First convert the data score to a percentage. Add its 60% contribution to the editorial percentage’s 40% contribution. A simulated 50% editorial input contributes 20 percentage points. Real assessments use the published criterion weights. Displayed inputs are rounded; contributions use the unrounded data score.
IG
Data component: 4.9/5
Editorial input: 76.5% (editorial review)
(4.9 ÷ 5 × 60) + (76.5 ÷ 100 × 40)
≈ 58.4 + 30.6
= 89.0% = 4.5/5
Assessment scope: traders:gb
Scope: UK resident offers; desktop public-information checklist uk-documentary-1.0, collected with AI assistance on 12 September 2026. No human reviewer, account opening, execution testing or mobile certification is claimed. A 100 checklist score means the named disclosure checks passed, not a perfect broker.
Brand signal: Trustpilot 3.7/5 -> (3.7-1)/4*100 = 67.5. Mixed products/countries; not UK-only feedback. Source: https://www.trustpilot.com/review/ig.com
Navigation: 3/4 checks. Cost headings, legal and support routes identified. Product choices were exposed as list items rather than semantic tabs in the inspected accessibility tree, so the semantic-controls check did not pass. This is a narrow accessibility observation, not a claim that switching is impossible.
Clarity: 3/4 checks. Product basis, fee categories and conditions were distinguishable. The minimum-spread table labels its values bps, creating a unit ambiguity for a reader comparing pip-based FX quotes; the unambiguous-units check did not pass. Source: https://www.ig.com/uk/charges
Documents: 4/4 discoverability checks. Local companies, customer agreements, complaints and execution-policy routes identified. Sources: https://www.ig.com/uk/terms-and-agreements https://www.ig.com/uk/customer-agreement
Limits: cookies can obstruct first access; no support contact was made. Document contents were sampled for identity and scope, not legally audited.
Editorial evidence ↗Reviews 67.5% · Navigation 75% · Clarity 75% · Documents 100%
Explore the data component →
OANDA
Data component: 2.9/5
Editorial input: 88.8% (editorial review)
(2.9 ÷ 5 × 60) + (88.75 ÷ 100 × 40)
≈ 34.4 + 35.5
= 69.9% = 3.5/5
Assessment scope: traders:gb
Scope: UK resident offers; desktop public-information checklist uk-documentary-1.0, collected with AI assistance on 12 September 2026. No human reviewer, account opening, execution testing or mobile certification is claimed. A 100 checklist score means the named disclosure checks passed, not a perfect broker.
Brand signal: Trustpilot 3.5/5 -> (3.5-1)/4*100 = 62.5. The global profile has country subprofiles; this audit consistently uses the global profile for all three brands. It is not UK-specific sentiment. Source: https://www.trustpilot.com/review/www.oanda.com
Navigation: 4/4 checks. Descriptive fee headings, legal and support routes, and a directly structured cost page without product-tab dependency were observed.
Clarity: 4/4 checks. The inspected page distinguishes fee types, sets out inactivity conditions and conversion basis, identifies v20/v20 MT4 applicability and links financing/spread information. This does not mean the fees are low. Source: https://www.oanda.com/uk-en/trading/our-charges/
Documents: 4/4 discoverability checks. OANDA Europe Limited is identified; contract, complaints and execution-policy links are listed. The sampled terms identify the company and CFD/spread-bet scope. Sources: https://www.oanda.com/uk-en/legal/ https://legal.oanda.com/oel/fxtrade_customer_agreement_oel/en
Limits: no account onboarding, support response or mobile journey was tested. The score measures this checklist, not full usability.
Editorial evidence ↗Reviews 62.5% · Navigation 100% · Clarity 100% · Documents 100%
Explore the data component →
Pepperstone
Data component: 4.3/5
Editorial input: 86.5% (editorial review)
(4.3 ÷ 5 × 60) + (86.5 ÷ 100 × 40)
≈ 52.0 + 34.6
= 86.6% = 4.3/5
Assessment scope: traders:gb
Scope: UK resident offers; desktop public-information checklist uk-documentary-1.0, collected with AI assistance on 12 September 2026. No human reviewer, account opening, execution testing or mobile certification is claimed. A 100 checklist score means the named disclosure checks passed, not a perfect broker.
Brand signal: Trustpilot 4.2/5 -> (4.2-1)/4*100 = 80. Global profile, not a UK customer sample. Source: https://www.trustpilot.com/review/pepperstone.com
Navigation: 4/4 checks. Cost headings, legal/support routes and semantic tab controls were exposed. A country prompt was encountered from the audit environment in Italy; this is reported as context, not attributed to all UK visitors.
Clarity: 3/4 checks. Standard/Razor/spread-betting pricing, additional charges and international-withdrawal exceptions were distinguishable. The cTrader explanation uses per unit while adjacent pricing uses lots; the unambiguous-unit-basis check did not pass. Source: https://pepperstone.com/en-gb/trading/costs-and-fees
Documents: 4/4 discoverability checks. Pepperstone Limited, UK contract, complaints and execution policy identified in the legal hub. A costs PDF was also retrieved. Sources: https://pepperstone.com/en-gb/about-us/policies/legal https://eu-assets.contentstack.com/v3/assets/bltd74737b907a31bbb/bltc19314db598b7ab0/UK_Legal_Costs_and_Charges.pdf
Limits: this browser check resolved the earlier full-page research-tool failure. No account onboarding, support response or mobile journey was tested.
Editorial evidence ↗Reviews 80% · Navigation 100% · Clarity 75% · Documents 100%
Explore the data component →
These criterion weights are managed in the editorial database. They apply to completed editorial assessments; simulated inputs remain explicitly labelled and are not derived from these criteria.
Published documentary audits have dated observations, cited sources and explicit scope. See the UK audit and checklist. Review-platform scores alone are not editorial assessments. Brand-wide review evidence must not be presented as jurisdiction-specific evidence, and evidence already scored in the data component must not be counted twice.