Navigating Dashboard Success: Insights from 150+ Deployments

After having overseen the delivery and lifecycle of 150+ dashboards, using Tableau, Power BI and SAC over the last 7 years, wanted to share my perspective on the key success factors for these data products.

?1.?????? Define clear business ownership of each of the dashboards.

• Each dashboard needs to have a named business stakeholder as owner who confirms every change and sets expectations with the intended audience.
• Having a named owner ensures accountability towards good quality requirements, address resolution of data/process issues.
• Named owner can be nudged to ensure that the right stakeholder and the right levels are engaged in the delivery process.
• Hopefully this also ensures that the investment in dashboard gets scrutinized for the value of the desired outcome.
• Of course, the delivery team needs to partner with the designated owner and help guide with the right questions to increase the chances of addressing the right business problem and therefore adaption and success.

?

2.?????? Robustness of the Data Supply chain

• Data supply chain needs to be automated.
• Reliable, that is it works every time.
• On time and in full. Delivery every time and ensures consistent data with the source.
• Any failure will erode the trust of the consumers on the insights being presented. You do not want your stakeholders second guessing if the data got refreshed properly. That’s where I am a strong proponent of simplification of the data supply chain. Additionally, if feasible, you should also explore investments in suitable data observability tools, so that your team is able to detect and resolves issues before the consumers detect it.
• The consumer need not understand and does not need to, the complexity of the data supply chain. That is what our expectation as consumers is as well in our day to day lives - get good quality products and service, every time and on time and reasonable costs.

?

3.?????? Operationalization by embedding in business processes.

• Operationalization ensures that the dashboard is key to the completion of a process and no longer optional.
• With a review cadence established along with the procedures defined to perform what, why and how, such dashboards become drivers of action and change.
• Ideally integrate recording, monitoring and updating of the accruing Actions Items along with the dashboard itself. This ensures accountability for the respective owners.

?

4.?????? Purpose driven - Tools such as Power BI, Tableau etc are highly effective and matured, to the level that they are able to support and translate the chain of thoughts into good visualizations. However, their purpose if not to produce pretty visualizations, but to support the business in making better decisions.

• Need to drive / support an action/decision.
• Right visualization with just enough complexity to convey the message should be strived.
• If it takes effort by the consumer to extract the insights, the dashboard won’t succeed, no matter how much visually appealing its is.
• No need to get swayed by complex visualizations that are difficult to decipher.
• Dashboards is not the answer for every report or information request. Consider alternative options as well.

?

5.?????? And of course, appropriate governance needs to be in-place, adherence to the design and development and delivery guidelines to be adhered to. You can refer to the best practice documents made available by the vendors. My broad learnings:

• As far as possible push down the business logic to the data platform layer. Let the visualization tools focus on visualization and those computations that are appropriate for this layer.
• Ensure that the dashboards are responsive, in seconds. Consider the geographic locations of the consumers and test their experience for that location. Refine accordingly.
• I now prefer to focus efforts on building governed data marts for single source of truth and only governed dashboards where the organization needs to ensure consistency across all the consumers. For the other dashboards, I prefer the business analysts or power users to create their own dashboards using their tools of choice by leveraging the governed data marts.
• Monitor the dashboards and present the usage to the relevant stakeholders. Identify dashboards that can be taken off to free up the compute and maintenance efforts. ?

?

6.?????? We need to be open to consider alternatives for other use cases, as required.

• I have been investing weeks exploring how LLM’s in conjunction with frameworks such as LangChain, etc shapes this space and proactively brings the alerts and insights to the consumers as per their preferred mode of interaction. This is such an exciting space and therefore should be on your plate…. now.
• From the decision makers perspective, nothing beats than getting only the relevant information and insights at the right time. Getting the answers to your questions based on your chain of thoughts, in your comfortable language, at your time, within seconds, on the preferred medium of interaction (Teams, slack, mails, What’s App etc) is clearly the next happening thing.

I am interested to know about your learnings and experiences. Looking forward to your views on the above.?