Smart metering can play a bigger role in tackling water loss

From meters to Megalitres

To reduce demand, smart meter data needs to be acted on, not just collected. Gary Adams urges analytics and engagement be factored in upfront.

This summer made the case for water resilience impossible to ignore. While the public may not follow reservoir stocks, the strain on the system is obvious to everyone. Population growth, climate change and the need to reduce unsustainable abstraction are placing increased pressure on water resources. Without further action, England could face a shortfall of five billion litres a day by 2055. Changing that depends on infrastructure investment and decision-making at a household level.

The installation challenge

Smart metering is the sector’s fastest lever to pull. During AMP8, water companies plan to install 10.4m smart meters across England and Wales, taking the total installed base beyond 12m by 2030. It’s one of the largest coordinated asset deployment programmes the sector has attempted and a critical enabler of demand reduction commitments.

Meters are often hard to reach, records are incomplete, and access is unpredictable. At scale, that means thousands of customer interactions per week that need to be done right first time, because a second visit multiplies cost and erodes public patience. Once installed, the meter must be commissioned correctly, communicate reliably, and keep working for years in a difficult below-ground environment. Get this wrong, and it counts on the installation report, but does nothing for demand reduction.

Data to action

Turning data into action is the step that actually saves water. Yet, we have spent years working through the complexity of how to install meters at this scale and are only now looking at how to drive value from the data. Some companies are further down this road than others, but, as a sector, we have not industrialised this enough.

A meter records consumption; it doesn’t reduce it. A programme can look successful on installation alone, while the demand curve barely moves. Getting from meters to megalitres is a different discipline, and one the sector has underinvested in.

Savings only follow when reliable data triggers the right operational or customer intervention. Turning meters into actual water savings takes three things: complete and timely data; analytics that can distinguish a leak from ordinary use; and a route to the customer that shifts behaviour.

A property should show near-zero flow overnight. A steady trickle usually means a leak, once you can separate it from legitimate use. This is a genuine step change: the small, continuous customer-side leaks that traditional meters could never surface are exactly the ones smart data makes visible, and they add up across millions of homes. That matters because Ofwat estimates between 1 and 1.5bn litres are lost through leaks in homes and businesses across England and Wales every day.

Contextual data

The same data that identifies continuous flow at a property can, when combined with DMA, network and operational data, improve a company’s understanding of where water is being consumed and lost. Smart metering needs to become part of the wider approach to leakage, network operations and water-resource planning.

Making that integration work is an organisational challenge, as well as a technical one. Too often, installation is treated as the core programme, while data, analytics and customer engagement are managed as downstream workstreams. If the intended outcome is demand reduction, all four need to be designed, funded and governed together. Operationally, that means treating each point of interest as an intervention with an owner, expected outcome and closed feedback loop.

The obvious objection is that data use cases can wait, because companies are already stretched, hitting installation and connectivity targets. But if the operating model for acting on the data is bolted on after millions of meters are installed, it gets retrofitted under pressure, at a higher cost, against a demand curve that still hasn’t moved. The cheapest time to build it is now, in parallel, while volumes are still ramping.

Customer engagement and behaviour

The same applies to the route to the customer. The sector cannot assume that installing a smart meter creates an engaged customer. Many won’t know it’s there or why the data matters. Engagement has to begin before installation and continue for the life of the meter.

Turning an alert into a sustained saving is a behavioural problem as much as a data one. We’re supporting the industry in solving this by pairing granular consumption data with behavioural science through our customer engagement capability. Consumer Council for Water research from 2025 found customers prefer to compare their current use with their own past usage rather than with similar households. That information needs to be simple, with reminders, incentives and tailored support to maintain engagement.

In one of our large-scale deployments, this approach reduced consumption by 3.5 megalitres a day (Ml/d). Households found leaks they didn’t know they had and stopped paying for water they weren’t using.

The AMP8 metering Price Control Deliverables go beyond installation volumes to cover connectivity and data completeness too. This matters because a meter that fails to trans mit consistently can’t support effective demand management. But connectivity and completeness are still measures of whether companies can get the data, not whether they can use it. Ofwat’s cross-industry Smart Water Metering capability maturity assessment, delivered with Baringa, found companies further ahead on getting the data than on using it. Governance and connectivity monitoring are progressing, but customer and demand-reduction use cases are only just emerging.

That distinction is no longer just an internal operational concern. Ofwat’s assessment now makes it visible, company by company. Those that close the gap between collecting data and acting on it will turn their meter investment into a measurable dent in demand.

The meters are coming; the real question for the sector is whether it has built the operating model to make them count.