ComparEdge
Home/Methodology
Research StandardUpdated May 21, 2026

How ComparEdge researches, verifies, and scores SaaS data

Every data point is independently sourced, cross-referenced across multiple platforms, timestamped, and scored for confidence before publishing.

methodology.json
6
Data layers / product
5
Review platforms
30-day
Verification cadence
3+
Sources / price point
0.95
Confidence threshold
0
Sponsored rankings
01Collect02Cross-ref03Score04Timestamp05Publish
Freedom Score methodologyStack Compatibility scoringEvidence-Based Expert TakesPublic changelogsConfidence badges

Data Sources

Every price is multi-source verified. Single-source claims are flagged; multi-source confirmations raise the confidence score.

Official Pricing PagesConfidence: High

The vendor's public pricing page is the primary source. Captured on a rolling 30-day cadence so list prices, tier structures, and feature breakdowns stay current.

Official Product DocumentationConfidence: High

Feature claims, plan limits, and API capabilities are cross-referenced against the vendor's own published docs. Discrepancies are flagged before publishing.

Aggregated User ReviewsConfidence: Medium-High

Pricing-related signals from five verified-review platforms: G2, Capterra, TrustRadius, PeerSpot, and StackShare. Hidden fees, surprise renewal patterns, and billing satisfaction trends flagged by actual users inform our confidence scoring. Ratings are cross-referenced across platforms to detect outliers.

Buyer Community ReportsConfidence: Medium

Public procurement discussions, hidden-fee reports, and negotiation data shared in trusted online communities. Used to validate or flag outliers in official pricing.

Complaint and Dispute RecordsConfidence: Medium

Pricing disputes and billing complaints from public consumer-protection channels. A signal for surprise fees, auto-renewal traps, and post-sale cost escalation. Products with high complaint rates receive a reduced confidence score.

Manual Editorial VerificationConfidence: High

Products flagged for volatility or community disputes receive a manual review from our editorial team. Manually verified entries receive a confidence score of 0.95 vs 0.75 for auto-verified.

AI-Assisted Analytics LayerConfidence: Supporting

An AI-powered analytics system processes structured product data at scale: flagging pricing anomalies, detecting cross-category patterns, and surfacing outliers for human review. This system supports the editorial workflow but does not determine rankings, scores, or verdicts. All final assessments are produced and validated by a human analyst.

Confidence Scores

Not all pricing data is equal. A price confirmed by multiple sources is more reliable than one scraped from a stale page. Every product in our database carries a confidence score shown on its pricing section. This score reflects data quality only: how recently and thoroughly pricing data was verified. It does not affect the product's editorial ranking or composite score.

High
0.95Manually verified by our editorial team within 30 days against the vendor's official source. Multi-source confirmed.
Verified
0.75Auto-verified against the vendor's official pricing page within the current 30-day cycle. No independent disputes detected.
Review needed
0.65Verification cycle overdue or community disputes flagged. Data may be accurate but should be confirmed before a purchase decision.
Estimated
0.50Based on limited or outdated information. Confirm with the vendor before relying on this price.

Confidence scores decay 0.1 per 30 days without re-verification. Manual verification resets to 0.95. Scores are visible on every product pricing section and factored into our overall product rankings.

30-Day Verification Cadence

Every product is assigned to a verification group checked on a rolling daily schedule. At each cycle, our pipeline visits the vendor's pricing page and compares current data against stored records. Changes surface within hours, not weeks.

Last updated
Changes only when pricing, plans, or features actually changed.
Last verified
Date our pipeline last checked this product. Updates every cycle regardless of outcome.
Revision ID
Unique identifier per cycle (e.g. CE-LLM-2026W20-74C). Encodes category, year, week, and a content hash.
We distinguish verification (checking) from updating (changing). Products that pass with no changes show "No changes detected."

True Cost Framework

List price is the starting point, not the total. B2B SaaS pricing regularly omits costs that appear after you sign. Our pricing pages include a "Beyond the subscription" breakdown covering the full cost of ownership.

License Fees
Base subscription at the right tier for your use case
Implementation
Professional services, data migration, initial setup
Training
User onboarding, certification, admin enablement
Premium Support
SLA tiers, dedicated CSM, 24/7 access upgrades
Required Add-ons
Features marketed separately but needed for core use
Integrations
Connectors, API usage overages, custom development

The "Beyond the subscription" block appears on every pricing detail page. Always request a total cost of ownership estimate before signing a multi-year contract.

Scoring Criteria

Each product receives a composite score from four weighted criteria. This score reflects editorial product quality and determines category rankings. It is independent of the Confidence Score, which measures data freshness only.

