Editorial methodology

1. Structured data first

Every broker, platform and fee is held with four facts: the value, where it came from, the URL to check it, and the date we last verified it. Pages read from that one source — we never retype a fee into an article or a component.

2. Verification

• Fees are checked against official broker pricing pages (linked as sources where public).
• Each figure shows its status: verified with date, or “confirming” while re-verification is pending. Confirming figures are labelled in the UI — never presented as fact.
• Material pricing changes trigger a re-check of comparisons, calculators and verdicts that cite them — and are logged publicly on the pricing-changes page.

3. Editorial ratings (x/5)

Ratings weigh pricing transparency, platform reliability, support quality, and cost for typical users of that product. They are editorial judgement, not measurements — two honest researchers can disagree by 0.2, and that's fine. App-store ratings are shown separately and labelled as such.

4. How we test each broker

Every review follows the same hands-on checklist before a score is set:

• Fees (weight: highest). Every charge on the official pricing page is recorded with source URL and date — opening, AMC, delivery, intraday, F&O, DP. We also place test orders where possible to confirm what the tariff really means.
• Platforms. We use the actual web and mobile apps: order placement, charting depth, watchlists, back-office reports. Screenshots in reviews come from real sessions.
• Support. We contact support as an ordinary customer (chat + email) and time the response. One good reply doesn't make a 5.
• Research & education. We read the research output and learning material as a beginner would — jargon density counts against it.
• Onboarding. We walk the full e-KYC flow and note every friction point, document ask, and delay.

5. How scores are computed

Each broker carries five category sub-scores (0–5, one decimal): fees, platforms, support, research, onboarding. The headline x/5 is an editorial synthesis, not an average — fees dominate for discount brokers, so a broker with ₹0 delivery and weak research can still score 4.6. Sub-scores are shown beside the headline precisely so you can audit the judgement: if you value research over price, re-weight them yourself.

6. Update cadence

• Fees: re-verified against official pricing pages at least quarterly, and immediately when readers report a change or a broker announces one. A monthly automated check fetches every official pricing page and flags content changes for review — see the pricing-changes log.
• Platforms & features: re-checked on every major app release we cover.
• Comparisons & calculators: rebuilt from the same records, so a fee update propagates everywhere within minutes — no stale copies.
• Every figure shows its status: verified with date, or “confirming” while re-verification is pending.

7. Affiliate separation

• Commercial relationships live in a separate affiliate-link table — updating a URL never touches editorial content.
• Ratings, sort orders and verdicts are set without reference to commission levels. Featured placements are editorial.
• Every monetised page carries an affiliate disclosure.

8. What we won't do

• No fake reviews or invented ratings. No Review schema on content that isn't genuine editorial evaluation.
• No programmatic SEO spam: every comparison page must add unique, useful differentiation or it doesn't ship.
• No personalised investment recommendations. Educational content only.

Corrections

Found a stale figure? Tell us with the official source link — verified corrections ship within days, not quarters.

Authorship

Research is published under the shared LetsInvest Research Desk byline — an editorial team, not an individual. We use no invented pen names; a named byline appears only when a real contributor signs the piece.

Last reviewed: 22 September 2026.