This page explains who produces the research on Marketpedia, how a broker score is actually calculated, and what we do to keep commercial relationships away from editorial judgement. It is deliberately specific: weights are published as numbers, the testing method is described step by step, and the things we refuse to do are stated plainly rather than implied.
The scoring section sets out the four fixed pillars applied to every firm without exception — cost at thirty per cent, execution at twenty-five, safety of client money at twenty-five and platform experience at twenty. Those weights do not move between reviews, which is what makes two scores on this site comparable in the first place. Underneath each pillar sits a measured input rather than an impression: sampled spreads, timed fills, verified licences, completed withdrawals.
Further down you will find the people behind the work described by role and responsibility, our editorial policy on corrections and re-testing, and a risk disclaimer that we would rather you read than skip. Research is completed before any commercial conversation happens, affiliate arrangements never change a ranking, and no broker sees a page before it is published.
If you think a number on this site is wrong, tell us. Corrections are published with the date and the reason attached, because a research desk that quietly edits its own history is not a research desk.