Proof of Value:The Crucial First Step of your Data Strategy Journey

A Guide to Understanding and Implementing Proof of Value for Your Organisation

In our last blog, we outlined some of the first steps businesses should consider before aligning their data and business strategies. One such step is to run a data strategy workshop. The aim is to bring key department heads into one room to present & brainstorm use cases that will revolutionise their decision-making process using data in their respective departments. The outcome as a group is to agree on which use cases to run a proof of value against.

To help you understand the Proof of Value process, we’ve written a checklist, leading you through the core concepts of this important process, as well as the benefits one provides.

Benefits of Running a Proof of Value

The key goal of the Proof of Value is to prove that the use cases selected are technically achievable and that they fulfil the use case business benefits. Although in some simple scenarios, proof of values may become a production-ready application compared to more complex use cases, which require more detailed technical planning and execution.

Wherever possible, the proof of value should engage many department members in the design and execution of each use case, especially the end users who will benefit from the final application. The two key benefits of this approach are 1. User adoption as the end users feel part of the process, and 2. You start identifying data literacy gaps and highlighting where end-user training may be required. Providing an analytics dashboard to an end user for specific use cases is all well and good, but if they don’t understand how to ask questions or hypothesise ideas to gain insight, the application is just an unutilised, static dashboard.

Here are the key considerations for running the Proof of Value:

• Who is the sponsor of each use case, and who sets the priorities?
• What are the success criteria of the proof of value?
• Agree on what KPIs will help resolve a business case.
• Agree on what the key measures are. (e.g. Margins sometimes can be calculated differently by different users).
• Where is the data, and how can it be extracted?
• Does the data have quality issues? This will help define what the prerequisites are for the production application and what temporary measures need to be put in place for the proof of value.
• Define how the data should be modelled. In most use cases, you may need to extract data from multiple databases and files; how are each one of these data files joined? While this sounds trivial, you must consider this seriously.

To help you get started, we’ve created a comprehensive data migration checklist outlining all the steps you need to take to start your data journey on the right foot; revealing how Data Technology can support you.