Research and verification

Gulf Business Review Research Methodology

How Gulf Business Review verifies Gulf company, sovereign investment, AI infrastructure, energy and country-level research.

Scope. This page explains how Gulf Business Review handles permanent research, datasets, rankings and evidence-heavy reporting. It complements the publication's Editorial Standards, which govern corrections, independence, bylines and assisted production.

Gulf Business Review covers six different national economies and avoids treating the GCC as a single statistical or legal market. Saudi Arabia, the UAE, Qatar, Kuwait, Oman and Bahrain have different fiscal structures, regulators, sovereign investors and project pipelines, so national evidence is used before regional aggregation.

The publication also separates announced AI and infrastructure ambitions from contracted, financed, under-construction and operating capacity. Large headline commitments are retained in the stage and unit in which they were disclosed.

Primary-source hierarchy

Sources are chosen according to the subject and the claim being made. The following source types normally receive priority, with the most specific underlying document preferred over a general summary.

  1. National statistical agencies, central banks, ministries, regulators and exchanges across the GCC
  2. Sovereign wealth fund, state-company and listed-company disclosures
  3. Operator announcements, planning material and official project documentation
  4. Cloud-provider and technology-partner infrastructure disclosures
  5. High-quality international or regional reporting where it verifies project status not otherwise public

Core research rules

Country and regional geography

GCC, country, emirate, city and free-zone figures are not interchangeable. Regional totals are used only when the source actually covers the region. A UAE figure is not presented as a Gulf figure without qualification.

Sovereign capital

Fund assets, announced commitments, portfolio-company investment and project financing are separate concepts. A sovereign-backed announcement is not treated as budgeted government expenditure unless the disclosure supports that classification.

AI and data-centre infrastructure

Cloud regions, physical data centres, GPU programmes and gigawatt-scale campuses describe different infrastructure. Announced potential, initial phase, under-construction capacity and operating capacity are kept separate. Missing MW is not estimated from investment value.

Energy and project delivery

Final investment decision, contract award, construction and commercial operation are different milestones. Project value, capacity and ownership stakes retain their disclosed units and dates.

Dataset and permanent-reference rules

Gulf AI & Cloud Infrastructure Tracker

Rows cover identifiable cloud regions, AI data-centre projects or sovereign-compute programmes with credible sourcing. Partnerships without a defined physical capacity remain in the tracker only when their status and lack of site-level MW are made explicit.

Country guides

Country pages use national statistical and regulatory sources first. Cross-border company groups are not automatically assigned to one country when the disclosure refers to the group as a whole.

Review dates, corrections and refreshes

Project and country reference pages are reviewed when official status, disclosed capacity, investment or operating dates change. Announcements remain dated to the original event, and later reviews distinguish what has actually progressed from what remains a target.

Where a material factual error is identified, the correction process follows the Editorial Standards. Research pages are not given a newer date simply to create an appearance of freshness.

Meridian Review Group Information Gain Standard

A publication-wide quality gate for research, rankings, explainers and evidence-led reporting. New pages should add verifiable information, structure or analysis that a reader could not obtain by merely restating the same source material.

Add a new fact, comparison or structure

Every authority page should contribute at least one material element beyond a generic summary: an original dataset, a ranked or normalised comparison, a historical series, a project-stage classification, a source reconciliation, a geographic breakdown, or a clearly reasoned synthesis across multiple sources.

Prefer primary evidence

Company filings, regulators, government datasets, official project disclosures, investor materials and other first-party records should establish factual claims wherever practical. Secondary reporting can add context, challenge claims or fill gaps, but should not replace an available primary source without a reason.

Preserve provenance

Numbers must retain their observation date, source and qualification. Announced investment is not treated as completed spending, planned capacity is not operating capacity, and group-wide figures are not silently assigned to one site, country or subsidiary.

Make definitions explicit

Rankings, trackers and databases must explain what is included, what is excluded and how ambiguous categories are handled. Terms such as largest, capacity, investment, employer, headquarters, analyst consensus and project stage should be operationally defined.

Show uncertainty instead of inventing precision

Missing values remain missing unless a defensible calculation is explicitly labelled. Estimates, ranges and inferred values must be distinguished from disclosed figures. Conflicting sources should be surfaced rather than averaged into false certainty.

Keep a time dimension

Materially changing subjects should retain review dates, historical observations or event histories so readers and machines can distinguish a current state from a previous disclosure.

Connect the evidence graph

High-value pages should link to the relevant authority hub, supporting analysis, persistent entities, methodology and primary-source evidence. New pages should strengthen an existing topic graph rather than exist as isolated search landings.

Do not publish thin variants

A new page should not exist solely because a keyword has a slightly different wording. Overlapping intent should be consolidated into the strongest canonical resource unless the new page has a genuinely different dataset, audience task, geography, time frame or analytical question.

Publication gate before a new authority URL is created

  1. What verifiable information or structure does this page add that is not already present elsewhere in the archive?
  2. Can the important quantitative claims be traced to a source and observation date?
  3. Are scope, definitions and exclusions clear enough for another analyst to reproduce the classification?
  4. Does the page distinguish disclosed facts from estimates, interpretation and commentary?
  5. Does the page belong to an existing authority cluster and link back into that cluster?
  6. Would merging this material into an existing canonical page produce a better resource than publishing a separate URL?

Default decision: if a proposed page cannot clear these questions, the preferred action is to strengthen, update or consolidate an existing canonical resource instead of publishing another thin URL.

Research layer

Research & Data

Permanent datasets, directories and reference material maintained under this methodology.

Signature cluster

Gulf AI Infrastructure, Sovereign Capital and Delivery

Gulf Business Review's signature cluster follows the region's AI buildout from national ambition to physical delivery. It connects sovereign-backed technology companies, investment institutions, cloud regions and data-centre projects with the power, financing and infrastructure required to operate them. Country pages keep Saudi Arabia, the UAE and the other GCC markets distinct while the project tracker separates announcements from construction and operating capacity.

Gulf Business Review Research Methodology | Gulf Business Review