KT Score · Methodology
TL;DR · Last reviewed 12 June 2026
The KT Score rates UK energy suppliers out of 100 using only three published datasets: Energy Ombudsman accepted complaints per 100,000 accounts (weight 45 percent), the Citizens Advice customer service star rating (40 percent) and the Ofgem enforcement record over a trailing 24 months (15 percent). Missing metrics are never imputed: their weight is dropped and the rest rescaled. The engine is deterministic, so identical inputs always return identical scores.
KT Score methodology. Data to Q4 2025. Last reviewed 12 June 2026.
The Kael Tripton Score (the KT Score) is a single number out of 100 for each UK domestic energy supplier. It is built only from published regulator and ombudsman data. It contains no opinion, no star ratings invented here, no user reviews, no commission, and no pay-to-rank arrangement. The phrase that sits behind it is plain: scored by data, not commission. This page sets out exactly how the number is produced, so any reader can reproduce it from the same public sources.
The three datasets
The KT Score uses three primary sources and nothing else. Each is a regulator or statutory ombudsman publication.
- Energy Ombudsman, Disputes Data. Accepted complaints per 100,000 domestic consumer accounts, published quarterly. Current vintage used: Q4 2025 (October to December 2025). Source: energyombudsman.org/complaints-data. The Energy Ombudsman reports per-supplier figures only for suppliers with more than 1,000,000 consumer accounts.
- Citizens Advice, domestic energy supplier customer-service star rating. An overall rating out of 5, published quarterly. Current vintage used: Q4 2025 (October to December 2025). Source: citizensadvice.org.uk supplier customer-service table.
- Ofgem, enforcement and compliance. Formal enforcement actions and guaranteed-standards records over a trailing 24-month window. Window used: June 2024 to June 2026, reviewed 12 June 2026. Source: ofgem.gov.uk investigations, orders and penalties.
No secondary coverage is used. Press articles, price-comparison sites, and commercial review platforms are excluded by design. Only the three sources above and gov.uk material sit behind any figure.
The formula
The KT Score is a weighted average of three normalised metrics, each scaled to a 0 to 100 range before weighting:
KT Score = round( 0.45 × A + 0.40 × B + 0.15 × C )
- Metric A, weight 45 percent: Energy Ombudsman accepted complaints per 100,000 accounts, converted to a percentile rank across the scored suppliers and inverted, so that fewer complaints produces a higher value.
- Metric B, weight 40 percent: Citizens Advice star rating, scaled from its 0 to 5 range onto 0 to 100 by multiplying the rating by 20.
- Metric C, weight 15 percent: an Ofgem enforcement value that starts at 100 and is reduced for documented regulatory action, as set out below.
The result is rounded to the nearest whole number and shown out of 100 with a vintage line, for example: KT Score 76 out of 100, data to Q4 2025.
How Metric A is normalised
Complaints per 100,000 accounts is a count where a lower figure is better. For each scored supplier the percentile rank is calculated as the share of scored suppliers with a higher (worse) complaints figure, plus half the share that share the same figure, expressed out of 100. This places the supplier with the fewest complaints near the upper end of the 0 to 100 band and the supplier with the most near the lower end, with the spacing set by rank rather than by raw gap. The population for this calculation is the set of scored suppliers that disclose the metric in the current quarter.
How Metric B is normalised
The Citizens Advice rating is already a like-for-like score out of 5 across suppliers. It is mapped linearly onto 0 to 100 by multiplying by 20. A rating of 3.78 out of 5 becomes 75.6; a rating of 2.28 out of 5 becomes 45.6. No other adjustment is applied.
How Metric C is calculated
Metric C begins at 100 for every supplier and is reduced only where a primary Ofgem source documents action inside the trailing 24-month window:
- minus 40 where Ofgem has imposed a formal enforcement action, meaning a financial penalty, a consumer redress order, or a published enforcement decision, dated within the window;
- minus 20 where Ofgem has flagged a guaranteed-standards-of-performance breach pattern within the window that falls short of a formal action.
Routine compliance-monitoring mentions, such as scheme-delivery notes in a Supplier Performance Report, are not treated as formal enforcement and do not reduce Metric C. Actions older than the 24-month window do not reduce Metric C; they are recorded in the supplier notes for transparency but carry no score effect. A supplier with no qualifying action keeps the full 100.
Why the weights sit where they do
The three weights reflect how directly each dataset measures realised consumer harm. Metric A carries 45 percent because accepted complaints per 100,000 accounts is the most direct published signal of problems that customers could not resolve with the supplier and escalated to an independent ombudsman. Metric B carries 40 percent because the Citizens Advice rating is a structured, like-for-like assessment of customer service quality across suppliers, measuring areas such as ease of contact, billing accuracy, and complaint handling. Metric C carries 15 percent because formal enforcement is a strong but infrequent and binary signal: most suppliers in most quarters carry no qualifying action, so a smaller weight prevents a single historic event from dominating an otherwise current picture. The weights are fixed across suppliers and across quarters; they do not vary by supplier, and any change to them is logged in the change log below.
A worked example
Take a supplier with a Citizens Advice rating of 3.78 out of 5, the lowest complaints figure among the scored group, and one qualifying enforcement action inside the window. Metric B is 3.78 multiplied by 20, which is 75.6. Metric A, as the lowest complaints figure, sits near the upper end of the percentile band; with five suppliers and no ties it resolves to 90. Metric C is 100 minus 40, which is 60. The weighted average is 0.45 multiplied by 90, plus 0.40 multiplied by 75.6, plus 0.15 multiplied by 60, which is 40.5 plus 30.24 plus 9, totalling 79.74, rounded to 80. The same three lines of arithmetic produce every figure in the league table, which is why the engine is described as deterministic: identical inputs return an identical score on every run.
The renormalisation rule
If a supplier is missing from a dataset, because it is too small or too new to appear, that metric is recorded as n/d (not disclosed) and is never imputed or guessed. When a metric is n/d, its weight is removed and the remaining weights are rescaled so they still sum to 1 before the weighted average is taken. A supplier scored on two of the three metrics is therefore scored fairly against those two, rather than penalised for a gap in public data. The current quarter has all three metrics present for every scored supplier, so no renormalisation is active, but the engine applies the rule automatically when a gap appears.
What the KT Score is not
The KT Score is a compression of public regulatory data into one comparable figure. It is not a recommendation to switch to or away from any supplier. It is not a prediction of the service any individual household will receive, since complaint rates and star ratings are aggregates. It is not a price ranking; tariff costs change every cap period and are covered separately. It is not financial advice and it is not a substitute for checking current tariffs, exit terms, and eligibility before any switch. It carries no commercial weighting: no supplier can pay to raise its figure, and no affiliate or commission relationship feeds into it.
Reproducibility and change log
The score engine is deterministic. Running it against the same recorded raw values returns the same figures every time. Each published figure traces to a recorded source URL and publication vintage, held alongside the data in the scoring file. When a new quarter is published, only the raw input block changes; the formula and weights stay fixed unless a methodology change is logged here.
- v1.0, 12 June 2026: first publication. Datasets at Q4 2025 vintage for Energy Ombudsman and Citizens Advice; Ofgem enforcement window June 2024 to June 2026. Five suppliers scored.
Disclaimer
This page and the KT Score are informational only and do not constitute financial, legal, or consumer advice. Figures reflect the published data vintages stated and may change when regulators publish new quarters. Always check a supplier's current tariff and terms before making any decision.