Ordnance Survey has warned that major projects across the UK’s telecoms, utilities, and financial services sectors are running five months late on average because of growing decision complexity and weak confidence in data.
The findings form part of OS’s Building Decision Confidence Report, based on research among 251 senior decision makers in UK organisations across utilities, telecoms, and financial services. All respondents managed projects worth £100,000 or more, and the data was collected between 20 and 27 March 2026.
The research found that more than one in five leaders, at 22%, regularly have to revisit major strategic decisions because of incomplete or inaccurate data. Those decisions cover multimillion pound investments in areas such as infrastructure expansion, outage response, insurance, and climate mitigation.
OS said the results reveal a growing confidence deficit at the heart of UK decision making. Three quarters of businesses, at 76%, said decision making has changed significantly over the past five years. Four in ten said decisions are more complex than ever, 35% said they take longer to make, and 37% said they are having to make significantly more strategic decisions than before.
At the same time, expectations of speed are rising. More than a quarter of leaders, at 28%, said they feel expected to move faster than ever. A fifth said the volume of data now available has left them with a less joined up view than they had five years ago, while 23% reported greater decision fatigue.
The issue is being amplified by AI adoption. OS found that 71% of leaders are not confident in the data used to generate AI insights, while more than a fifth are not confident in the data used to train the models behind them. The company said this creates conditions for errors to be learned and repeated while costs and delays accumulate.
Four in ten respondents said better access to high quality, trusted data would directly enable faster decision making, while 34% said trusted location data specifically helps them make more confident decisions and reduce costly project delays.
Tina Kennedy, Chief Customer Officer at Ordnance Survey, said: “Businesses today are making the most consequential decisions of a generation, and they’re making them faster than ever before – from planning resilient infrastructure for a changing climate to accelerating next-generation network rollouts. But moving quickly is not the same as getting it right. When the data driving those decisions is incomplete or inconsistent, organisations pay the price twice – once in the bad decision, and again in the cost of remaking it.”
She added: “As AI increasingly influences strategic decision-making, the quality and provenance of the underlying data that powers it becomes even more critical. That’s why organisations working in these critical industries, and in an increasingly complex world, need to look at location first.
“When leaders have trusted, accurate location data, they have the confidence to move quickly and decisively, connecting the dots between operational, customer and network intelligence in ways that fragmented or incomplete data simply cannot. That is what enables organisations to get decisions right the first time.”
The report exposes a practical limit to AI driven transformation. Businesses have more data than ever, but volume is not the same as trust. Fragmented, inconsistent, duplicated, ageing, or poorly governed data can slow decisions rather than accelerate them. AI can make that weakness more visible because models depend on the quality, context, and provenance of the information feeding them.
The stakes are high in utilities, telecoms, and financial services. These sectors make capital intensive decisions that affect networks, resilience, service levels, regulation, risk, and customer outcomes. A delayed infrastructure project can affect capacity, outage response, and regulatory commitments. A delayed financial services project can affect risk controls, customer service, compliance, or product delivery.
The five month delay finding also points to a governance problem. Many organisations have invested in analytics, dashboards, cloud platforms, and AI tools, but still struggle to create a single trusted view for important decisions. That weakness often sits between business functions rather than inside one department. Operations, finance, technology, risk, customer teams, and external suppliers may all hold different versions of critical data.
Fragmented data is already weakening customer and loyalty work in financial services, where AI personalisation depends on a reliable view of the customer: Fragmented data slows AI loyalty gains. OS’s research broadens that problem into infrastructure, utilities, and strategic planning.
The management task is not simply to buy better tools. Trusted data requires ownership, standards, stewardship, auditability, integration, and clear accountability for quality. Leaders also need to know which data is good enough for operational decisions, which requires further assurance, and which should not be used for automated or AI influenced outputs.
As AI becomes more embedded in planning, data confidence will sit inside risk management. Fast decisions based on weak data create rework, regulatory exposure, cost overruns, and customer harm. Slow decisions caused by low trust can be just as damaging. Advantage will sit with companies able to move quickly because their data foundations are strong enough to support confidence.
You must be logged in to post a comment.