BrandInsightAI Insights
Energy · a live data story

Energy Suppliers,
according to AI

Every week we ask ChatGPT, Claude and Gemini the questions real UK consumers ask about energy suppliers — and record exactly which brands they put forward. Right now, Octopus Energy owns the conversation.

0 AI answers this run · 0 brands named · measured, never estimated

The podium

Octopus Energy is the default answer.

Named in 18% of energy suppliers answers, typically in position 2.5. British Gas follows at 13%, EDF Energy at 8%.

British Gas
13% of answers
Octopus Energy
18% of answers
EDF Energy
8% of answers
The full board

Every brand AI names, ranked.

Visibility is the share of answers naming the brand; citation share is its slice of all brand-owned pages cited; read→cited is how often its pages convert a read into a citation.

#BrandVisibilityAvg. positionCitation shareRead→cited
1Octopus Energy 18% 2.5 10%
2British Gas 13% 1.1 20% 38%
3EDF Energy 8% 2.2 14% 75%
4E.ON Next 8% 3.2 14% 25%
5OVO Energy 7% 4.0 10% 0%
6Green Energy 5% 1.7 0%
7Good Energy 3% 1.0 6% 0%
8Ecotricity 3% 2.0 0%
9ScottishPower 3% 4.0 8% 0%
10Utility Warehouse 3% 6.0 4%
11Boost 2% 2.0 0%
12Shell Energy 2% 4.0 0%
13Fuse Energy 0% 10% 40%
14So Energy 0% 2%
15Utilita 0% 2%
What AI wants here

Energy Suppliers is decided on attraction.

Being named is visibility. Underneath it, every answer describes a brand in three ways — how much it stands out, how much it belongs, and how much it is returned to. In energy suppliers, Octopus Energy leads attraction at 109 against Bulb at 65, indexed on the category at 100.

Attraction
Octopus Energy109
Bulb65
44-point spread · tracks being named 0.63
Attachment
Octopus Energy97
Bulb55
42-point spread · tracks being named 0.77
Affinity
Octopus Energy102
Bulb64
38-point spread · tracks being named 0.80
Octopus Energy: Attraction 109, Affinity 102, Attachment 97
Octopus Energy
Magnetic but unbought
OVO Energy: Attraction 91, Affinity 91, Attachment 87
OVO Energy
Fading
British Gas: Attraction 84, Affinity 91, Attachment 90
British Gas
Fading
AttractionAffinityAttachment

Each whorl lights one ring per index threshold cleared. See the full energy suppliers desire edition →

What AI searched for first

Before it answers, it goes looking for Green Energy.

An assistant does not answer from memory alone. It runs its own web searches first, and whoever it searches for is the shortlist it then chooses from. Across 79 searches in energy suppliers, Green Energy is named in 100% of the ones that name a brand at all. ChatGPT Web does not report its searches, so these shares are measured against the ones that do.

Green Energy 100%
Claude
Searches per question1.4
Names a brand0%
Names a year38%
Goes to a site0%
Gemini
Searches per question1.8
Names a brand3%
Names a year14%
Goes to a site0%
ChatGPT
Searches per question1
Names a brand5%
Names a year37%
Goes to a site79%
site:ofgem.gov.uk Renewable Energy Guarantees of Origin REGO electricity suppliers 100% renewable UK
ChatGPT
Ofgem energy price cap standing charge September 2026 UK why standing charges rising
ChatGPT
energy supplier financial hardship support UK how to apply OFGEM citizens advice
Gemini
UK prepayment energy key top up online traditional meter official supplier
ChatGPT

These are the assistants' own searches, read from their answers — not what anyone typed into a search engine.

Who AI trusts here

The referees of energy suppliers.

The independent sources AI models cite when answering energy suppliers questions — the places a brand needs to be seen, reviewed and ranked.

ofgem.gov.uk 60 cites
citizensadvice.org.uk 23 cites
energyplus.co.uk 13 cites
uswitch.com 9 cites
moneysavingexpert.com 8 cites
gov.uk 7 cites
energysavingtrust.org.uk 6 cites
cse.org.uk 5 cites
smartenergygb.org 5 cites
ecoflow.com 3 cites
Read is not cited

EDF Energy converts reads into citations. E.ON Next doesn't.

When Claude or ChatGPT opens a brand's own pages while answering, does the page make the cited-sources list? EDF Energy converts 75% of reads; E.ON Next just 25%. That gap is a content problem, not a visibility problem.

