Case Study: Belén Found a Cheaper Way to Save Water – With Smiley Stickers

Rate hikes and awareness campaigns made little difference. But being compared to the neighbors – with a smiley or a frowny face on water bills – cut water use by 5%.

Belén had a water problem that was already expensive, and getting worse.

Key Takeaways: How Belén Used Behavioral Science to Save Water

Give people a reference point.
Most households in Belén did not know whether their own water use was high or low – until they saw how much their neighbors used.

Keep it real, local, and not too broad.
Comparisons to nearby neighbors worked, while broader, city-wide comparisons did not.

Take a little off the top.
Stickers and postcards attached to existing utility bills cost almost nothing and resulted in easy water savings – before making more drastic and costly interventions. No rebates or new infrastructure, just peer pressure.

The ROI:
The intervention cost was reported at $0.10 per household, and the water savings cost about $0.00025 per gallon saved.

Want a framework for turning behavioral research into utility programs? Read how community-based social marketing helps conservation teams remove barriers and make water-saving behaviors easier to adopt.

Article: Community-Based Social Marketing: Increase Water Conservation with Limited Rebate Budgets

Households in the municipality were using about 25% more water than the national average. Projections suggested there’d be shortages by 2030 unless there was either more supply or less demand.

The city had already tried the usual conservation strategies. In late 2012, Costa Rica approved a large rate increase. Consumption dipped for a month, then climbed back. Education and awareness efforts didn’t help. People already agreed that saving water was important, but most did not see their own household as the issue.

In 2014 the city of Belén tried something new. They collaborated with the World Bank and ideas42, a US-based nonprofit that applies behavioral science insights to complex social problems. Their plan: compare household water use to their neighbors.

Why Belén Took The CBSM Approach

The researchers began with focus groups.

People said conserving water mattered. But not many had ideas on how they could lower their consumption. Most didn’t know whether their monthly use was high, low, or ordinary. There was no basis for comparison.

That was a compelling insight. You don’t know if you’re using too much unless you know how much everyone else is using. Humans are social, tribal creatures after all, and there is social pressure to avoid being an outlier and using more than your fair share.

Maybe, all other things being equal, people would be more conscious of their water use if they knew whether they used more than those around them. Social comparison is a powerful tool.

Community-Based Social Marketing (CBSM) makes changing behavior a straightforward and repeatable process.

CBSM is a type of social marketing for a specific group of people. It meets people where they live, literally and psychologically – an evidence-based framework for designing conservation programs.

CBSM focuses on three questions:

Which specific behavior has the greatest impact?
Who performs it?
What barriers prevent changing that behavior?

Then it designs programs that remove those barriers individually. The goal is to close the gap between what people intend to do and what they actually do.

In Belén, the impact they were going for was reducing excessive water consumption. The barrier was that people didn’t know if they used more or less than others in their community. To overcome that barrier – tell them.

How the Pilot Worked on the Ground

The trial included all 5,626 individually metered residential households in Belén. In July 2014, households were randomly assigned to one of three treatment groups or a control group.

The first treatment used neighborhood comparisons

One group received a sticker on the water bill showing how their household compared with the average household in its immediate neighborhood. Households using less than average got a blue sticker with a smiley face. Households using more got a yellow sticker with a frowning face and a few tips for cutting back.

The second treatment used citywide comparisons

A second group got the same basic design, but the comparison was to the city average rather than the neighborhood average.

The third treatment added a planning prompt

A third group received a postcard that asked them to compare their use to local averages, set a goal, and think through specific steps they could take to reduce consumption.

The whole intervention was simple – and cost-effective. The utility did not use the usual conservation tools like new software, educational materials, awareness campaigns, rate hikes, or laws.

Staff could add the stickers to bills by hand and social pressure would do the rest.

What Worked Best?

People liked the stickers. The smiley and frowning faces added a light touch of approval and disapproval. Enough to steer the message, but not so much that it felt like finger-wagging.

The neighborhood comparison carried more weight than the citywide comparison. Why? It could be that a citywide average is too abstract. It doesn’t feel like your community – it feels like statistics.

Your neighbors are different. You see them regularly. They’re human beings, not numbers.

