6 steps to build a data-driven company, according to experts

Trends can be observed and patterns identified from the vast amounts of customer and operational data that businesses are generating. And those who know how to take hold of and use those insights to chart a course for their company will be ahead of the game.

Becoming data-driven is not a ‘nice to have’

There are handsome rewards for companies that insert big data and analytics into their operations. According to PWC, these organizations enjoy 5% more productivity and 6% more profitability than those that don’t.

Meanwhile, Harvard Business School found companies that make data-driven decisions are more confident in those decisions, proactive, and able to realize cost savings. The confidence piece is backed up by founder of Off the Shelf Analytics, Adrian Coy, who told The Next Web: 

“Being data-driven actually enables organizations to take more risks because they know they can quickly see when it’s not working and course-correct when necessary.”

Sainsbury’s, one of the biggest supermarket chains in the UK, had to rush to meet demand as shoppers panic bought essentials including flour and tinned pork. Group CIO Phil Jordan told diginomica that his retailer was the first to use data insights to identify elderly, disabled, or vulnerable customers and ensure they had access to buy their groceries.

In addition to serving the bottom line, data insights can support decision-making when it comes to other areas of the business such as employee engagement and corporate social responsibility. This can lead to a better understanding of the state of mind of the staff body and the effectiveness of different CSR initiatives.

The problem is, transforming data-driven insights into impactful business decisions isn’t a simple process. In fact, a recent study by NewVantage Partners found that just 24% of respondents thought their organization was data-driven (down from 37.8% the previous year). 

According to experts, if you want to build a truly data-driven business, you can’t miss these six steps:

Step one: Create a roadmap

The first step is to create a solid foundation by thoroughly understanding the data you’re working with and creating a roadmap for insight. This involves defining how you understand, measure, and segment customer needs and behaviors, as well as how you build models to predict future behaviors, and define the right KPIs, measures, and processes.

Mike Bugembe, author of Cracking the Data Code, stated in his book that the absence of a clear strategy can lead to time and money being spent on collecting and analyzing data that looks interesting and just might prove to be useful. He wrote that generic exploration is great if your company has the time and human resources to spend on it, but if not the company could be left data-rich but insight-poor. “This puts any expected ROI at risk because data is only valuable when it is used appropriately to deliver real results,” Bugembe wrote.

Step two: Foster a culture of data literacy AND data democratization

Next, it’s important to, “put the right metrics in front of the right people,” said Coy, through reporting and analysis tools. People within the organization need to be able to understand the right KPIs – which should have been set out at the foundation stage – and define the right way to track them. This speaks to the need to have a high level of data literacy within the organization, in addition to a culture of data democratization where non-technical users can access data.

Coy said that business intelligence tools are useful at this stage and people should be coached on how to use them and understand what decisions to take, not just which buttons to press to get the report. Some of the tools used for data analytics are made by Qlik, Tableau, Thoughtspot, PowerBI, Looker, Sisense, Spotfire, Yellowfin, Targit, DataRobot, and Snowflake, among others.

Ross Perez, head of international product marketing at Snowflake told The Next Web, “These tools are vital to accessing one’s data, asking questions of it, evolving a strategy, and ‘actioning’ that strategy.”

Step three: Develop data storytelling skills within your team

To underscore the essential nature of data education, Perez said that as important as it is to ask questions of data and allow it to tell a story, it’s equally important to teach people who work with data how to tell a story with data effectively. 

Good data storytellers have the ability to convince and persuade. Studies have shown that charts and graphs can make people change their minds more easily than words can. Furthermore, according to Andy Cotgreave, technical evangelist director at Tableau, one of the best ways to keep an audience engaged is to present data in a way that is easily understood by everybody.

Step four: Gain C-level buy-in

The fourth step is to create a top-down culture for the senior management so they can ask for the evidence behind recommendations and find out how the results will be tracked. C-level buy-in to the idea of being data-driven is crucial, as is the direction that senior-level executives can give.

A McKinsey report found that executives who recognize the importance of data are open to receiving educational workshops. This display of leading by example embeds the culture of data literacy among other members of staff. And leaders who are fully on board with a data-driven culture can ensure that data is as accessible as possible, one of the tenets of data democratization.

Without support and understanding from the executive level, insights teams can be left rudderless, without a clear course to follow or precise priorities of which problems to solve and which data to analyze. This can result in an insights initiative falling flat as they fail to gain traction or deliver on what was expected.

Without strong data leadership, wrote Bugembe, organizations can fall behind their competitors in terms of winning, serving, and retaining customers.

Step five: Build a ‘consultative’ analytics capability 

Following that, build a ‘consultative’ analytics capability able to invent approaches to answer the big important business questions. This is the point at which you look for patterns everywhere and see which data points have a positive or negative correlation. According to Coy, this is important because consultative analytics can help businesses to figure out the right questions to ask of the data and the right requests to make. 

Through the lens of the pandemic, an ordinary question might be around the increase or decrease in sales but a question formed through a consultative approach would be ‘what do I need to do differently?’.

Step six: Assign accountability 

Finally, make sure there is someone who takes accountability for the data. This may take the form of a chief analytics officer or chief data officer. 

While the results are manifold and data democratization means putting responsibility in the hands of many, Coy said that, ultimately, there needs to be clear accountability so that data is used in the right way. He said: “All this depends on the data being done right. Bad data is worse than no data. Make sure there is someone responsible for it.”

Although Perez wouldn’t explicitly say if there was a right or wrong way to glean data insights, he did say that some activities were more beneficial than others.

One of the less advantageous data activities, he said, is to choose a business strategy and then find data retrospectively to back it up. “Even worse, is using data in a way that obfuscates the truth so that some trends appear stronger than they are and other trends disappear entirely,” said Perez. “It’s really counterproductive.”

Conversely, he said that, if people are trained to understand and use data appropriately, they’ll be able to enjoy the full benefits the data-driven approach can bring.

To understand how data can transform your business, check out, “The Evolution of Data in the Cloud: The Lynchpin of Competitive Advantage”, research sponsored by Snowflake.