Review Methodology

How We Research and Verify Facts

We prioritize evidence in this order:

1. First-party product documentation and official pricing pages
2. Official product pages and announcements
3. Our own hands-on testing and direct product interface evidence
4. Reputable independent sources and expert reviews
5. Community evidence (used cautiously, and never alone for material claims)

We rely on first-party sources especially for pricing, plan limits, and official feature availability — but a vendor’s own marketing claims about performance or quality are not treated as independent proof; those get verified or tested separately.

We don’t treat the following as sufficient evidence for important claims: unverified AI-generated claims, unsourced listicles, anonymous claims, outdated reviews, or vendor marketing presented as if it were independent evidence.

When sources conflict, we check publication dates, product versions, regional or plan-level differences, and — where possible — test directly rather than simply picking whichever source supports the conclusion we’d prefer.

How Hands-On Testing Works

Where we test a product directly, we record what plan or account type we tested, when we tested it, what tasks we performed, what we observed, and what limitations we ran into. We don’t rely on memory alone, and we don’t claim “we tested” or “in our experience” language unless genuine hands-on testing actually happened. If direct testing wasn’t possible for a given review, we say so rather than implying otherwise.

Our Review Types

Different kinds of pages serve different purposes, and we apply the appropriate approach to each:

How We Score Products (the 100-Point System)

Every product we score is evaluated against a weighted 100-point framework:

Each category is rated on a 0–10 scale (0 = absent/unusable, 5 = adequate, 10 = exceptional and genuinely difficult to earn), then converted to weighted points:

Weighted Category Score = (Raw Score ÷ 10) × Category Weight

The final score is the sum of every category’s weighted points.

Some product categories (for example, keyword research tools vs. technical SEO crawlers) may use adjusted weightings so the score fairly reflects what actually matters for that category — but the total always remains 100, the rationale is documented, and affiliate relationships never justify a weight adjustment.

Important: Missing information is never automatically scored as zero. If we couldn’t verify or test a particular dimension, we disclose that limitation rather than penalizing the product for our own access constraints — or mark that dimension not applicable where it genuinely doesn’t apply.

What a Rating Does and Doesn’t Mean

Final scores map to a rating band:

A high score does not automatically mean “best for everyone,” “best value,” “best for beginners,” or “best overall” — those labels require their own evidence-based justification tied to a specific use case. A product can score well for agencies and poorly for solo freelancers; that’s not inconsistency, it’s context.

We also track evidence confidence (High / Medium / Low) as something separate from the score itself. A product can score highly while our confidence in some of that evidence is only medium — for example, if hands-on access was limited. Where evidence is genuinely too thin to responsibly assign a precise number, we say so instead of publishing a score that implies more certainty than we have.

How We Handle Pricing

For every material pricing claim, we record the pricing date, currency, billing frequency, plan name, and major limits, sourced primarily from the official pricing page. We distinguish clearly between “starts at,” “custom pricing,” “promotional price,” and standard listed price — and we never present an annual-billing-equivalent monthly figure as if it were the standard monthly price. Regular and promotional pricing are always clearly separated.

How We Write Strengths and Weaknesses

Pros must be specific and verifiable — “supports X workflow, reducing Y manual step” rather than a vague “lots of features.” Cons must be honest and material — we don’t manufacture a fake weakness just to look balanced, and we don’t hide a genuine limitation because of a commercial relationship. A review doesn’t need an equal number of pros and cons; the goal is accuracy, not artificial symmetry.

How Comparisons and Alternatives Work

Comparisons: The products being compared are evaluated against identical criteria — we don’t select criteria to favor a preferred option. We only declare a clear “winner” when the intent is well-defined, the criteria are appropriate, and the evidence gap is genuinely meaningful; otherwise we frame it as “best for X” vs. “best for Y” rather than forcing a universal winner.

Alternatives: A product only qualifies as an “alternative” if it has real functional overlap with the original — not because it’s popular, has an affiliate program, or shares the same broad industry category.

Affiliate Relationships and How They’re Disclosed

Where SEORAF may earn a commission on a link, that relationship is disclosed clearly and in plain language, near the top of the page and again near the relevant call-to-action — never buried only in a footer. Full detail on how and where disclosures appear is covered in our Affiliate Disclosure page.

Some things we never do, regardless of commercial relationship:

• Fake urgency or manufactured scarcity
• False “lowest price” or “official deal” claims we haven’t verified
• Fabricated testimonials, user counts, or review scores
• Hidden affiliate relationships
• Paid ranking manipulation

A call-to-action like “Start Free Trial” or “Save X%” is only used when that trial or discount is currently verified to be active.

How We Use AI — and What We Never Let It Do

AI tools may assist SEORAF’s team with research organization, draft structuring, and content planning. AI is never permitted to invent product features, pricing, test results, screenshots, customer experiences, performance claims, or affiliate relationships. Every material factual claim must be traceable to actual evidence — if it can’t be, it isn’t presented as fact. Material claims are verified by a human against authoritative sources before publication.

How and When We Update Reviews

Pricing, plans, and features change — so our reviews aren’t static. We reverify content on a regular cadence and revisit a review sooner when something material changes: a pricing update, a major feature change or removal, a company acquisition, or a product shutdown. Reviews and pricing-heavy pages generally carry a “last reviewed” date so you can judge how current the information is.

Corrections

When we find something in a published review that’s outdated, unsupported, or wrong, we verify the correct information, update the content, recalculate the score if the correction affects it, and record the change. We don’t quietly leave inaccurate claims live because a page happened to perform well.

Corrections & Accuracy: info@seoraf.com
Please include the relevant page URL and a brief explanation of the issue so that it can be reviewed appropriately.

What You Should — and Shouldn’t — Assume From a SEORAF Review

You can assume:

• Our score and verdict were reached independently of whether we earn a commission on that product.
• Material claims were verified against a documented evidence hierarchy, not assumed.
• Genuine limitations are disclosed, not hidden.
• Where we say we tested something hands-on, we mean it.

You shouldn’t assume:

• A high score means the product is right for your specific situation — check the “who it’s for” guidance in the review itself.
• A numerical score, on its own, captures every nuance of a decision that also depends on your budget, team size, and workflow.
• SEORAF has covered every product in a category — absence from our site isn’t itself a negative signal about a product.

If you have a question about a specific review, or you believe something we’ve published is outdated or inaccurate, please contact us using the confirmed SEORAF editorial/corrections contact method.