Scale up your data usage | Think Evolve Solve

Scale up your data usage (without scaling your team)

October 11, 2022

Despite investments in time, effort, and expertise, many organisations are still struggling to realise the value of their data.

Practical challenges like poor data quality, technical-debt and skills shortages grind data projects to a halt. Cultural challenges like a fear of data, a lack of trust and a lack of clarity on responsibilities reduce data adoption.

All of these factors make data uniquely difficult to manage. Most organisations haven’t yet found a successful approach, resulting in data remaining underutilised at best, and at worst, unmanaged. In the ‘information age’, this is often the biggest missed opportunity business are facing.

Why is data difficult to manage?

Today almost everyone’s job involves creating, using or interpreting data. Despite this, the majority of business teams believe they are unqualified to work with data, and are reluctant to participate in its management. It’s often unclear who is ultimately responsible for data, as it is such a vital part of every business unit.

Technology adoption means more data than ever is siloed or hidden, making it near-impossible for business teams to manage independently. As a result, data work is often handed to technical or IT teams, who don’t have the business context to manage it.

Many organisations seek a CDO (Chief Data Officer) with a blend of technical and business skillsets to centralise management. However, these highly qualified individuals still depend on a huge distribution of stakeholders to complete their objectives, and can flounder without a clear management approach.

Get the whole business involved

The below tactics from the ‘data supply chain approach’ enable CDOs to define responsibilities and critically measure performance so organisations can overcome data management challenges and accelerate their data programs.

Step 1: Define Accountability

Get the business team involved. Data begins and ends with the business team – people who don’t have ‘data’ in their title. They’re at the front line of creating and interpreting data in their daily responsibilities, yet they’re almost always excluded from the planning of data management and design of data solutions.

Get these teams involved in data management by defining their impact as ‘suppliers of data’. This includes explaining how the data they create is utilised in business planning and the possible outcomes if this data is incorrect.

Sharing this information and introducing the concept of ‘data supplier’ enables the CDO to define the desired quality standards and format for data, as well as involve business teams that supply data in the remediation process when quality requirements are not met.

Focus IT on tech, not data. IT, data engineering, and technical team members should never be responsible for data. These teams don’t understand the business context for how data is created or used. Their skills are costly, and best focused on building infrastructure to move and integrate data, with a clear brief from CDO.

Step 2: Bridge Understanding

Successful CDOs are a bridge between the business and IT. They are the middleman that creates a shared understanding and facilitates collaboration between business teams as data suppliers, IT skills as infrastructure providers, and business teams as data consumers.

To do this, CDOs need to clarify responsibilities, build common language use, and create predictable, controlled processes that keep team interactions minimal and efficient. The data supply chain approach defines a clear step-by-step process that draws inspiration from traditional supply chain management. This has been utilised with great success across multiple industries.

Step 3: Measure Performance

Get full visibility. Help business and IT teams become fully accountable for their part in data processing by defining and establishing clear measurements. Create metrics to analyse things like data quality, supplier performance, and time elapsed between data receipt and incorporation into a data product. These enable CDOs to understand the parts of the process that require attention and to take corrective action when a team is unable to meet their quality standards.

Step 4: Integrate siloes

Manage data before it is integrated. Overcome an uncoordinated ‘spider web’ of data integration by streamlining data processing independent of source.

To do this, define each dataset in the business as an ‘asset’, a raw material that will be utilised to create multiple data products downstream. By considering the quality requirements for each asset, companies can streamline processing for easy use downstream.

When data assets are defined, it’s easy to create a centralised ‘funnel’ that coordinates data quality and preparation in an organised way before it is stored. The resulting ‘information layers’ of quality-assured information can be integrated in systems or utilised in data solutions without duplication of effort downstream.

Step 5: Encourage collaboration

Use a workflow tool. Support collaboration between business teams and data teams using a workflow tool that manages the interface between these two teams. These tools can help by providing clear accountability, and providing clarity on the task at hand. Notifications and measurements can also prevent tasks from falling between the gaps and cut out challenging email back and forth to remediate issues.