51 Questions to Help You Create Business Value with Data
“To ask the right question is already half the solution of a problem.” — Carl Jung
Everyone is talking about how data solutions need to create business value.
But how does this actually happen?
One popular paradigm is the “Data — Information — Knowledge — Wisdom” pyramid. In this model, raw data with context, enriched with meaning, and ultimately analysed, can give us insights into the past or future. This model describes how we can use data for analytical purposes within a business to create value.
But data can be more than just a resource for various models or a requirement for financial reporting. Google, Facebook, Airbnb, and Uber are all data-driven companies. For them, data is not just an addition to their business but the enabler of their core operations. They build applications based on data. This also influences how they build their systems.
Now, not every organisation is a technology company, and treating data in the same way is probably not feasible. Each data initiative requires a custom approach tailored to the given setting.
Over the last 7+ years, I have been a consultant, data scientist, data engineer, and solution architect in more than 10 projects for different organisations. During this time, I’ve gathered and analysed requirements, implemented models, designed architectures, built data platforms, developed data pipelines. I’ve also presented to top management, aligned with stakeholders, mentored juniors, established standards and best practices, automated processes, estimated budgets, etc.
To do my job, I need to bring together information from different areas: organisational framework, technical landscape, business requirements, financial constraints. Over the years, I have put together many different lists of questions to be used for documentation, workshops, meetings and presentations. The answers turned concepts, plans and eventually products.
In two of my last articles, I covered some questions related to solution architecture and tool choice, but I felt there were many more important ones to ask when discussing data solutions in general. This led me to this article, in which I summarise 51 questions I believe can help us deliver business value with data.
Not every question is worth asking or debating in every setting, but having a framework around this allows me to choose the most important ones, build upon them, or even answer them myself so I know where I stand.
It’s up to you to decide if it’s worth reading through every question or keeping it as a backlog for later use.
Enjoy!
Business and Strategic Goals
What are the short- and long-term goals of the business, department, or executives?
How does data support these goals? Is it used for decision-making, reporting, or as a core component of an application?
Do these goals span multiple departments? How much alignment between stakeholders is required?
What does success look like for this initiative? How will we measure the effectiveness in achieving the goals?
Who is supporting this data initiative? Who are the sponsors, and how can they support us?
Is this a project with a clear scope and end date, or is it an ongoing initiative? What are major milestones along the way?
Do we have use cases that can provide the required return on investment (ROI)?
Are there strict budget constraints and hard deadlines that we must adhere to?
What are the regulatory or compliance considerations related to this initiative?
What are the biggest risks associated with this initiative, and how can they be mitigated?
Project Scope and Requirements
Are we migrating from an existing data solution, or is this a greenfield project?
Are we developing Proof of Concepts (PoCs), Minimum Viable Products (MVPs), or productive solutions right from the start?
Is this the first and only data use case, or are we building a platform for multiple use cases?
Are we building a data warehouse, data lake, or lakehouse? What is the reason behind choosing one over the others?
What kinds of use cases are we going to implement? Are they focused on batch processing, streaming, Online Analytical Processing (OLAP), or Online Transactional Processing (OLTP)?
How much historical data do we need to store? How fast do we need to be able to retrieve it?
How quickly does data need to be available? What are the benefits of having data available immediately?
How critical are these systems? Are they directly integrated into production processes, such as manufacturing?
Can we talk to the source application team? Can we modify the source application if needed?
How critical is data accuracy and consistency? Is it acceptable to lose singular data points?
Data Sources and Technology
What source systems are we connecting to? Are they streaming data sources, RDBMS, cloud-based, or on-premise?
Are we working with structured, semi-structured (e.g., JSON, XML), or unstructured data (e.g., text, images, audio)?
Where is the data generated? Is it through user input or machine-generated data?
Do we have specific high-performance requirements? Would we benefit from distributed computing?
What are the main data security and compliance requirements we need to consider?
How do we plan to handle data quality and data governance?
Are there specific tools or technologies that the organisation prefers or has standardised on?
Are there any existing integrations that need to be maintained or modified?
How scalable does the solution need to be to accommodate future growth?
What transformation processes are required before the data is usable?
Are there plans for integrating third-party data sources?
Team and Expertise
How many people will be working on each use case, and what are their roles and levels of expertise?
What is the role of the data team within the organisation?
What is the data maturity level of the company?
Do we have the expertise to customise, implement, and maintain the solutions?
How well can the team define and provide requirements? Will we need to extract and define every detail, or is the team capable of providing well-defined specifications?
Do we need to hire additional resources or consult external experts?
How will the data team collaborate with other departments, such as IT, marketing, and sales?
How will roles and responsibilities be defined and communicated within the team?
How will the performance and productivity of the team be measured?
User and Stakeholder Considerations
Who are the primary users of the data solutions being developed?
What are their key requirements and expectations?
What are the primary concerns or pain points of users related to the current data processes?
How will we ensure that the data solutions meet the needs of all stakeholders?
What feedback mechanisms will be in place to continuously improve the data solution?
How much training or onboarding will end-users require?
Do we need to provide different access levels or data views for different user groups?
How will we gather and incorporate user feedback into future iterations of the data solution?
What are the user’s expectations for system uptime and availability?
How can we measure and encourage user adoption?
How will changes and updates to the data solution be communicated to users?
Conclusion
Navigating the complexity of IT and data projects requires extensive experience, as well as both a high-level understanding of the business and the ability to dive deep into the technical details. It involves asking the right questions and drawing the appropriate conclusions for each situation.
These 51 questions are just a starting point for anyone considering a data project. There are certainly many more to ask, and I look forward to hearing from you about what you believe might be missing.
All the best,
Eduard