive.

Start thinking about how data and AI will be used, especially with customers. Consider the meaning of “ethical AI” for your company and determine what needs to be in place to accomplish that. Continue to monitor the use of AI so that inappropriate usage can be stopped early.

Biases built into data and AI could become an issue

We also must remember that AI technology is coded by humans. Data generally comes from humans, which naturally have bias tendencies. With humans at the root of the technology, unconscious biases can get coded without getting noticed.

A notable example offered by several participants was AI commonly used by HR departments. The system culls resumes before being seen by humans and can eliminate viable candidates based on biases coded into the algorithms. How do you make sure the technology that you're using doesn't have bias in it?

Action

Raise consciousness of biases and unexamined assumptions, and assert the importance of avoiding them in systems. Create training programs that employees must complete to help cut through biases and identify new ones.

This is not only helpful when it comes to building more inclusive products and experiences but it is imperative when it comes to recruitment. If your employees conducting interviews are aware of their biases they can make better hiring decisions.

It really comes down to the fact that you need more multidisciplinary people making decisions within your organization, “If your algorithm is a mirror of humanity, you failed and your algorithm is biased.” – Kyle Hundman, Data Science Manager, American Family Insurance.

Staffing for needed data skills is a challenge

People, process, and technology are the fundamentals of a successful business and people are at the forefront of a company’s success. Your employees continuously push the business forward, but as the pace of technology outstrips the pace of training, IT leaders are seeing glaring skills gaps.

Data science skills are at a premium. Hiring people with the right skills can be challenging because of the high salaries being sought and the tendency to jump jobs to take advantage of better offers. So how do organizations remedy this?

Action

Reinvest in employees, when possible, by conducting training programs to upskill them to new roles. Your staff will view this as a resume-building opportunity and a chance to work at the leading edge of data science.

Potential candidates are also looking at enterprises based on their technology stack when making decisions on their next career move. Are you using the technologies that will continue to help them grow that muscle from their career? Look at it through both lenses in terms of attracting and retaining talent and then reskilling your existing talent. It's a balancing act.

In summary, the participants brought much insight to how data is driving decision making and the need to tread data as a product. But at the core of the discussion was one key insight. Moving the data needle to “insight-rich” is not solely an IT undertaking. Collaboration across business lines is essential.It extends accountability for data and decision making from that information to the business. Business units can provide the use case and priorities on data initiatives most important to business leadership, offering a taste of insight-richness that makes a difference.