Micro Focus is now part of OpenText. Learn more >

You are here

You are here

How to rethink what data-driven means in your business culture

Mike Perrow Technology Evangelist, Vertica

Somewhere, you've probably come across the famous statement, often attributed to Peter Drucker: "Culture eats strategy for breakfast." That is, a strong business culture—including employees' personal and professional behavior, attitudes, expectations, and track record—is a clearer marker for business success than are the best-laid plans (i.e., strategy).

While strategy is needed, consistency, integrity, and leadership make the bigger difference when you consider results among various companies over time.

But is it possible that a well-established culture could be the enemy when a business strives to become more data-driven?

After all, the confidence of most business leaders lies in their past successes. What happens to that confidence when they're suddenly confronted with data suggesting they move the business in a way that runs counter to their instincts? Those executives need to learn to trust the data.

You can help by building a data-based decision making culture where executives learn from experience that they should trust the data. First make sure you fully understand the culture challenge that lies ahead. Then take the steps below to build this new culture in your organization. A key lesson here: Don't be afraid to go big.

The problem in getting to data-driven: Just about everything

NewVantage Partners' (NVP) 2021 survey of 85 Fortune 1,000 industry-leading firms in financial services, healthcare, life sciences, and retail found that the sample organizations are universally invested in data and AI.

However, the results aren't so good. Only 24% of those businesses report that they are indeed data-driven; that's down from 38% last year. What's the problem?

In his Harvard Business Review write-up of the survey results, NVP founder Randy Bean put it this way: Some 92.2% of mainstream companies report that "they continue to struggle with cultural challenges relating to organizational alignment, business processes, change management, communication, people skill sets, and resistance or lack of understanding to enable change."

So, even though data and AI are clearly part of the strategy, is Drucker's adage about the relative power of culture just as true as ever? Maybe. But there's a different way to look at the data-driven challenge.

'Strategy vs. culture' is too simplistic

With the availability of data analytics, leaders no longer have to rely on gut instincts based on experience when deciding about trends to exploit. The data itself, when collected and analyzed properly, can show those things to business intelligence and data science teams, which in turn can report the findings to the corner office. It's then up to leadership to incorporate what the data shows into their business decisions.

So the issue isn't "strategy vs. culture," but rather "experience culture vs. data culture." Either way, "culture" is always part of the equation. The challenge lies in getting experienced executives to trust that the data "knows" as much, and often more, than they do.

It's not enough to show them how truly data-driven businesses such as Amazon, Netflix, and Facebook are eating their competitor's lunches. They need to see data-driven success in their own companies. So how should those who want to focus more on data drive toward positive results?

Go big with your data-driven initiative

Any time an organization decides to move from an old to a new way of doing things, many folks advocate taking baby steps, moving gradually to prove that the new approach has merits.

Too frequently, that careful approach leads to failure.

Why? Because while one team gets assigned to a "safe" project—i.e., one that the business can afford to have fail if the new methods don't pan out—all the other teams continue plodding along as usual, with little interest in the success of the experimental project. The larger organization learns nothing and, worst case, a failed experiment just confirms bias against any kind of change.

Experts often suggest another approach: Choose a project that will have a big impact on the business, where stakes are high enough that success will really matter. This is the often-recommended path when changing software development methods from waterfall (heavy up-front requirements) to agile (incremental and results-oriented).

This approach includes putting your high-profile contributors on the job—ones who can lead others in the new direction—while reporting progress and making regular adjustments along the way.

Bean makes a similar recommendation in his analysis of why businesses struggle to become data-driven: They need to start with high-impact use cases. "By starting where there is a critical business need, executives can demonstrate value quickly through 'quick wins' that help a company realize value, build credibility for their investments in data, and use this credibility to identify additional high-impact use cases," Bean advises.

You can glean similar advice from other recent business journal articles; see the section about data scientists in this Harvard Business Review article, and the part about engaging key data stakeholders in this CIO Magazine article

Renew your focus on the data

Taking on a big, high-profile project to demonstrate valuable, data-driven results is easier said than done. In the not-too-distant past, transactional data resulting from sales and other business processes was considered valuable only as a record that those processes occurred, and after a period of time that data was typically purged.

When big data techniques came along and important trends could be understood from that same data, given the right kinds of analysis, those records suddenly became a gold mine for many businesses.

Big picture: The chief data officer (CDO) or any leader guiding an organization toward a data-driven culture should encourage the various teams across the organization to think differently about the data itself. Here are two principles that can help drive change toward a data culture.

Treat data as an asset, not simply a byproduct

With today's focus on data analytics, data science, and all things big data, this may seem like an obvious bit of advice. But the key is understanding how data becomes valuable to your business, which means figuring out what's possible to learn from any data source.

And that means leaders in a data-driven transition need to align business problems with the data that's most likely to hold the answers. 

The next steps may involve some coordination with IT to ensure the data is accessible in its entirety, not just as a sample. Data truly becomes an asset when analytic processes can discover patterns based on the fullest datasets available, even when that means working at petabyte scale, or larger.

Refine your data sources as much as possible to ensure data is clean and usable 

This means acknowledging that not all data is usable in its initially acquired state. Becoming data-driven means not just investing in data scientists and business analysts, but also in operational data specialists who can ensure that an organization is collecting useable, accurate, and clean data in the right format.

Refining data sources can be as simple as ensuring accuracy in data being collected—for instance, your phone calls reminding customers about an appointment should always begin with confirming the customer's mailing address. Or the data refinement can be much more complex—e.g., if IoT data collection requires some sort of filtering to provide more structure to the data prior to analysis, data engineers will likely need to be involved.

Individual contributors on your larger team can get quickly frustrated when poor data quality in day-to-day processes leads to relatively simple mistakes, such as getting an address wrong. When this same data is taken in aggregate for a big data project, it's easy to see why the processes that produce that data in the first place need to be retooled.

By the way, this need for "clean" data might also guide which initial projects should be selected, since cleaner, more trustworthy data will have the greatest likelihood of having a big business impact in a data analytics initiative.

Be patient; you may be taking on a culture that's old and deep

If you work for or lead a large company, the culture is deep and often decades old. Listen to how people talk. Employees typically adopt the same phrases that are unique to their business, and language is one of the most telling aspects of any culture, business or otherwise. Maybe everyone is willing to help the organization become data-driven.

But does everyone know what "data-driven" means? Some experts recommend creating a data glossary, given that everyone on the team needs to be speaking the same language.

The barriers to adopting a data-driven culture will naturally be different for different organizations. It's important to remember that the organization's data is just as unique and holds value that smart leadership can exploit. That’s easy for born-on-the-cloud startups to understand.

But businesses that have been successful for decades may find it hard, at first, to accept that the data team needs a seat at the decision-making table. That's where a successful, high-impact initial project can make a difference.

If you read the literature on achieving a data-driven culture, you'll find plenty of advocates for all the soft skills that need to be preserved in the process (here, for example). There shouldn’t be any arguments with that. Becoming data-driven doesn’t mean setting aside inspiration, thoughtfulness, team building skills, and other ingredients that successful organizations have always embraced.

The point is, for a business to become data-driven, it needs to blend a heavy dose of data into its culture—especially at the top, where the business decisions get made. Data will not only make the strategy better. It will also help the business better understand its processes as leadership hones them in the quest for new levels of success.

Keep learning

Read more articles about: Enterprise ITData Management