We’ve Made Dashboards Beautiful — How Do We Make Them Trustworthy?
With the proliferation of self-service business intelligence (BI) tools and the subsequent democratization of data across the enterprise, trusted business reporting is more critical than ever. It needs to be baked into your data strategy.
Here’s a scenario that you have probably seen at your own organization: The CEO receives five different reports from five different teams, and the metrics don't align. Lack of trust breeds low-quality, unreliable insights. Delays in decision-making easily result.
Your employees who rely on data to do their jobs — people we call "data citizens" — need to know not just how to locate data but also that they can trust it. They need to understand that the data is authorized, so "data brawls" or debates don't ensue because the source of the data isn't trusted. They need to know if the data on hand is appropriate to share or if it needs to be protected. They need to know there is a single source of truth for business terminology, so inconsistent business terms and KPI and metric calculations don't lead to lack of trust in the analytics used in executive reporting and dashboards.
At my company, we grappled with similar challenges as our company was scaling up and we were installing our leadership teams. And, like many enterprises, instead of simply being informed, we were overinformed — we had too many metrics, too many insights. We wanted to use our dashboard to track specific metrics — our revenue growth and velocity of development, for instance.
This is why data governance is an imperative for enterprises. Data governance ensures that the reporting and insights generated can be trusted, all the more acute when data access is spread far and wide throughout the enterprise.
An enterprise data catalog with embedded data governance capabilities will streamline your business users’ ability to locate trusted data, ensure consistent terms and metrics, identify the data authority, enable certification of key reports and KPIs/metrics, and ultimately enable confidence in your data’s veracity and relevance. It will empower your business users with a process that is automated and straightforward to help speed innovation within your enterprise.
We set up a trusted business reporting process as a means to better manage growth in our organization. As part of that process, we created metric certification standards to confirm consistency and quality assurance to all data users.
We began by determining which metrics are essential to our enterprise, and then we established what it would take for those metrics to be trusted.
To secure an enterprise-wide understanding of our metric certification standards, we created a hierarchy of watermarks. Reports are "stamped" with a bronze, silver or gold watermark based on their coverage of trusted business reporting requirements. A bronze watermark, for example, demonstrates how metrics have been defined and calculated. A silver watermark delineates that the metrics are also connected to specific self-service BI reports (Tableau, in our case). A gold watermark assures the data lineage.
This provides another layer of quality control, giving users more confidence when making decisions because of their trust in the reports and in the process that produced those reports. This process is easy to set up and popular with stakeholders in the business. It creates top-down demand (as executives want to make decisions based on reports and data they trust) as well as bottom-up buy-in (as insight producers want to get to the "gold standard" that ensures their work will be seen and acted upon). The process is a stepping stone to data intelligence, where you apply the full value of your data to your business objectives.
Here are five steps to propel your enterprise's trusted business reporting forward, enabling the necessary confidence for your data users to execute insights and make decisions:
1. Establish standards for data quality for a report to be trusted and certified. Determine the necessary attributes essential to your enterprise in terms of accuracy, lineage and identifying the data authority.
2. Create a central catalog of reports to increase awareness of what is available and the associated KPIs and metrics. Establish intuitive workflows to manage how metrics, KPIs and reports are verified, governed and certified, expediting access to available reports and preventing redundancies and reports going forward.
3. Create a unified view of all data assets, a single source of truth. This includes a business glossary, a shared language, with approved definitions of terms, KPIs and metrics (with the supporting business logic) to facilitate the understanding of reports and accelerate trusted decision making.
4. Establish accountability for data. Identify the critical data elements (CDEs) that are essential to key reports and ensure proper governance of these CDEs. Establish authoritative data sources and trace lineage from these data sources to the reports to ensure clarity and trust. Have the authority formally certify reports and KPIs/metrics as "fit for purpose" and show that the assets can be trusted from the bottom up, accelerating trust for decision makers.
5. Collaborate across your business. Collaborate across business functions to define data, build context and maintain information as your business changes.
Business is moving faster than ever. The digital transformation isn’t stopping — in fact, we are at the very beginning of this journey. The use of artificial intelligence and machine learning throughout the business ecosystem is poised to explode. But you can't move forward without trusted information at your fingertips. To drive value from your data, the data needs to be trusted. It needs to be essential. It needs to be actionable. Trusted business reporting is paramount to securing the quality of your decisions and the actions you take on behalf of your enterprise. There is no time to waste; we are moving rapidly to the now-or-never point.
Forbes Feature - Stijn "Stan" Christiaens is Founder and Chief Data Citizen at Collibra.