Research methodology
Evidence before inference.
How ExecGraph turns public industrial signals into accountable pursuit intelligence.
ExecGraph combines public records, operator disclosures, professional profile data, facility datasets, and analytical models. Every public research surface should identify the source class, evidence level, relevant date, author, reviewer, and the difference between a verified fact and an informed conclusion.
Current coverage snapshot
Coverage is measured. Completeness is not assumed.
Counts show the available production graph at the time of the snapshot. They do not convert every field into confirmed evidence.
Business records in the energy vertical. This does not mean every field is independently confirmed.
Records with a public professional profile URL available for source review.
Operators, contractors, suppliers, service firms, and related organizations in the graph.
Facilities with at least one EIA, OGIM, capacity, or imagery verification flag.
Turnaround events with both a source URL and a recorded verification timestamp.
Public accountability
The reader should know what supports the claim.
A polished page is not evidence. The source, date, classification, and review path are what make the conclusion usable.
ExecGraph separates the record itself from the analytical labels built around it. A verified title can sit beside an inferred facility assignment. A probable event can be commercially useful without being represented as a confirmed operator schedule.
When the evidence is missing, the public answer remains unavailable. The gap becomes research work rather than a value invented to complete the page.
Evidence classifications
Confidence is a label, not a promise.
The classification describes the strength of the available evidence. It does not guarantee a future event, commercial result, or customer outcome.
Confirmed
confirmedDirectly supported by a named primary source, current public record, or verified first-party record.
Probable
probableSupported by multiple consistent indicators, but not independently confirmed as a final fact.
Possible
possibleSupported by limited or early evidence and should not be treated as a committed outcome.
Inferred
inferredAn analytical conclusion derived from source-backed facts, industry patterns, or ExecGraph taxonomy.
Illustrative
illustrativeA product example or simulation. It is not a live customer result, measured outcome, or verified facility value.
Unavailable
unavailableThe available evidence is not sufficient to support a public conclusion.
Contact and role records
Keep the source record separate from the commercial interpretation.
Names, titles, companies, and public professional sources remain separate fields.
Facility assignment may be explicit, location-supported, title-supported, inferred, or awaiting review.
Role, seniority, and buying-center labels are analytical classifications and may differ from an employer's internal terminology.
A public profile source does not make every enriched field confirmed.
Facility and event records
Treat timing and scope as evidence questions, not calendar filler.
Facility identity and capacity may be supported by government, operator, imagery, or infrastructure datasets.
Turnaround records retain a source, confidence class, and verification date when evidence is available.
Probable and possible events are not presented as committed operator schedules.
Missing evidence remains unavailable rather than being replaced with a fabricated value.
Modeled commercial values
Estimate is not outcome.
Equipment populations, installed values, annual MRO values, turnaround scope values, opportunity values, payback periods, and recommendation examples remain estimates or illustrations unless a page identifies a verified first-party source. They are not customer-reported results, commitments, or guarantees.
Data freshness
Dates belong beside the evidence.
A current page can contain older source records. ExecGraph exposes the latest available update and verification dates rather than replacing record-level dates with the page request time.
Editorial review and corrections
See the evidence. Challenge the claim. Correct the record.
Automated systems may collect sources, classify claims, and propose revisions. High-risk publication remains human-gated. Material factual errors can be submitted publicly and are recorded when they change the meaning of a page.