- What does self-service mean in your organisation?
- What are the business goals behind enabling scalable self-service?
- How do you ensure data access leads to value, not noise?
- What technical and cultural enablers matter most?
- How do you align central teams with local ownership?
- What does good knowledge management look like in practice?
- How do you maintain trust, quality, and shared language at scale?
Do not confuse dashboards with decisions
Most users want simple visualisations, but that does not mean they are getting value. As one participant put it, “We give people data to explain the obvious.” Self-service must be measured by the decisions it improves, not the volume of charts produced.
Define the “who” behind self-service
Several leaders noted that different users need different models to achieve their goals. Not everyone wants to explore raw data; some just need the right insight at the right moment in the workflow. Successful implementations design around roles, not just tools.
Synchronised autonomy beats top-down control
Rather than enforcing standards, the most mature organisations foster alignment through joint ownership. Data leaders need to stop thinking of themselves as the data police and lean into a synchronised relationship with the business units. The cultural shift is slow but essential.
Build for discoverability, not just access
Multiple leaders raised the issue of duplicated effort and siloed insights. “People build the same reports without knowing they exist elsewhere.” Without metadata, catalogue usage and visibility, self-service can deepen fragmentation and hinder progress.
Self-service starts with safe, governed data sets
Empowering business users does not mean abandoning quality. Data leaders explained they often utilise a hub-and-spoke model to push out well-defined data models and empower users to solve local problems rapidly.
Training is foundational
One leader flagged a clear gap: “We’ve got tools and data sets, but lack formal training. People are guessing their way through.” Without intentional upskilling, especially around interpreting and contextualising data, self-service fails to deliver.
Culture is the differentiator, not platforms
A number of well-known platforms were widely in use across the different organisations, but no data leader saw them as silver bullets. “The platform comes empty. If you don’t bring the people and the energy, nothing happens.” The workaround to this is convincing the business that stewardship is a value-add, not a chore.
The carrot matters more than the stick
Ownership and stewardship need legitimacy, not mandates. Data leaders cannot successfully assign someone a data role and expect results – they have to see the value. Leaders have found that co-prioritising the backlog with stewards worked better than telling them what to do.
Use perceived value to guide investments
To make the case for platform spend and literacy work, one leader surveyed users with a simple prompt: “What would you pay per month for this report?” Responses were used to simulate ROI in business terms, which is more credible than theoretical benefit statements.
Resist the illusion that tech equals engagement
As one CDO warned, “You need to be careful. There are many data teams who think that just through the deployment of technology and process, users will get engaged.” Real engagement comes from giving users something they cannot get elsewhere: richer insights, faster decisions, and simpler processes.
Practical Next Steps
- Establish a shared definition of self-service. Avoid vague language and define it by who uses it, what for, and what is governed.
- Segment your users. Match enablement to role and readiness as not everyone will require the same level of access or tooling.
- Invest in catalogue usability and metadata. Findability is as critical as access, so focus on reducing duplication and increasing transparency.
- Pair governed datasets with practical training. Deliver ready-to-use assets and show users how and when to use them to achieve success.
- Avoid platform burnout. Prioritise human behaviours and incentives over features and rollouts.
- Involve data owners in backlog shaping. Co-ownership creates accountability far more than job titles do.
- Survey perceived value regularly. Use signals and insights from real users to guide investment and priorities.
- Design for culture early. Embed legitimacy and control from the start, before rollout begins.