Features (30%)

Verified against official documentation. We assess breadth (use cases covered), depth (implementation quality), and accessibility (features available across all plans, not just enterprise).

Pricing (25%)

Verified from the vendor's official pricing page. We evaluate value-for-money within the competitive set: what you get per dollar. Hidden fees, per-seat scaling, and feature gating are factored in.

User Ratings (25%)

Aggregated from verified public review platforms. Products with more verified reviews carry higher confidence in our analysis. Vendor-hosted testimonials and sponsored reviews are excluded.

Ease of Use (20%)

Derived from user-reported setup and usability ratings on the same review platforms, supplemented by onboarding quality and documentation coverage where data is available.

Public Changelog

Every pricing change, plan addition or removal, and feature update is logged in a public, timestamped changelog. You can see exactly what changed, when, and what the previous value was.

pricing_change

Price of an existing plan increased or decreased

plan_added

A new pricing tier was introduced

plan_removed

An existing plan was discontinued

feature_added

A significant feature was added

verified

Checked manually: no changes found

description_update

Product description updated

Machine-readable at /api/changelog. Queryable by product slug, entry type, and result limit.

Expert Analysis

Each product carries a distinct editorial assessment authored by Oleh Kem, Founder & Editor. The analysis draws from verified pricing data, user-reported friction points, feature sets, and comparative positioning within the competitive category.

Structured from product data: Each assessment is built from verified data: documented use cases, user-reported drawbacks, pricing tier, and competitive positioning. The output reflects what the data shows, not vendor marketing.
Category-aware: Assessments account for how the product sits within its competitive set. A $20/mo CRM is evaluated differently than a $20/mo analytics tool.
Author attribution: All expert analysis carries a linked author profile for Oleh Kem. JSON-LD schema on every page identifies the author as a Person entity.
Revision-tied: When pricing or key features change, the assessment context is revisited as part of the 30-day verification cycle.

Alternatives Intelligence

Each product's alternatives page is a structured intelligence report, not a generic list. Every section is populated from verified data.

1
Overview:Rating source (G2 priority), pricing tier, free plan status, alt-type classification.
2
Benchmark Snapshot:Category-specific performance data for 319 products (those with available public benchmark data). LLM: Arena ELO, HumanEval, MMLU. Security: MTTD, agent CPU %. Payments: auth rate, dunning recovery. Crypto: execution latency, fees. Each category has a unique metric set. Visible when 2+ alternatives have benchmark data.
3
When to Stay:Data-driven analysis of when the original product remains the stronger choice, based on its verified strengths and documented user feedback.
4
Per-Alternative Expert Evaluation:Each alternative receives a 2–3 sentence editorial assessment covering differentiation, pricing delta, and documented user friction points.
5
By Use Case:Workflow-segmented buckets (e.g. "For Coding Agents", "Open Source / Self-hosted") filtered from product feature tags, benchmark data, and pricing tier.
6
Top Picks Leaderboard:Leading alternatives ranked by G2/Capterra rating, pricing tier, and verified review platform coverage.
7
Feature Comparison Table:Semantic HTML table with category-specific columns. LLM columns: API Access, Open Source, Multimodal, Fine-tuning.
8
Product-Specific FAQ:Questions cover pricing trade-offs, migration paths, and category-specific buyer concerns. Supplemented by topics sourced from verified buyer discussions on G2, TrustRadius, and PeerSpot.
9
Sources & Data Trail:Verified review URLs for G2, Capterra, TrustRadius, PeerSpot, StackShare. Official pricing URL. Methodology attribution. Last verified date.
10
Editorial Block:Expert analysis by Oleh Kem with AuthorTooltip biography card and JSON-LD Person schema attribution.

Performance Benchmark Data

ComparEdge collects independently verified benchmark data for 319 of 495+ products (those with publicly available benchmark data across all 44 categories). Each category uses a unique metric set sourced from public vendor documentation, independent test labs, and community benchmarks. Data verified May 21, 2026.

AI / LLMArena ELO, HumanEval pass@1, MMLU accuracy, context window, TTFT latency, cost per 1M tokens
AI CodingSWE-bench Verified (%), HumanEval pass@1, autocomplete latency (ms)
Cloud HostingUptime SLA (%), edge nodes, free bandwidth (GB/mo)
Security (EDR/CSPM)Agent CPU %, MTTD/MTTR (min), agentless coverage (%)
PaymentsAuthorization rate (%), dunning recovery (%), checkout latency (ms)
CryptoExecution latency (ms), maker/taker fees (%), MEV protection, supported chains
Password ManagersKDF iterations, autofill accuracy (%), open-source flag
37 more categoriesVPN speed retention, SEO DB size, analytics event limits, HR country coverage, FinOps savings %, and more

Benchmark Snapshot is visible on alternatives pages when 2 or more products in the same category have benchmark data.

