Ensure business value by adopting a data product process
As we wrote in our earlier Insights post Why Organizations Struggle to Derive Value from Data, a McKinsey Global Survey showed that a mere 15% of organizations believe they are fully capitalizing on their investments in data and analytics.
Over the years, we have gained insights into how a multitude of companies, ranging from quite immature to world leading, are working with data, analytics and AI. A major finding we can see when we compare the companies that are creating high business value from their data and analytics efforts with those who fail to do so, is if the companies have a structured and measurable process when it comes to identifying and solving business problems.
The main problems we see for the companies that do not have a structured approach and process for solving business problems and delivering data products that create value are:
- Focusing on a solution and/or a technology
- No in depth understanding of the problem to solve
- Do not understand what outcome to achieve and which metrics the effort should affect
In order to help companies to achieve business value and make their data and analytics effort pay off, we have developed a data product process which will ensure that data and analytics efforts are value-focused and maximises chances of success (alternatively, stopped early which also is completely acceptable).
Background
Our Data product process is based on a model called triple diamond, which is a product development model. There are good material from Zendesk around how they use it and more in depth information about the steps etc
Link: The Zendesk Triple Diamond
Based on years of experience within the Data and Analytics space we have altered the model to fit more into the data and analytics process and created material that could be used in driving the development of data products.
Data Edge’s Data product process
We outline the process in a few major steps which we see as part of a data product lifecycle. Since there are many different definitions of data products on the market, we need to be clear what we mean in this context. In this context we use it more as a concept that could be ML-model, a report or an insights work for a specific decision, which means that the process timeline and depth of each step will be highly dependent on the problem and solution. The main focus for us with this process is to solve a business problem in the best way possible with the support of data and analytics.
This is not a straight process that moves solely from left to right. As learnings happen you might need to revisit earlier phases, such as the problem discovery phase. However we think there are important areas and questions that should be answered before confidently moving into the next phase.
Examples of important topics to cover during the use case identification phase:
- What is the high level business problem/opportunity
- What business outcome/metrics are you targeting
- What stakeholders do you have and are they engaged
- Business value potential and certainty
- Feasability (Technical, data, process changeability, compliance)
- Strategic alignment
In this phase you do not need to have the perfect answer to every question, but it is really important that you have thought about these aspects and can create a common view on this opportunity.
Each phase has different important aspects to consider and work deeper with.
Concrete example
Benefits
Maximized chance of solving a value adding problem or stopping the work early without too much time waste
- The process and support material makes sure that the most important parts have been worked through before coming too far in the development process
- It ensures that we early on understand what metrics we are trying to improve and what value potential we see. (This might be one of the most important parts in creating a data-driven organization)
Easier communication and prioritization across your data and analytics function(or company)
- In our view, many times organization have a hard time understanding what use cases are being worked. With a clear process you can visualize all use cases you work on as well as where they are in the process.
- This can help both when communicating with stakeholders and set the expectation level as well as increase your understanding of what effort will be needed in the different cases and serve as ground for prioritization.
Support and shared best practices across your organization which leads to more efficient and effective work
- Creating a common way of working and process when delivering data products makes it easier to create supporting material and knowledge sharing across the team.
Conclusion: Make sure that your data and analytics initiatives are creating business value in an effective and efficient way
If you feel that you are not fully creating the business value you think you should, ask yourself if your team is working in a structured way that both maximizes the possibility for value creation but also creates long term impact on capabilities which will make you become faster and faster in delivering value driving use cases!
Data Edge has a proven track record of helping organizations to create value by solving real business problems with data, anlytics and AI. If this article has sparked your interest and you think we can achieve something together, let us know!