Energy & Environment · Pilot template
Home energy report rollout
Reduce household electricity use by sending a monthly social-comparison energy report to a random subset of utility customers.
Timeline
12 months (to measure persistence past initial novelty).
Per-participant cost
$3–$8 per household per year for report production, mailing, and analytics.
What we already know
Opower-style home energy reports — comparing a household's usage to similar neighbors — have been tested across 8M+ households in dozens of utility programs. Average effect: ~2% reduction in household electricity use, persistent for years rather than fading. Small per household but large at scale, with cost-per-kWh-saved competitive with traditional efficiency programs.
Hypothesis
Customers receiving monthly social-comparison energy reports will reduce their electricity consumption by 1–3% relative to customers not receiving reports, with the effect persisting for at least 12 months.
Intervention
Send each customer in the treatment arm a monthly mailed or emailed report comparing their previous-month electricity usage to (a) similar nearby households, (b) the most efficient nearby households, and (c) their own past usage. Include 2–3 tailored conservation tips.
Comparison / control
No report. Standard billing communication only.
Outcomes
Primary
Average daily kWh consumption over the trial period.
Secondary
- Enrollment in energy-efficiency programs (insulation, appliance rebates)
- Customer-service contacts about the report
- Cancellation / unsubscribe rate
Required sample size
Computed for α = 0.05, 80% power, balanced allocation. Pick the row whose baseline best matches your jurisdiction, then size at the MDE you can defend.
| Baseline | MDE | Per arm | Total |
|---|---|---|---|
| 30 kWh/day mean (SD 10) | 2% (0.6 kWh) | 4,400 | 8,800 |
| 30 kWh/day mean (SD 10) | 1% (0.3 kWh) | 17,500 | 35,000 |
Need a custom value? Use the sample-size calculator.
Randomization
- Identify residential utility accounts with at least 12 months of historical billing.
- Stratify by census-block-group and prior consumption decile.
- Randomize 1:1 within strata.
- Begin sending reports in the same monthly cycle for all treatment customers.
Data collection
- Monthly billing data from the utility's existing meter / billing system.
- Customer-service contact logs (especially complaints / opt-outs).
- Enrollment in any energy-efficiency program during the trial.
IRB / ethics
Treated as program evaluation in most US jurisdictions. The report contains the customer's own data and aggregated comparisons; no individual neighbors are identified. Standard utility privacy rules apply.
Common pitfalls
- Boomerang effect: customers below the comparison group may increase use unless the report includes a sanction-of-approval cue (smiley face, positive frame). Always include.
- Self-selection out: opt-out rates of 5–10% are normal; report them and analyze ITT.
- Multi-family buildings: meter-to-household matching is harder. Either exclude or pre-validate.
Anchor evidence — registry entries this template draws from
Adopt this template
Tell us your context and we'll adapt the template to your jurisdiction and help you launch. No obligation. Small pilots are typically free.
Mechanism profile
/mechanisms/social-norms →Evidence synthesis
/evidence/energy-conservation-feedback →Other templates
Browse all templates →