Stack Compatibility

The Stack Compatibility widget on alternatives pages estimates how well a product integrates with tools already in your stack. The widget is interactive: select tools you use, and it returns a fit score (0–100%) split into native integrations and middleware-required pairs.

How the score is computed
1
Category chip options
The widget shows integration categories derived from the product's own data: not a generic category-level list. A voice AI tool shows Video Conferencing, CRM, AI Video; a payments tool shows CRM, Accounting, Website Builders.
2
User selection
User selects categories from their existing stack. Each selection is matched against verified integration data for that product.
3
Fit score
Score = (native integrations selected / total selected) × 100. Native = confirmed in product data. Remainder = "May need middleware" (Zapier, Make, custom API).
Integration data sourcing

Integration compatibility data is derived from product descriptions, feature documentation, and vendor integration pages. We identify named tools each product connects with (e.g. HubSpot, Stripe, Zoom, Datadog) and classify them within our 45-category taxonomy.

For products with limited public integration documentation, compatibility is estimated based on category-level pairing patterns established across the full catalogue. The confidence badge indicates the evidence quality for each product.

Confidence badge
High confidenceIntegration compatibility confirmed from vendor documentation and official integration directories.
Medium confidenceSupported by partial vendor documentation. Some compatibility assessments draw on established patterns for this software category.
EstimatedBased on established integration patterns for this software category. Treat as indicative; verify with the vendor before committing.
Integration data is sourced from vendor documentation, official integration directories, and verified by the ComparEdge editorial team. The widget surfaces native connectors where confirmed; pairs without a confirmed native integration are flagged as potentially requiring middleware.

Vendor Lock-In Score

The Freedom Score (0–100) quantifies vendor exit risk across 5 dimensions, developed from data analysis across 495+ SaaS products. Formula: Score = 100 − Σ risk points. Higher score = more portable, lower switching cost.

5 Risk Dimensions
0–30
Data Portability
API presence + bulk export formats detected from vendor feature data. Both → 0 pts. API only → 10 pts. Export only → 18 pts. Neither → 30 pts.
0–25
Contract & Pricing
Free tier → 0 pts. Multiple paid plans, no free tier → 6 pts. Annual-only billing → 15 pts. Enterprise-only (custom pricing) → 25 pts.
0–20
Technical Standards & Ecosystem
SSO/SAML availability tier assessed from verified pricing plan documentation: business-plan SSO = accessible (reduces lock-in), enterprise-only SSO = not accessible (full penalty). Integration surface assessed from vendor feature documentation. Rich integrations + accessible SSO → 0 pts. No SSO + minimal integrations → 20 pts.
0–15
Market Alternatives
Live count of catalogued alternatives in the same category at query time. 20+ → 2 pts. 10–19 → 6 pts. 5–9 → 10 pts. 2–4 → 13 pts. 0–1 → 15 pts.
0–10
Migration Complexity
Weight-based detection in feature and description data: formula engine (3 pts), data mesh / cell-linking (3 pts), automations (2 pts), custom workflows (2 pts), custom fields (1 pt). Capped at 10. Widget note names each detected signal by product-specific feature name.
Confidence Badge System

The confidence badge is specific to Lock-In scoring: it reflects whether the data needed to assess SSO availability, integration depth, and migration complexity is fully verified for this product. It is independent from pricing verification status.

High confidenceProduct manually verified with complete pricing plan documentation and full feature coverage. All five dimensions scored from verified data.
Medium confidenceProduct verified with sufficient data for most dimensions. Technical standards and migration complexity scores are directional.
Low confidenceLimited structured data available. Treat the score as indicative and verify technical specifics with the vendor.
Freshness date: reflects the most recent verified update to any product data: pricing, features, or documentation: whichever was revised most recently.
Methodology developed from analysis of 495+ vendor pricing pages, feature matrices, and migration documentation. Updated as new products and categories are added to the catalogue.

Editorial Independence

ComparEdge is editorially independent. Scores and rankings are determined solely by our evaluation criteria, not by commercial relationships, affiliate fees, or vendor requests.

No paid placements or sponsored rankings
No vendor-supplied content used in scoring
Affiliate links (where present) do not influence rankings or confidence scores
Vendor corrections are reviewed on factual grounds only and do not affect editorial rankings

Corrections and Updates

Pricing changes fast. If you have found outdated or incorrect information, we want to know. Every report is logged, reviewed, and resolved within 48 hours.

Found an error?
Use our contact form to submit a correction. We investigate every report.
Report an inaccuracy
Are you the vendor?
Contact us to flag outdated data through our verified editorial channel.
Submit a vendor correction