Somewhere in England, probably not far from you, a road is being dug up again.
The third time in as many weeks.
The permit is valid, the reports look clean, the dashboards look ok. But the underlying data had already predicted this might happen. No one joined the dots.
That’s the gap between recording information and using it.
We Already Have the Data
Utilities collect more data than ever before. Every permit, reinstatement, FPN, Section 74 and job record feeds into something. The problem isn’t access, it’s action.
Most of what’s collected still looks backwards. Reports explain what went wrong instead of helping teams stop it from happening again. Dashboards show activity, not risk. Data ends up being a mirror, not a guide.
Why Utilities Struggle More Than Most
This industry doesn’t move fast. Long contracts, old systems and strict rules make every change feel heavier than it should be.
Some of the core tools still in use were built before smartphones existed.
So when people talk about predictive data or automation, the idea sounds simple but the reality is hard. It is not just about new software. It is about trust, culture and ownership.
The Data Maturity Ladder
Think of maturity as how well a company turns data into decisions.
| Level | Description | Example |
| Reactive | Data is collected because it has to be. | Permit breaches logged manually at month end. |
| Controlled | Data is visible but not trusted. | Dashboards show delays but reports still drive decisions. |
| Diagnostic | Data explains why issues happen. | Analysis links overruns to missing stop notices. |
| Predictive | Data warns what might happen next. | Alerts show permits at risk before expiry. |
| Prescriptive | Data triggers automatic action. | System reschedules jobs before fines are issued. |
Each step is less about technology and more about behaviour, moving from counting events to preventing them.
Where Are You Really On The Ladder?
A quick test:
- Can you name every active work site without exporting to Excel?
- Can you see which permits are at risk before they expire?
- Can your data explain why issues happen, not just where?
If any answer is no, you are still operating below diagnostic level.
The Real Blockers
Most utilities get stuck in the first two stages.
Not because they lack data, but because the basics are not steady.
- Systems do not connect properly.
- Data quality is poor: missing USRNs, bad dates, weak location data.
- Trust breaks once people spot errors. They go back to Excel.
- Legacy contracts and vendors make integration political, not technical.
Until those problems are fixed, prediction is guesswork.
What Better Looks Like
You can tell how mature a team is by how it reacts when something slips.
- Reactive: run another report.
- Diagnostic: find the cause.
- Predictive: say, “we already saw it coming.”
That is the difference between explaining failure and preventing it.
Building the Base
There is no quick way up the ladder. It starts with boring work.
Fix the basics. Get permit, works and reinstatement data right first. Check quality before analysis. A wrong USRN or missing status will break everything above it.
Keep things consistent. One format, one field list, one standard across all systems and contractors.
Show risk, not volume. Knowing which jobs are close to breaching SLAs matters more than listing every open one.
Learn from patterns. If the same streets or crews keep triggering fines, the issue is not the report, it is the routine.
Automate small things first. Alerts, expiry warnings, basic checks. Start with what saves minutes, not months.
None of it is glamorous. Most of it is maintenance. But prediction only works once the foundations hold.
Why This Matters Now
Regulators expect faster answers. Costs are climbing.
AI is getting cheaper, but good data never is.
The space between reactive and predictive is widening.
Standing still now means falling behind later.
Final Thought
Every hour spent cleaning bad data is an hour not spent improving delivery.
The teams that move up the ladder make better calls and fewer excuses.
Most utilities should already be at least diagnostic. Many are not.
Until data quality, ownership and procurement align, they will stay reactive.
The ones that fix it now will save millions and cut disruption.
The rest will keep paying to learn the same lesson again.
By Duncan Bosworth
Duncan Bosworth is Principal Consultant at Skewb. To speak with him about this topic in more detail, you can connect with him on LinkedIn.
