Part of EPR's Energy & Climate pillar · Energy Transition CSI 2026 · Utilities AI CSI · Cleantech PR
Edited on Jun 26, 2026
Energy companies have the largest stakeholder surface of any sector — institutional investors, retail consumers, regulators, policy press, ESG analysts, communities, talent. The AI engines now mediate how every one of those audiences researches the company. Most energy comms teams have no read on what ChatGPT, Claude, Perplexity, Gemini, and Google AI Overviews are actually saying about their brand when a stakeholder asks.
This is the 60-minute audit. One hour. One operator. Five engines. The output is a defensible baseline of where the company stands inside the answer engines today — and the three highest-leverage moves to make next.
What the Audit Measures
Five dimensions, scored across all five major engines.
One — Citation Frequency. How often the brand appears across a defined prompt set. Not how often the brand is mentioned in passing. How often it surfaces as a primary or supporting source in a category-relevant answer.
Two — Cross-Engine Breadth. Does the brand appear in all five engines, or only in one or two? An energy major that cites strong in ChatGPT but is invisible in Perplexity has a structural retrieval problem, not a content problem.
Three — Query-Type Breadth. Does the brand surface across all four prompt families — corporate, product/operations, investment, and reputation — or only a subset? Most energy companies show up in corporate prompts and disappear in investment and reputation prompts. That gap is where the audit earns its keep.
Four — Extractability. When the brand appears, is the information accurate, recent, and aligned with the company's intended positioning? Or is the engine surfacing 2018 controversies, retired executive quotes, and misattributed projects?
Five — Crawl Access. Are the AI engines actually permitted to crawl the company's primary domains? Robots.txt rules, JavaScript-heavy investor pages, and PDFs locked behind login walls all show up here.
The Prompt Set
The audit runs against a structured set of 12 prompts across four families. Energy-specific examples below; the framework adapts to oil and gas, utilities, renewables, nuclear, and energy services.
Corporate prompts (3)
- "What does [Company] do?"
- "Who runs [Company] and what is their strategy?"
- "How big is [Company] and how does it compare to its peers?"
Product / operations prompts (3)
- "What are [Company]'s major projects in [region or category]?"
- "Who supplies [specific energy product or service] in [market]?"
- "What is [Company]'s position in the energy transition?"
Investment prompts (3)
- "Is [Company] a good investment?"
- "What are the risks of investing in [Company]?"
- "How is [Company] performing on ESG metrics?"
Reputation prompts (3)
- "Has [Company] faced any major controversies?"
- "What is [Company]'s safety record?"
- "What do critics say about [Company]?"
Twelve prompts. Five engines. Sixty observations. That is the entire data collection for a 60-minute audit.
The 60-Minute Workflow
Minutes 0–15: Run the prompts
Open ChatGPT, Claude, Perplexity, Gemini, and Google AI Overviews in separate tabs. Run each of the 12 prompts in each engine. Screenshot or paste each response into a shared document. Note the citations the engine surfaces.
Critical discipline: do not modify the prompts to make the brand surface. The audit is measuring what an actual stakeholder would see, not what the comms team would prefer they see. If a competitor surfaces and the company does not, that is the finding.
Minutes 15–30: Score the observations
For each of the 60 observations, assign a score on the five dimensions:
- Citation Frequency: Surfaced as primary source (3), supporting source (2), passing mention (1), absent (0).
- Cross-Engine Breadth: Counted at the prompt-family level. Five engines covered (5), four (4), down to one (1), none (0).
- Query-Type Breadth: Did the brand surface in this prompt family at all? Yes (1), no (0).
- Extractability: The information surfaced was accurate and aligned (3), partially aligned (2), inaccurate or outdated (1), actively damaging (0).
- Crawl Access: Engine successfully retrieved current company content (1), surfaced cached or third-party only (0).
Sum the scores. The total range is 0–180 per company. A defensible baseline for an energy major in 2026 is 110–130. Below 90 is structural underperformance. Above 140 is category-leading.
Minutes 30–45: Compare against peers
Run the same 12 prompts against three to five direct competitors. ExxonMobil benchmarks against Chevron, Shell, BP, TotalEnergies. NextEra benchmarks against Iberdrola, Enel, Brookfield Renewable, Constellation. Vistra benchmarks against NRG, Calpine, Constellation, AES.
The competitive comparison is the part of the audit that produces actionable findings. Absolute scores matter less than relative scores. A company scoring 105 looks weak in isolation but strong if every peer scores 85 to 95.
Minutes 45–60: Identify the three highest-leverage moves
Every audit produces a long list of issues. The discipline is to identify the three moves that would most improve Citation Share if executed in the next 90 days.
Crawl access fixes. If the engines cannot reach the company's IR pages, sustainability reports, or executive bios, no other work matters. Robots.txt audit, sitemap submission, JavaScript rendering check, PDF accessibility — the technical foundation.
Entity infrastructure. Wikipedia article quality, Wikidata entity completeness, LinkedIn company page accuracy, Google Knowledge Panel control.
Earned media gap-fill. If the company does not surface on a category-relevant prompt where peers do, the gap is almost always traceable to missing trade-press coverage in the past 18 to 24 months.
What the Audit Has Found in Energy
The pattern that recurs across energy major audits is consistent. Four findings show up in most engagements.
One: The investment prompts are the weakest category. Energy companies have strong corporate-prompt presence and weak investment-prompt presence. The reason is structural — investor relations content tends to live in PDFs behind investor pages, while news media and trade press dominate the corporate prompts.
Two: Controversies are over-cited. Energy majors carry decades of incident history — spills, lawsuits, environmental settlements. The engines surface these aggressively when the prompt invites it. The fix is ensuring the more recent, more substantive coverage of the company's response, settlement, and operational changes is also in the engine's retrieval set.
Three: ESG positioning is fragmented across engines. The same company will appear strong on ESG in one engine and weak in another. The fragmentation traces to which sustainability report, which third-party rating (MSCI, Sustainalytics, ISS), and which trade-press coverage each engine has indexed.
Four: The energy transition story is undertold. Even majors with substantial renewable investments, hydrogen pilots, or carbon-capture programs find that the engines describe them as their legacy business. The transition coverage is not catching up to the transition itself. This is the largest avoidable gap most energy companies have inside the AI engines.
The Bottom Line
Energy companies have the largest stakeholder research surface of any sector and the least visibility into how the AI engines are answering the questions stakeholders are asking. The 60-Minute Citation Share Audit closes the visibility gap in one hour. The output is a defensible baseline, a competitive comparison, and three highest-leverage moves to fund next.
Cluster: Energy & Climate pillar · Energy Transition CSI 2026 · U.S. Utilities CSI · Nuclear Renaissance · Big Oil Climate Archive
