Analytics maturity describes the degree to which an organisation has developed the data infrastructure, analytical capability, governance frameworks, and cultural practices required to use data as a genuine input to decisions at scale. Our research applies a five-level maturity model, Descriptive, Diagnostic, Predictive, Prescriptive, and Autonomous, to a sample of mid-market and enterprise organisations across nine industry sectors, producing a detailed picture of where the broader organisational population currently stands and where the steepest improvement opportunities lie.
Level one, Descriptive analytics, represents the baseline capability of knowing what happened. Organisations at this level have basic reporting infrastructure: dashboards showing historical performance metrics, standardised reports distributed on a scheduled basis, and the ability to answer backward-looking questions about business performance. The majority of our sample, 61% of mid-market organisations and 44% of enterprise organisations, is operating primarily at this level, despite significant investment in more sophisticated analytics tooling.
The gap between tool adoption and maturity level is one of the most striking findings of our research. A large proportion of organisations that have deployed advanced analytics platforms, business intelligence tools, data warehouses, and in some cases machine learning infrastructure, are using those platforms for descriptive reporting that could have been delivered with significantly simpler technology. The platforms have been purchased; the organisational and process changes required to use them at their designed capability level have not been made.
Level two, Diagnostic analytics, adds the ability to understand why something happened. Organisations at this level can investigate deviations from expected performance, identify contributing factors through drill-down analysis, and segment data to isolate the sources of variation. Approximately 24% of mid-market organisations and 31% of enterprise organisations in our sample operate reliably at this level. The primary barrier to reaching it from level one is data quality, diagnostic analysis requires clean, consistent, granular data that many organisations do not yet have.
Level three, Predictive analytics, enables probabilistic statements about what is likely to happen. This level requires statistical modelling capability, sufficient historical data for model training, and the operational infrastructure to generate and distribute predictions at the frequency and latency required for them to be actionable. Approximately 11% of mid-market and 19% of enterprise organisations in our sample have reliable predictive analytics capability in at least one business domain. The barriers at this level are primarily talent and data, skilled modellers and clean, labelled historical data are both scarce.
Levels four and five, Prescriptive and Autonomous analytics, remain rare in our sample, with meaningful capability found in 4% and 1% of organisations respectively. Prescriptive analytics recommends specific actions rather than simply predicting outcomes. Autonomous analytics closes the loop entirely, implementing decisions without human review for defined decision classes. Organisations at these levels have typically invested for five or more years in building the data, talent, and infrastructure required, and operate in domains where decision frequency, data availability, and outcome measurability make the investment viable.
Data culture is the single most powerful predictor of analytics maturity level in our research, outperforming technology investment, team size, and organisational scale. Organisations where leadership consistently uses data in decision-making, where analytical findings are treated as credible inputs rather than challenges to intuition, and where data literacy is treated as a core competency show maturity levels significantly above those predicted by their technology investment alone. Culture is also the most difficult characteristic to change, and the one that receives the least structured attention in analytics improvement programmes.
The practical improvement agenda for organisations seeking to advance their analytics maturity differs substantially by current level. Moving from level one to level two requires data quality investment, not new tooling. Moving from level two to level three requires analytical talent and historical data curation. Moving from level three to level four requires decision workflow redesign and the organisational trust in model outputs that only comes from sustained track record. Understanding which barriers are relevant at the current maturity level prevents the common mistake of investing in the wrong constraint.
