A utility cuts electricity use without touching prices, rebates, or technology — just by telling households how they compare to their neighbors

问题

Getting people to voluntarily use less of something costs money (rebates, subsidized equipment) or requires them to already care about the abstract goal (saving the planet, saving money on a small monthly bill)

背景

换别人会怎么做

Raise electricity prices or fund rebates for efficient appliances -- the two standard levers utilities had for reducing energy use, both of which are expensive, politically difficult, or only reach the customers already inclined to act, and straight information campaigns urging conservation had a poor track record on their own.

他们看到了什么

Yates and Laskey saw that a simple raw comparison of a household's usage against its neighbors could backfire -- prior studies had found a 'boomerang' effect where below-average users, learning they already used less than typical, felt licensed to use more. The fix wasn't abandoning peer comparison, it was pairing the comparison with a small normative signal (a smiley or frowning face) that kept even already-below-average households nudged toward further improvement instead of relaxing.

那一手

Starting with the Sacramento Municipal Utility District and Puget Sound Energy in 2008, Opower mailed households periodic Home Energy Reports comparing their electricity (and in PSE's case, gas) usage to their 100 most similar nearby neighbors by home size, with no rebate, rate change, or new equipment attached — only the comparison itself, softened with an emoticon to signal whether usage was better or worse than the norm.

为什么管用

Because most people care how they compare to a relevant peer group more than they respond to price signals or abstract appeals to save the planet, showing a household's usage against its 100 most similar nearby neighbors supplied a motivation that cost the utility almost nothing to generate -- no rebate, no rate change, no new equipment. Adding the normative emoticon closed the specific failure mode straight comparison data had shown in prior studies, so even customers already using less than average kept a reason to improve rather than coast, which is why the effect held for 7 to 12 months rather than fading immediately.

值了多少

A 2009 NBER analysis of the two field experiments, covering roughly 75,000 households, found sustained reductions in energy consumption of 1.2% at Puget Sound Energy and 2.1% at Sacramento Municipal Utility District, with the effect holding for 7 to 12 months after the reports began — cutting energy use at a cost per kilowatt-hour saved that was competitive with, and often cheaper than, building new generation capacity.

什么时候会失灵

This mechanism depends on the peer group being genuinely comparable (similar home size, similar local climate) -- a badly matched comparison group undermines the credibility of the signal and can produce no effect or even resentment. It also risks the same boomerang effect it was designed to avoid if the normative signal is removed or if a population becomes numb to the comparison over time, and it caps out at whatever behavioral slack exists in a given household's usage; it cannot substitute for genuine efficiency investment once the easy behavioral gains are exhausted.

后来呢

Opower scaled the model to utilities serving more than 50 million households worldwide before its 2016 acquisition by Oracle for $532 million, and its combination of descriptive social comparison with a normative nudge to counter the 'boomerang' effect became the reference design cited across the behavioral-economics and energy-policy literature for using peer comparison, rather than price or persuasion, to shift consumption at scale.

资料来源

1. [1]Evidence from Two Large Field Experiments that Peer Comparison Feedback Can Reduce Residential Energy UsageNational Bureau of Economic Research (Ayres, Raseman & Shih), 2009nber.org
2. [2]Opower's Alex Laskey and Daniel Yates want to transform the electric industryHarvard Magazine, 2015harvardmagazine.com

收下它

同一路聪明

• 2010 · American ExpressAmex didn't advertise its card on the biggest shopping weekend of the year — it invented a whole new day inside it
• 1962 · AvisAvis told customers outright it was only #2, and the confession made everything else it said believable
• 2025 · AXA France (with Publicis Conseil)AXA covered domestic violence like a house fire by changing three words in a contract millions of people already had