- Encouraging companies to take a longer-term view of their customer ratings is helping them to spot – and remedy – previously unrecognised issues.
- Red Olive has helped the industry to optimise capital outlay and allocate resources more efficiently by predicting stresses across its network.
- By minimising customer disruption, Red Olive is helping Thames Water to improve its standing within the industry and secure government funding.
- Water providers who have worked with Red Olive, better understand their network assets and can perform proactive, intelligent maintenance and investment based on patterns identified in its performance data.
- Using IBM SPSS software, Red Olive can accurately forecast issues before they become problems.
Customer satisfaction matters
The UK’s privatised water industry is custodian of a national asset: a network of sewers and mains more than 200 times the combined length of all the UK’s motorways. These need to be maintained and extended to meet consumer demand – and if the consumer isn’t happy with the service they’re receiving, water firms can be fined or have their public funding reduced.
By accurately forecasting likely issues based on previous failures, network stress and customer profiling, Red Olive is helping UK-based water providers to forestall issues before they arise, and reducing the likelihood of consumer dissatisfaction attracting penalties from the regulator.
Although collecting the initial data and analysing it to build each model takes time and investment, the savings that can be delivered off set the cost many times over, while the response it generates from their customers enables firms in competitive, regulated markets, to benchmark well against their peers.
When Thames Water, Britain’s biggest supplier, wanted to improve its customer satisfaction scores it, too, turned to Red Olive, which specialises in helping businesses understand their data. Its long history of working with subscription services, like utilities, made Red Olive the ideal partner for a company that needed to identify patterns to reveal what it was doing well, and what could be improved.
Selecting an appropriate window for analysis
From hygiene to profi ts, everything in the water industry is regulated by Ofwat, which judges each operator’s performance on the basis of satisfaction. It compels water companies to conduct regular surveys, which benchmark each one against its peers on a five-point scale.
When it brought in Red Olive, Thames Water was looking to use this data – and the supplementary surveys it conducts on its own behalf – to help it better understand how what it did as a business impacted how customers felt about the organisation and its service offering.
The raw data, in which the weekly satisfaction score could deviate by more than 10%, suggested that the company’s actions almost immediately had an impact on its customers’ feelings. However, when Red Olive took into account the statistical limitations in the data it could prove to Thames Water that the aggregate rate of change was almost nil. This convinced the company to be less reactive to ongoing fluctuations and instead focus on broader business activities.
However, when Red Olive presented this clarified data to Thames Water, it was able to identify an earlier drop in customer satisfaction that it could address. By looking back, it could see that that the dip coincided with the introduction of a new process for handling incoming customer calls. With this in hand, it set out a practical improvement plan to remedy that part of its business.
Identifying pinch points
Not content to stop there, Thames Water next wanted to know how the outcome of those calls, which often involved maintenance or changes to its capital assets, would affect its ongoing score.
Red Olive therefore performed extensive key value driver analysis with the company’s Developer Services department which, among other things, deals with requests from builders to install water mains, and from individuals who want to extend their home on land that crosses the water network.
Red Olive broke down the department’s work by job type, allowing for the different sample sizes that each one returned, so that it could calculate average customer scores for each task. Comparing these normalised results highlighted which areas were under performing, and thus dragging down the company’s overall rating.
With this data in hand, Thames Water could apportion its resources more intelligently, rather than splitting them equally across each part of its business. Additional staff were allocated to tasks that would have the greatest quantifiable effect on the eventual customer rating and, equally importantly, the company knew not to make changes where it was already achieving optimal scores.
Predictive maintenance through data analytics
Each of these measures focused on better serving the customer, but there’s a second, far larger part of a water company’s business that it also needs to maintain: the pipework, treatment farms, pumping stations and other infrastructure that it uses to deliver its service.
The portfolio is so geographically disparate that it can be difficult to maintain an up to date log of the state of every asset. This means that equipment can break down, causing an interruption to service, contamination or other problems seemingly without warning.
Water company service areas typically cover several counties. Thames’, for example, touches Basingstoke, Guildford, Dartford and points in between, so no two sites will put exactly the same stresses on the hardware, which means that manufacturer’s quoted lifespans are often little better than a rough guide. Once on site, the lay of the land, quality of the water, throughput, contamination and a host of other variables need to be taken into account by any water company, and will greatly affect the actual working life of the asset in question.
Utilities therefore need a more accurate means of forecasting technical disruption than relying on manufacturers’ estimates, which Red Olive provides by profiling each site individually and applying the stresses that it encounters to the asset in situ.
By gathering data from points across the network and analysing issues over time, Red Olive grades each pump and pipe to predict when it’s likely to fail based on historical data. Comparing these results with the manufacturer and model of each asset then enables it to make specific purchase recommendations for each part of the network, taking into account the state of the water or waste being processed at those points. A similar process can then be applied to environmental threats, such as flooding.
Protecting customers from environmental threats
It’s difficult to predict weather-based flooding, but that’s less oft en the case with overflows, which can be caused by backed-up sewers and drains. This kind of flooding can enter customers’ homes and gardens, harms the water company’s reputation, and can lead to both fines and a reduction in its Ofwat grant, on top of the cost of making good.
Red Olive has developed models for predicting likely flooding. These models categorise both static data, such as the lie of the land and number of bends in a pipe, and dynamic data, which includes damage caused by tree roots or the need to clean a silted runoff . These are further sub-divided to take into account whether or not a factor can be controlled. So, while weather is both variable and uncontrollable, the frequency with which you clean a pipe, while still variable, is entirely under the water provider’s remit, and therefore predictable.
By identifying how these variables correlate to incidences of flooding, Red Olive can build a series of predictive models. These break down a utility firm’s service area into smaller zones, and compares factors such as land usage and the size of the pipes to derive a risk level for each.
By repeating this over different time periods it can plot a heat map that highlights locations in which the risk of a flood is increasing. On this basis, the company can proactively allocate maintenance teams to remove sediment from drains and check for pipe damage before the flood risk reaches danger levels.
Securing a better return on your investment
Understanding how customers react to a particular scenario helps companies like Thames Water to allocate resources more efficiently. The result is a satisfied consumer base, and improved performance scores consistently attracting a higher level of funding.
Thames Water – and its peers – are able to proactively allocate their spending across the network, too, both by dispatching maintenance workers to potential problems before they arise, and sourcing hardware that they know is best suited to each point on a case by case basis.
For many companies, the savings delivered by this alone are enough to offset the cost of building the models that they now work from – even before the additional funding is taken into account.
Minimising ongoing costs
Water companies maintain geographically diverse asset portfolios. The ability to predict where problems arise helps them to minimise ongoing costs.
Red Olive profiled each site on the water network to build an accurate map of the vulnerabilities that applied to each location, which affect the working life of each asset in use.