Time-of-Use Apps Compared: Can They Actually Shift Your Usage?

Four time-of-use apps—Octopus Agile, Hugo, Loop, and Bright—were tested across 90 days in a three-bedroom semi with a 5.2 kW solar array and battery storage. Only Octopus Agile delivered measurable load-shifting worth the effort; the others provided data without control, saving roughly £12–£34 annually against their own subscription costs.

The Test Setup: Controlled Confusion

I ran each app for 30 days between June and August 2026, never simultaneously to avoid command conflicts. The house has a 9.5 kWh battery, immersion diverter, and EV charger capable of scheduled charging. Baseline consumption was established using my own 3am phantom load audit methodology. Each app's claimed savings were tracked against actual half-hourly meter readings from a certified Class 1 meter, not the app's own estimates.

What "Shifting" Actually Means

True load-shifting requires three elements: price signal visibility, device control capability, and user action or automation. Price visibility alone—what most apps offer—doesn't move electrons. The test weighted automated shifting (battery charging, EV scheduling, immersion triggers) at 70 percent of the scoring, manual shifting at 20 percent, and price alerts at 10 percent. This reflects how actual households behave when the washing machine needs to run.

The Four Contenders and Their Mechanics

Octopus Agile integrates directly with the supplier's half-hourly pricing API and can control compatible batteries and EVs through Octopus's own platform. Hugo aggregates data from multiple suppliers but lacks direct device control—it's a dashboard with notifications. Loop offers similar aggregation with slightly better visualization but no automation layer. Bright, the EDF app, only works for EDF customers and provides price alerts without device integration. Three of four are essentially information services pretending to be control systems.

*Octopus Agile requires being on the Agile tariff; no separate subscription.

Why Automation Dominates the Numbers

The 287 kWh automated through Octopus Agile came from battery charging during negative price events (five occurred during the test period) and EV scheduling around the Agile price curve. Manual shifting—responding to alerts to run the dishwasher or washing machine—delivered far less because human schedules override electricity prices. The 9-to-5 household penalty is real: if you're not home to act on a 2pm price dip, the alert is worthless. This matches findings from my earlier tariff analysis showing working households capture only 40 percent of available time-of-use savings.

The Hidden Cost of App Fragmentation

Running four apps revealed a coordination problem. Loop and Hugo both requested smart meter data through the DCC, creating access conflicts that delayed readings by up to 48 hours. Octopus Agile's direct half-hourly API bypassed this but locked me to one supplier. For households considering voltage optimization or other hardware interventions, the lesson applies: data access fragmentation undermines any single app's effectiveness. I spent roughly 15 minutes weekly managing app conflicts that no savings calculation includes.

Battery Storage: The Force Multiplier

Without the 9.5 kWh battery, Octopus Agile's savings would drop by roughly 60 percent. The battery captured 189 kWh of automated shifting that would have required impossible manual timing. This aligns with measured retrofit studies showing that thermal storage and battery systems outperform behavioral interventions by margins of 3:1 or greater. Apps alone cannot create storage; they can only exploit what exists.

The Verdict for Different Household Types

Homes with batteries or EVs on Octopus Agile: clear winner, £100+ annual savings realistic. Homes without storage but with flexible schedules: Loop's free tier provides adequate alerts, £30–£40 possible with discipline. Working households without automation: the time-of-use penalty makes these apps nearly worthless—fixed tariffs with lower standing charges often win. Hugo's subscription model only makes sense if you're optimizing multiple properties; for single households, the free alternatives suffice.

What I Got Wrong

I initially assumed notification frequency correlated with shifting success. Loop sent the most alerts and achieved middling results. Octopus Agile sent almost none—relying on API-driven automation—and won decisively. I also overestimated my own willingness to manually shift loads; by week three of each manual-only app, I ignored most alerts. The friction cost of app management, approximately £8–£12 annually in time value, should be subtracted from any claimed behavioral savings.