Comparison Methodology Disclosure
By RankFast Editorial · Sources reviewed
Affiliate disclosure: We may earn a commission from Outrank purchases through our links, at no extra cost to you. Our recommendation is an editorial assessment of workflow fit, not an independent performance award.
Methodology and evidence limits: The initial desk review covered seven vendors’ public product and pricing pages on the date above. Six additional vendor comparisons were reviewed November 13, 2026, bringing the collection to twelve comparisons across thirteen vendors. Individual articles retain their own review dates. This is not a hands-on campaign experiment. Feature and price claims link to their vendor sources. Vendor descriptions are not independent verification; plans and features may change.
Performance dataset status: No reproducible campaign cohort, site list, article sample, measurement dates, or raw ranking/quality results are available to support our former benchmark claims. Verified performance sample size: 0. Campaign period: not available. There is no result dataset to download. We therefore do not report ranking windows, quality scores, detector pass rates, user counts, or combined G2/Capterra ratings.
Reproducible benchmark protocol — proposed, not executed
- Pre-register a cohort of 10 consenting English-language sites across SaaS, ecommerce, and services. Disclose domain age, baseline traffic, topic authority, selection rules, and exclusions. Use 10 distinct matched query briefs per tool per site: 100 articles per tool, 700 across seven tools. Randomize comparable, non-overlapping queries within each site to avoid duplicate-content competition.
- Before generation, publish the exact start/end dates, tool versions, plans, prompts, query sets, difficulty source, and review budget. Hold editorial effort, publishing cadence, internal linking, and promotion constant; record deviations and all rejected drafts.
- Observe every published page for 12 weeks from its publication date. Record indexing separately from ranking; measure days to a sustained top-10 position (seven consecutive daily observations) for the assigned query, with fixed country/device and rank-tracker settings. Include non-ranking pages as censored observations, not successes.
- Have two reviewers, blinded to tool, score factual accuracy, completeness, source support, and readability from 1–5 using a published rubric. Report disagreements, editing time, actual subscription/add-on/labor costs, medians, uncertainty, and results by site and tool. Detector output is not proof of quality or Google compliance.
- Publish anonymized per-article rows and daily observations: site ID, cohort, query, tool/version, prompt, generation/publication dates, index date, country/device, daily rank, reviewer scores, editing minutes, cost, exclusions, and missing-data reasons. Release the rubric and analysis script alongside results before making comparative performance claims.
Search policy context: Google’s AI-content guidance (February 8, 2023). This guidance is not evidence that Outrank or any competitor ranks faster.