In what could put service activations and mergers and acquisitions (M&As) alike at risk, research commissioned by artificial intelligence (AI-)powered geospatial software provider IQGeo has found that only 13% of respondents said that more than three-quarters of their network records are accurate.
The inaugural Ground truth survey 2026: Telecoms report was conducted on behalf of IQGeo by Vanson Bourne, an independent research firm, among 500 senior decision-makers at telecom network operators in the US, Canada and the UK. Two-thirds (66%) of respondents held senior management or board-level positions, and 82% worked at operators serving more than a million subscribers. The respondent base spanned fibre, fixed broadband, mobile, integrated and wholesale operators.
Putting the research into broader context, the study noted that physical network data formed the bedrock of telecom network operations. Moreover, it stated that knowing where every cable, duct and splice point sat, and keeping that record current as field engineers complete repairs and upgrades on infrastructure, was what makes it possible to operate a network.
Yet, warned IQGeo, with thousands of field engineers touching infrastructure daily, relying on manual data entry and disjointed legacy systems to log changes, records could lose sync with reality quickly. It added that in an era of accelerating consolidation, that gap had a direct cost through acquirers conducting technical due diligence not being able verify capacity, model upgrade costs or confidently price what they’re buying.
Among the key findings was that all operators reported business impact from inaccurate physical network data in the past 12 months, with project and service activation delays most common (71%). This toll on companies also extended beyond day-to-day operations, with 78% saying the quality of network asset data has a critical or major influence on valuations during M&A.
Furthermore, 90% of operators revealed that they required engineers to take photos of completed infrastructure work to verify it was completed properly, but just 11% of operators verified them, leaving only 31% of operators confident the work was completed to standard.
Visual AI was the leading field work verification method in the UK (72%) and Canada (71%), but lagged significantly in the US (45%). The average network management system was found to be 13 years old, suggesting that while networking technology has continued to evolve, the systems managing them haven’t kept pace. Fixing physical network data was the top problem operators want AI to solve as 63% targeted autonomous networks by 2030.
Another worrying finding was that paper persisted in the field. Nearly half of respondents (48%) were still using paper notes that were then manually entered into digital network management systems later. Just under a quarter said spreadsheets, computer-aided design (CAD) files or paper-based documentation formed their primary system for managing their network records. This introduced delays and transcription errors undermining the accuracy of network data.
Data updates were noticeably lagging, with only 7% of respondents saying that updates happen immediately once a field worker completes work on the network. Just about half (49%) said these updates happen within one week. Data silos, poor systems integration, data synchronisation issues and manual processes were all cited as reasons for the lag.
Interestingly, operators using a single integrated geospatial system reported a faster update of their system of records (4.5 days), compared to operators using multiple systems (5.3 days).
In a way to resolve issues, 91% of respondents said improved physical network data would help reduce construction and maintenance rework, and 92% said it would improve use of existing infrastructure, reducing the need to overbuild. With 94% believing ESG requirements will shape telecom network operations over the next five years, the report found that making the most of existing infrastructure is becoming just as much an environmental obligation as a commercial one.
“Operators, AI systems, investors and acquirers all need accurate, complete records of the physical network to make decisions. This research shows most operators are a long way from that. Systems are ageing, records are split across multiple platforms and changes in the field are still taking days to flow back to the system of record,” said IQGeo chief technology officer David Cottingham.
“When we asked operators where AI deliver the most value, the number one answer was improving the accuracy of their records, ahead of predictive maintenance and network optimisation. Operators are recognising that physical network data is the foundation for the next wave of telecoms transformation and consolidation.”
