Standard agency test scenario
Unless a page states otherwise, every product is evaluated for five active client campaigns and five team users with different responsibilities. The workload includes keyword research, content production, technical auditing, traditional rank tracking, AI-search monitoring, and client reporting, including at least one local and one national campaign when supported.
This reference case prevents a one-user entry plan from looking cheaper than the configuration an agency actually needs.
The 100-point scoring model
Multi-client agency operations
20 ptsWorkspaces, permissions, users, projects, collaboration, and repeatable delivery.
Traditional and AI-search visibility
15 ptsSearch surfaces, AI mentions and citations, local visibility, history, and reporting clarity.
Ability to act on findings
15 ptsResearch, briefs, optimization, auditing, prioritization, and execution.
Client reporting and delivery
15 ptsWhite-label controls, reports, dashboards, exports, and explanations.
Cost at agency scale
15 ptsRequired cost for the same five-client, five-user scenario—not the cheapest plan.
Data depth and usefulness
10 ptsCoverage, freshness, granularity, limits, and decision usefulness.
Integrations and portability
5 ptsFirst-party data, exports, APIs, automation, and portability.
Usability, onboarding, and support
5 ptsSetup, learning curve, documentation, training, and support experience.
We define 0–5 thresholds before scoring, retain raw observations separately, keep unrounded calculations, and display rounded totals. We do not change the weights after seeing which tool wins unless the change is documented and every product is rescored.
Evidence hierarchy
- Direct hands-on observation in the tested account.
- Exports, screenshots, or saved outputs from the tested workflow.
- Official documentation or pricing with an access date.
- Verified customer-review patterns with source and sample size.
- Credible third-party analysis used as context.
- Vendor marketing claims labeled clearly as claims.
Documentation can establish that a feature is listed. It cannot establish usability, data quality, connector reliability, or support quality. Those claims require observation.
Agency-scale pricing protocol
We record billing frequency, annual-discount assumptions, taxes, add-ons, seats, clients, projects, keywords, crawl volume, credits, and overages. Costs are normalized for the same workload and shown per active client where the source data allows it.
A missing price or limit remains unknown. It is never converted to zero because that would reward incomplete evidence.
Rankability ownership and bias controls
Rankability owns this publication and is evaluated using the same fields, evidence rules, and scoring process as every other product; specialist tools may still win their categories.