The Number One Reason Your Data Strategy Will Fail (And How You Can Build a Framework to Prevent it From Happening)

Eugene Khazin

Principal and Co-Founder, Prime TSR

I’ll be the first to admit, creating a data-driven culture within an organization is easier said than done. When done properly, yes, companies can use data to make better decisions with proper insights. But, the question is, why do companies not live up to the expectations they set for themselves when it comes to data strategy?

I have a theory on why this happens: the IT and Business teams struggle with executing a clear data strategy framework together. Business leadership has their ideas on what they want to achieve, and IT has their own ideas of how to achieve it. This causes short term disagreements and long-term lingering data issues that sometimes never get fixed.

Here’s a prime example from a previous project:

When IT and Business Have Different Ideas About How to Solve a Problem

A customer support system had significantly degraded performance, and reps struggled to do basic tasks. The business was impacted because this restricted their ability to serve their customers.

The business team filed a high-priority incident report with the IT team. The IT team was under a lot of pressure to fix it, and honestly, they were also getting a lot of the blame about what happened in the first place.

The issue, like most enterprise IT issues, wasn’t a simple one to fix. In this case, the system, based on business requirements, had to store 13 months of data that all needed to be accessed from the core systems. The sheer amount of data slowed down the systems and eventually started impacting production IT systems.

Here was the problem: as IT was not the OWNER of the data they couldn't ACT to just get rid of the data and fix the problem. They knew what they had to do to fix the issue right then and there, but couldn’t pull the trigger. The business team didn’t want to lose that data, so they also struggled to give direction to the IT team.

The result?

Business and IT quickly agreed that only 5 months of real-time data needed to be stored, and thus, solved the problem almost instantly. Once the business and IT teams agreed on how the data needed to be accessed, based on real-life business requirements, things started to make a lot more sense.

This client had a strong data governance policy and talented data stewards who are dedicated to fixing data-related issues in the organization. The data stewards made a sincere effort in organizing and cleaning the problematic data. They had deep insights into how the data was used, so they were able to get both parties to agree on the next steps.

It was a quick decision, and it happened so quickly for one reason:

Data Stewardship.

In fact, I’ll take it one step further. Without a Data Steward, your project has a higher probability of failing. Not having a formally-assigned data steward is by far one of the major reasons why many data strategies aren’t executed properly.

Data Steward Is an Emerging Role Every Company Should Consider Hiring

What is a data steward? A data steward is the data evangelist within a company who sits between IT and Business. Data Governance focuses on overarching policies and procedures, and data stewards monitor and enforce the data rules. Think of them as Data Cops.

The need for a Data Steward stems from IT having little functional knowledge of the data, and the business team having little knowledge of the technical data details. Data Stewards fill this gap.

You will see these roles prevalent in financial institutions, but not many other companies have it, and I think that will change soon. Healthcare and insurance companies can especially benefit from strong data stewards.

The cost of having a dedicated data steward is easily justified by the positive impact on efficiencies, opportunities, and speed that are achieved from this role.

A major benefit of a data steward is guaranteed faster, easier access to data for your entire company. This means you can build reports with access to the right data at the right time without spending weeks or months trying to get access to the data in the first place. And in the case of my first example, you’ll prevent data issues from happening before they occur.

Data stewards manage and oversee an organization’s data assets to help provide business users with easily accessible high-quality data across an organization. Data stewards are integral to the success of any BI reporting system.

When data stewards and IT get along well and have a defined data governance model to work from, many data quality and system performance issues get fixed or never happen in the first place.

Most importantly, data stewards have to work with IT to achieve their goals.

Here’s a quick overview of Business and IT responsibilities with respect to data stewardship:

Business and IT both share a common goal: Establish a system of quality data which can help build great analytics reports to improve the organization.

It is important to have data stewards fall within the business units rather than solely within IT. However, data stewards and IT should maintain a collaborative relationship based on data needs. Data stewards need data skills, so begin your search by looking at your current “go-to” experts. Preferably, this person is already in-house and understands the current data situation within the organization.

How IT and Business Can Implement Data Stewardship

Establish realistic data governance goals.

From my perspective and experience, the quality of data in development environments should be no less than 85 percent and 99 percent in production. The higher the data quality in all systems, the better performing it will be.

You should be able to answer the question, “Who owns this data?”

“Who owns this data?” is a question I ask during the beginning of many BI implementations. And most of the time, without fail, I get blank stares or insufficient answers. Every project should have a data steward as an owner.

IT’s job is to build the systems and reports in a way that Business wants them, and it needs clear focus, not just from the functional teams but from the data steward, on how they want to handle the data and metadata. Without a clear vision, many of these projects will fail before they even begin.

Assign metrics to data stewards and make IT accountable for supporting their needs.

Being an owner of the data is not sufficient to fully understand if the data steward is doing their job properly. Being an owner also means they are accountable for increasing the quality of the data over the months/years. A sample metric an owner can be accountable for is “Data Quality in our Customer Relationships Management production system should increase from 95% to 99% in 12 months.“

IT should have clear direction about their role in the data quality. Whatever it takes to help the business, they should do it.

The next step is to hire a data steward—here’s how you do it.

My recommendation is to find a single person or group of individuals who have already been exposed to the data within your organization, which means you should most likely hire internally for this role. They don’t necessarily need the official title of “Data Steward,” but they do need to be part of the data governance framework.

A successful data steward ideally provides recommendations on hiring the specialist who brings the right toolset and process improvements to get the data to the desired state.

To summarize, data quality is a big issue for many companies. Establish a clear data governance program, bring on a strong data steward, and work with IT on a daily basis to support your high-quality data needs.

How Prime TSR can help.

We've been lucky enough to see this data story unravel many times. We’ve helped top brands create robust data strategies, and we’ll be happy to help create your successful data strategy.