As organizations invest more heavily in data, expectations change. Data teams are no longer judged solely on their ability to deliver pipelines, dashboards, or models, but on whether those capabilities can be sustained as the organization grows.
At this stage, progress often slows. Delivery continues but responsiveness declines and confidence in outputs becomes harder to maintain. The causes are rarely a lack of talent or technology. More often, they stem from practices that once enabled speed but now introduce friction and risk.
Across industries, the most effective data teams share a common trait. They are disciplined about which processes they allow to persist as complexity increases.
In this article we’ll look at some of the common issues that stop data teams from going from good to great, and explore some options to get your team to that level.
Bespoke scripts and narrowly scoped solutions usually emerge for sensible reasons. They address immediate needs quickly and can appear to reduce pressure on delivery teams.
However, as these solutions accumulate, their hidden costs become more apparent. Each script adds undocumented logic, informal ownership, and dependencies that are difficult to trace. Over time, this erodes visibility into how data is transformed and where failures might originate.
This pattern closely mirrors how technical debt develops in large systems. Gartner’s analysis of how organizations manage technical debt at scale show that accumulated, unmanaged shortcuts reduce adaptability and increase operational risk as platforms grow in complexity. The longer these patterns persist, the harder it becomes to change systems safely.
More mature data teams therefore limit reliance on one-off fixes. They prioritise shared, governed patterns that can be extended and modified without increasing fragility. While this approach may appear slower initially, it enables faster and safer change over time.
What great teams do instead
Manual data exchange remains widespread, particularly through spreadsheets and CSV files shared between teams. While this approach offers short-term flexibility, it introduces ambiguity into data workflows.
When data is manually copied, uploaded, or modified, lineage becomes difficult to establish and accountability unclear. As a result, investigating discrepancies or validating accuracy takes longer, and confidence in reported figures diminishes.
Research into the data-driven enterprise of 2025 shows that organizations only realise consistent value from analytics when they can trust the data flowing through their systems. Manual processes make that trust harder to maintain as decision-making becomes more distributed and time-sensitive.
For this reason, high-performing data teams treat manual handovers as transitional rather than foundational. They invest in automation and observability to ensure that data flows remain transparent, auditable and resilient as usage grows.
What great teams do instead
Shadow tools are often considered a governance problem but they typically emerge for more practical reasons. When official platforms are slow to adapt or difficult to use, teams seek alternatives that allow them to move forward.
Over time, this leads to fragmented logic, duplicated effort, and inconsistent results across the organization. Attempts to eliminate shadow tools through stricter controls alone tend to be ineffective, as they do not address the underlying causes.
We’ve found that adoption improves when platforms balance governance with usability. Organizations that reduce friction in core systems see higher compliance and less fragmentation, not because teams are constrained, but because supported workflows better reflect how work actually happens.
In doing so, mature data teams reduce reliance on unsanctioned tools while preserving autonomy.
What great teams do instead
High demand for data is often interpreted as a sign of success. Teams remain busy responding to requests, addressing issues, and reconciling definitions across stakeholders.
However, studies of high-performing technology organizations consistently show that constant reactivity is not correlated with long-term effectiveness. According to Forrester’s State of Enterprise Data and Analytics, businesses with established governance and automation are consistently less mired in manual firefighting and more focused on systematic improvements.
More effective data teams monitor indicators such as lead time for change, frequency of rework, and stability of pipelines. These measures provide a clearer picture of whether the team is enabling the organization or simply absorbing its complexity.
Over time, this shift in focus supports more predictable delivery and greater confidence among stakeholders.
What great teams do instead
Incremental improvement can create a misleading sense of maturity. Teams naturally compare their current state to where they started, rather than to external benchmarks or peers operating at a similar scale.
Without structured comparison, it is difficult to identify which practices are genuinely strong and which persist largely because they are familiar.
Benchmarking introduces an external perspective. By comparing operational practices across areas such as alignment, agility, governance, and enablement, teams can better understand where progress has stalled and where change is likely to have the greatest impact.
What great teams do instead
Maturity in data operations is defined less by capability than by discipline. The most effective teams understand which practices no longer serve them and take deliberate steps to replace them with more sustainable approaches.
For organizations seeking that clarity, structured assessment can provide useful context. Our new Data Operations Maturity Assessment evaluates how teams operate across strategic alignment, operational agility, data fluency, governance, team enablement and AI-readiness.
It offers a practical way to understand which behaviours support scale and which ones may be holding teams back. Spend 7 minutes completing it and come away with a personalized score and bespoke steps to level up your data operations.