PilotStack
Our Process

How We Evaluate Software

Last updated: July 2026

Quick Answer

Every PilotStack page runs through five stages: selection, sourcing, scoring, consistency, and upkeep. Each tool carries nine recorded category scores on a 1-5 scale, and the overall rating on a page is their mean rounded to one decimal. Figures come from sources we hold; facts without a source are omitted or marked unverified. Commercial relationships are disclosed where applicable and are not intended to determine coverage.

TL;DR

  • Five stages: Selection, Evidence & Sourcing, Scoring, Consistency, Publication & Upkeep
  • Nine equally weighted category scores on a 1-5 scale; overall is their mean
  • One recorded source per figure, reused everywhere a tool appears
  • Unsourced facts are omitted or shown as unverified — never asserted
  • Commercial relationships do not determine ratings

Key Takeaways

  • Coverage is guided by reader usefulness and editorial priorities
  • Ratings are arithmetic you can check on any review page
  • Conflicting internal sources result in no published value, not a guess
  • Certification statuses without a source read as "not verified"
  • Last-reviewed date visible on every page
  • Corrections are made in place and dated

This page describes how PilotStack pages are produced and scored. It is the short version of our Review Methodology.

Stage 1:Selection & Scope

  • •Cover categories where readers are actively evaluating options
  • •Record the same fixed set of fields for every tool: pricing model, deployment, API availability, migration complexity, security entries, integrations, and category scores
  • •Document the editorial criteria used for inclusion and exclusion
  • •Do not sell inclusion or exclusion as an editorial outcome

Stage 2:Evidence & Sourcing

  • •Every figure on a page comes from a source we hold: our recorded dataset, vendor documentation, or a published pricing page
  • •Review counts, ratings and pricing are recorded as of the review date shown on the page
  • •Where two sources we hold disagree, neither value is published
  • •Certification and compliance statuses with no source are shown as unverified rather than asserted

Stage 3:Scoring

  • •Nine equally weighted category scores on a 1-5 scale: Features, Usability, Pricing, Support, Security, Integrations, Performance, Documentation, Scalability
  • •The overall rating is the mean of those nine scores, rounded to one decimal
  • •The calculation is published and checkable on any review page
  • •Ratings are based on the published scoring criteria rather than commercial placement

Stage 4:Consistency

  • •A tool's rating, review count and category scores are held once and reused everywhere that tool appears
  • •Category, comparison and related-content pages read from the same recorded values
  • •Facts we cannot source are omitted instead of being restated in a new form

Stage 5:Publication & Upkeep

  • •Every page carries a last-reviewed date
  • •Pages are updated when material changes are identified
  • •Corrections are made in place and the last-reviewed date is updated
  • •Pages with unsupported recorded figures are corrected or, where appropriate, removed

What We Do Not Do

  • We do not sell editorial ratings or placement
  • We do not allow vendors to preview or approve pages before publication
  • We do not describe access arrangements as evidence of editorial quality
  • We do not publish a figure we cannot tie to a source we hold
  • We do not present an unverified certification or company fact as confirmed

For the full scoring rubric and sourcing rules, see our Review Methodology page.