People continuously read one another for cues about what is normal, what is acceptable, and where they stand. The realization you use more than your share might go against your own sense of fairness.

The other side of that same coin is the realization that others are spending less on water than you are, and that feels unfair.

The planning prompt also worked. Asking people to set a goal and think through a few concrete actions helped move the intention closer to behavior.

While it didn’t seem to have a big effect on those households using excessive water, it did have an impact on people who were already conservation-minded.

Some households were already using less water than their peers when the program started. Critics of the program feared these households would use more when they discovered they were using less than their neighbors.

For these households, setting a conservation goal and planning out steps to save more water further reduced their water use.

What Changed as a Result of the Pilot

The effects were modest in percentage terms and meaningful in practical ones.

Households that received the neighborhood comparison reduced water use by between 3.7% and 5.6% relative to the control group
The planning-prompt treatment produced a similar reduction, between 3.4% and 5.6%
The citywide comparison did not produce a statistically significant effect
Both the neighborhood-comparison group and the planning-prompt group reduced consumption by about 1.5 cubic meters (almost 400 gallons) per month
The reductions lasted for more than four months after the single intervention
If scaled across all households in Belén, the study estimated savings of about 6,720 cubic meters per month
The average cost was around ten cents per household

The study also places the pilot in the context of the municipality itself: Belén had 21,633 inhabitants, 99.3% of dwellings had water service, and average water use per household was around 27 cubic meters per month, about 1.25 times the national average.

Why This Matters for Policy and Capital Planning

Belén’s pilot is useful because it shows how little an intervention sometimes needs to cost to produce a measurable result.

A small nudge can have a big impact.

A 5% reduction may not solve a big supply problem. But factoring in the extremely low cost and relative ease of implementation, it bought Belen time to prepare for more drastic expansion.

A utility facing growth, drought, or imminent capital costs benefits from a fast and easy way to cut water use. It’s a cheap way of reducing the current and future load on the system with relatively little effort or cost – which not only postpones the need for costly expansions, but makes them more efficient when they eventually make them.

The pilot also widens the policy conversation a little. Price increases and awareness campaigns are still the default tools in many places. Belén shows there is an easy option when the real problem is not resistance, but weak feedback and lack of context.

How Can This be Replicated?

Start with the households using the most water.
If you want the biggest impact quickly, begin where the waste is concentrated. Pull billing data and identify the highest-use households or neighborhoods first.

Make the comparison local.
Belén’s results were strongest when households were compared to nearby neighbors rather than to the city as a whole. People care more about how they stack up against the households around them than against an abstract average.

Keep the signal simple.
Nothing beats the immediacy of a smile or a frown. There’s a reason why people use emoji – it has more emotional impact than a chart full of numbers.

Keep the delivery simple.
One reason this pilot is still useful is that it did not depend on expensive software or a major redesign of the billing system.

Measure against a control group if you can.
Belén’s results are persuasive because they came out of a randomized trial, not a before-and-after guess.

Remember This

Belén shed light on how much you use compared to your neighbors, and the numbers moved. While 5% is a modest reduction, it was essentially free. It saved thousands of gallons at a cost of 10 cents per household.

Social pressure is a powerful tool! Enough to get people to change their habits – without actually telling them to change their habits. If you tell people to change, they resist. But by shredding a little light, give or giving a little info, that change will feel like it was their own idea.

And it’s a lot harder to resist your own idea than someone else’s.

The Numbers

How much?	For what?
5,626 households	Individually metered residential households included in the trial
3.7% – 5.6%	Reduction from neighborhood comparison treatment
3.4% – 5.6%	Reduction from planning-prompt treatment
No significant effect	Citywide comparison treatment
1.5 cubic meters/month	Reduction for neighborhood-comparison households
1.5 cubic meters/month	Reduction for planning-prompt households
More than 4 months	Duration of the observed effect after one intervention
6,720 cubic meters/month	Estimated savings if scaled across all households
$0.10 per household	Approximate cost of the intervention
About $0.00025 per gallon saved	Cost per gallon saved based on a $0.10 per-household intervention cost and about 1.5 cubic meters of monthly water savings.