A "Water Score" to promote smart water use

In the past two decades we have seen an explosion in the use of Credit Scoring by the financial services industry. Virtually every bank, mortgage company, auto lender, credit card issuer and insurance company relies upon these statistical measures of credit worthiness. And in recent years we've seen consumers themselves becoming more and more aware of and focused on their scores as well. TV commercials, personal finance advocates, books and web sites abound to educate the consumer on the meaning of their credit score and how to improve it.

Credit scores look at an individual's credit history and attempt to predict the credit-worthiness of that individual going forward. Certain score cut-off points are well-known (eg. "good" credit starts at 640, etc) so that individuals have a sense of where they stand in the credit "pecking order." And knowing this information has proven to be a strong motivator for many people to improve, or maintain, their good standing. As one leading financial guru says: "Knowing your credit score can be empowering-if it's low you can take steps to improve your credit worthiness and if it's high you may be able to use it as leverage when shopping for a loan." It is this "self awareness" aspect of Credit Scoring that is so intriguing to me as I think about water consumption.

Hoping to motivate people in a similar fashion to conserve water, we recently introduced a scorecard-based service for creating a "Water Score" on each water utility account. Numerous statistical characteristics are derived from each account's usage history and historical data at the 'customer type' level. These characteristics are fed through a user-editable scorecard that can be tuned to generate a "water score" for each account.

By examining these characteristics we can accurately segment the customer base to identify those customers who are using high volumes of water and those whose meters may be stuck. Credit scores evaluate a much richer set of variables and are doing something much more difficult; predicting future behavior based on past behavior. The Water Score is simply ranking customers into five broad categories which could be described as "people with likely bad meters", "people who are using a very small amount of water", "people whose water usage is ok", "people with somewhat elevated water usage" and "people with really high water usage." We're not impacting the customer's water bill or his/her ability to get service in any way; we're simply providing a feedback mechanism to alert customers who may want to consider ways of reducing their water usage.

The Water Score is also a great way to guide customer interactions. Customers whose usage scores on the low end (grey) should be the focus area for bad meters. Customers scoring "yellow" and "red" can be focused on for conservation messaging and customer outreach. A separate model detects possible customer leaks.