Octopus Cosy heat-pump 99% claim: what it really means

Octopus published the claim on 2 September 2026 using anonymised data from thousands of its Cosy heat pumps. The measured period ran from 1 September 2025 to 31 August 2026 and compared actual Cosy heat-pump performance with the cost of producing the same heat from an 85%-efficient gas boiler.

The result matters because it moves the discussion beyond a modelled home. It still needs careful reading. The claim is about running cost inside Octopus's stated comparison, not the purchase price or a guaranteed saving for a particular property.

What Octopus actually compared

The press release says Cosy systems in the measured fleet produced more than four units of heat for each unit of electricity used. Octopus reports a fleet coefficient of performance of 3.97 for the year. It then compared the cost of that heat in two ways:

• the measured electricity used by Cosy heat pumps, charged at the Cosy Octopus rates that applied during the period
• the gas needed to produce the same heat with an assumed 85%-efficient boiler, charged at the matching Flexible Octopus gas rates

Octopus says customers on Cosy saved £227 a year on average in this dated fleet comparison. That is an average across the dataset, not a fixed credit or a forecast for a new installation.

Gas standing-charge savings need separate attention. Octopus's running-cost methodology includes them only for customers who removed their gas supply. A home that keeps gas for cooking, a fire or another appliance still pays the gas standing charge.

Why the live dashboard may show a different number

The 99.6% and £227 figures belong to the year stated in the 2 September release. Octopus also runs a live Cosy fleet dashboard, where the period, tariff inputs and rolling performance can move.

A nearby percentage or average on the dashboard does not make the press release wrong. It is a different snapshot. Read the date range and methodology beside any live figure rather than treating 99.6% as an evergreen result.

What the claim does not include

The headline is about heating running cost within that comparison. It does not settle the whole project decision. A household still needs to consider:

• the heat-pump purchase price, grant eligibility and any finance cost
• radiator, underfloor-heating, cylinder, pipework or electrical work
• maintenance, disruption and the condition of the heating system being replaced
• the home's heat loss, insulation, comfort settings and hot-water routine
• the actual postcode tariff rates and whether the household can use Cosy's cheaper periods

This is why the strong conclusion is narrower than the headline. A well-designed and well-run heat pump on a suitable tariff can cost less to run than the gas comparison across almost all of Octopus's latest measured fleet. It does not prove that any heat pump will beat any boiler in any home.

COP is not the same as a full-year result

COP shows the heat produced for each unit of electricity at a particular point or over a stated period. Seasonal performance factor, or SPF, covers performance across a longer period with changing weather and hot-water demand. SPF is usually more useful when planning an annual bill.

Do not apply the fleet COP of 3.97 directly to your quote. Ask the installer for the expected annual heat demand and seasonal performance of the proposed design. The assumptions should include flow temperature, room temperatures, hot-water use and any radiator or underfloor-heating changes.

Energy Saving Trust also says heat-pump running costs depend on radiator sizing, the electricity tariff and how the system is controlled. It notes that a time-of-use tariff can reduce costs, while a standard electricity tariff can leave a heat pump slightly more expensive than a newer gas or oil boiler in some homes.

How to sense-check a heat-pump quote

1. Ask for the design heat loss and expected annual heat demand.
2. Ask for the proposed seasonal performance factor, not only a best-case COP.
3. Check which radiator, cylinder, pipework and electrical changes are included.
4. Compare current postcode rates for Cosy and at least one simpler tariff.
5. Model the bill with and without the gas standing charge if you may keep gas connected.
6. Keep the Boiler Upgrade Scheme and finance decision separate from annual running cost.
7. Compare the whole household bill rather than one cheap tariff window.

Cosy currently has three cheaper periods and a higher 16:00 to 19:00 peak. The schedule only helps if the home can shift useful heating or hot-water demand without sacrificing comfort. Check the live Cosy Octopus page and your postcode quote rather than copying an old rate into the calculation.

Check a full heating season after installation

One mild week or one cold spell is not enough to judge the system. Keep a simple record across the heating season: heat-pump electricity use, controller SPF where available, room comfort, hot-water settings and the Octopus statement.

Controller, thermostat and home-automation data can explain what the system did. Octopus billing still depends on smart-meter data and the tariff terms. Reconcile any dashboard estimate with half-hourly readings and the bill before blaming the tariff or the heat pump.

If electricity use is much higher than the design forecast, check flow temperature, weather compensation, hot-water cycles, immersion-heater use and whether the rooms reach their intended temperature. A tariff cannot rescue a badly specified or badly controlled system.

What the October electricity VAT cut changes

The Government says VAT will be removed from eligible household electricity bills from 1 October 2026 while domestic gas remains at 5%. It expects suppliers to apply the change automatically, including to fixed electricity tariffs. That should strengthen the electricity side of a heat-pump comparison during the funded period.

It still does not make every installation economical. The tax change lands alongside the October price-cap period and household results vary with usage, tariff, gas standing charges and seasonal performance. Check the live account quote rather than applying a flat percentage to an older example.

The bottom line

Octopus's 99.6% result is strong evidence that its measured Cosy heat-pump fleet can beat the running cost of its stated gas-boiler comparison. The useful question is not whether every home will save exactly £227. It is whether your quote can support the same ingredients: sound system design, good seasonal efficiency, a tariff routine that fits and a clear decision on keeping gas connected.

Use the fleet claim to ask better questions. Let the property survey, installer calculations, current postcode rates and total project cost decide whether the heat pump works for your home.