EDF Energy · read 4×75%
Fuse Energy · read 5×40%
British Gas · read 8×38%
E.ON Next · read 4×25%
Full data table
#BrandVisibilityAvg. positionCitation shareRead→cited
1Octopus Energy18%2.510%
2British Gas13%1.120%38%
3EDF Energy8%2.214%75%
4E.ON Next8%3.214%25%
5OVO Energy7%4.010%0%
6Green Energy5%1.70%
7Good Energy3%1.06%0%
8Ecotricity3%2.00%
9ScottishPower3%4.08%0%
10Utility Warehouse3%6.04%
11Boost2%2.00%
12Shell Energy2%4.00%
13Fuse Energy0%10%40%
14So Energy0%2%
15Utilita0%2%
Which AI you ask

Three models. Three different answers.

Ask about energy suppliers and the brand you are shown depends on the assistant: Claude → British Gas, Gemini → Octopus Energy, ChatGPT → E.ON Next. Their top-five lists share just 25% of names.

Claude recommends
British Gas
named in 15% of its answers
Gemini recommends
Octopus Energy
named in 35% of its answers
ChatGPT recommends
E.ON Next
named in 5% of its answers

Brands whose visibility swings most between models — the bar per model, widest gap on the right.

Octopus Energy 30pp
E.ON Next 20pp
British Gas 15pp
OVO Energy 15pp
ScottishPower 10pp
Green Energy 10pp
Utility Warehouse 10pp
ClaudeGeminiChatGPT
The data behind this chart
BrandClaudeGeminiChatGPTSpread
Octopus Energy15%35%5%30pp
E.ON Next0%20%5%20pp
British Gas15%20%5%15pp
OVO Energy0%15%5%15pp
ScottishPower0%10%0%10pp
Green Energy0%10%5%10pp
Utility Warehouse0%10%0%10pp
EDF Energy10%10%5%5pp
Good Energy5%5%0%5pp
Ecotricity5%5%0%5pp
What AI reads here

Editorial carry energy suppliers.

Every page AI opened while answering, classified by what kind of page it is. Editorial supply the most citations; Product pages convert best — 91% of the ones models opened were actually cited.

Editorial
150 cited · 37 read only
80% of reads cited
How-to guides
149 cited · 17 read only
90% of reads cited
Comparisons
38 cited · 9 read only
81% of reads cited
Other
20 cited · 16 read only
56% of reads cited
faq
18 cited · 2 read only
90% of reads cited
Documentation
16 cited · 22 read only
42% of reads cited
News
10 cited · 5 read only
67% of reads cited
Product pages
10 cited · 1 read only
91% of reads cited
The data behind this chart
Page typeCitedRead but not citedConversion
Editorial1503780%
How-to guides1491790%
Comparisons38981%
Other201656%
faq18290%
Documentation162242%
News10567%
Product pages10191%
What gets cited

Answer up front is what separates cited from ignored.

Of the energy suppliers pages models actually opened, how much more often each feature's pages ended up in the cited sources. Answer up front is worth 10 percentage points; cites its sources pages fare 46pp worse. Measured within the read population, so selection can't flatter it.

Answer up front +10pp
Structured data +4pp
Named author +2pp
Comparison table -7pp
Updated this year -15pp
Hard statistics -21pp
Cites its sources -46pp
The data behind this chart
FeaturePages withCited %Pages withoutCited %Difference
Answer up front33982%24572%+10pp
Structured data17481%41077%+4pp
Named author26979%31577%+2pp
Comparison table13673%44879%-7pp
Updated this year25470%33084%-15pp
Hard statistics21565%36986%-21pp
Cites its sources4035%54481%-46pp
Citation readiness

EDF Energy builds pages AI can quote.

Readiness scores each brand's own pages on the structure models reward — a direct answer up front, tables, hard numbers, named authors, structured data, recency, specifications. Outfox the Market is the widest gap: well-built pages that still aren't converting into citations.

BrandReadinessPages citedPages profiled
EDF Energy 38 26% 35
Fuse Energy 29 47% 17
Octopus Energy 24 18% 68
Good Energy 24 40% 15
OVO Energy 24 20% 35
So Energy 24 20% 10
E.ON Next 21 41% 34
British Gas 21 34% 35
Outfox the Market 20 0% 5
Ecotricity 20 0% 5

Readiness is the share of eight measurable structural signals present on a brand's profiled pages. It is a build-quality measure, not a ranking of the brand.

What the answers answer

The questions energy suppliers pages have to settle.

Read straight off the pages AI cited: the consumer questions cited content answers most often. Content that fails to settle these is content models have no reason to quote.

What is a prepayment meter?
What is the energy price cap?
What is a smart meter?
What is a standing charge?
Why are standing charges so high?
What are time-of-use tariffs?
How do smart meters work?
What is a green energy tariff?
Keep exploring

Nineteen more categories.

Method: 20 unbranded consumer prompts per category, spread across the buying journey, run weekly against ChatGPT, Claude and Gemini. Read→cited uses Claude and ChatGPT only (Gemini does not separate the two). Correlational observations of model behaviour at the time of the run. Data from the run completed Sun, 13 Sep 2026 06:39:59 GMT.

scroll ↓