Equipment Insider //Heavy equipment intelligence Account

Methodology version 1.1

Trust is a
data structure.

Every useful conclusion starts with a source, survives review and keeps its sample size visible.

Source hierarchy

Fact, observation and opinion never share a label.

Level 1

Manufacturer evidence

Official product pages, specifications, brochures, manuals and first-party media establish identity and published configuration facts. They do not establish reliability or owner satisfaction.

Level 2

Market evidence

Dealer inventory, quote-backed prices, verified transactions, auction results and rental rates establish observed market conditions. Each observation keeps date, geography and configuration context.

Level 3

Field evidence

Owner forums, long-form operating videos, technician discussions, fleet reports and direct submissions may establish experience. Repeated hearsay and generic opinions do not count as owners.

Level 4

Editorial analysis

Equipment Insider conclusions combine eligible evidence under the rules below. Editorial analysis is always labeled and can change when the evidence changes.

Two separate public signals

A comment is not a star review.

Owner Score and Market Sentiment answer different questions and never masquerade as each other. A reviewed observation publishes immediately; confidence grows with the sample.

1–5Owner Score

Explicit ratings submitted by owners, operators, renters, mechanics, and fleet managers. We show review count, verified count, confidence, dimensions, and buy-again percentage.

0–100Market Sentiment

Qualified public owner/operator evidence analyzed by theme and polarity. We show signal count, source diversity, freshness, and confidence—not stars.

1Directional

A single reviewed observation publishes with a single-source warning.

2–4Limited sample

The number is useful, but individual reports can move it sharply.

5+Developing

Direct Owner Score reaches moderate confidence as independent ratings accumulate.

10+Broader signal

Market confidence also considers community and source-type diversity.

25+High confidence

Large, diverse samples remain the target—not a gate that hides early evidence.

Owner Score and rankings

Six dimensions. Bayesian-adjusted rankings.

01

Performance

Productivity and task execution within the intended class.

02

Reliability

Uptime, failures, repair recurrence, and operating-hour context.

03

Operator experience

Controls, visibility, comfort, and fatigue over real operating days.

04

Serviceability

Inspection, maintenance, repair access, and documentation.

05

Ownership value

Ownership costs relative to useful output and resale strength.

06

Dealer & parts support

Parts availability, dealer responsiveness, and field support.

Adjusted score = review weight × raw score + prior weight × peer-group mean.

A tiny perfect sample cannot outrank a mature excellent sample. No ranking publishes with fewer than five eligible comparable products. Peer groups use equipment type, objective size/capacity band, generation, power or hydraulic class, status, and relevant use case.

Every score and ranking stores the methodology version, inputs, evidence counts, confidence, and historical snapshot used at publication.

Pricing intelligence

The number manufacturers hide needs context.

Price types stay separate

MSRP, dealer asking price, quote-backed price, transaction price, used asking price, auction result, rental rate and editorial estimate are distinct evidence types.

Configuration comes first

Cab, high flow, tracks, buckets, attachments, freight, setup, rebates and warranties can materially change the number. Unknown inclusions lower confidence.

One price still counts

A single reviewed listing publishes as one dated observation. It is useful evidence, but not mislabeled as a national market average.

Ranges need support

A range publishes only when comparable observations support it. Every price keeps its source type, date, condition and known inclusions.

Media policy

Discovery is not permission.

Published product images must be first-party manufacturer media, licensed press assets, permitted embeds or owner-submitted media with tracked permission. The current Bobcat catalog uses Bobcat-hosted product media for editorial identification with attribution. Gallery assets remain unpublished until reviewed.

AI and editorial disclosure

Automation collects. Review publishes.

Software and AI may discover pages, normalize specifications, identify possible duplicate owners, extract candidate price context and surface issue themes. Machine-generated work enters a review state. It cannot publish a score, recurring issue or market-price conclusion by itself.

Corrections

The record is allowed to improve.

Source pages are content-hashed and observed over time. Material changes create a change event instead of silently rewriting history. To report a factual error, email editor@equipmentinsiderhq.com with the product URL, disputed field and supporting source. Requests are reviewed, not automatically accepted or published.