Trust centre · Methodology

A verdict you can inspect.

No single lab routine fits a phone, a cloud platform and an AI service. Our common framework keeps the evidence, criteria and limitations legible.

Effective: 17 July 2026Applies to: editorial reviews and comparisonsVersion: 1.0

Core principles

  • Test or research the decisions that matter to the intended user, not just the easiest specifications to measure.
  • State the basis of coverage and do not turn limited access into a full-review claim.
  • Compare like with like, using a declared audience, price context and review period.
  • Keep editorial scores distinct from community ratings and commercial popularity.
  • Record material limitations, conflicts, supplied access and changes after publication.
Demonstration content is not evidence.Sample ratings, seeded activity, fictional testimonials and prototype review cards are explicitly labelled. They must not be cited as a DopeTech test result or user experience.

The review workflow

1

Define

Identify the audience, use cases, competitors, test period, version and criteria before forming the verdict.

2

Research

Check specifications, pricing basis, documentation, known issues, vendor claims and relevant independent sources.

3

Evaluate

Run repeatable scenario-based checks where access permits, and record conditions that could affect results.

4

Challenge

Look for failure modes, long-term costs, accessibility gaps, privacy trade-offs and security caveats.

5

Conclude

Weight criteria for the stated audience, document evidence and separate facts from editorial judgement.

6

Monitor

Revisit material changes in price, software, service quality, safety, support or availability.

Physical product testing

Relevant checks may include setup, build and repairability, ergonomics, accessibility, performance, battery or energy use, connectivity, companion software, support, privacy and value. Review pages should identify the model, software version, test duration and material accessories where they affect the result.

Benchmarks are useful only with context. When measurements are published, the page should name the workload and meaningful conditions. Results from a vendor-controlled demonstration are labelled as a demonstration, not an independent benchmark.

Software and service evaluation

A service can change without a new box on a shelf. Relevant criteria may include onboarding, core workflows, integrations, reliability evidence, admin controls, accessibility, data handling, security documentation, support, contract constraints, exit paths and total cost.

Enterprise security or compliance claims are attributed to current vendor documentation or named independent reports. A DopeTech review is not a penetration test, legal opinion or certification. Pricing comparisons state the plan, billing period, currency and assumptions where practical.

How scoring works

The overall score is an editorial synthesis for the audience and price context stated on the page—not a mathematical claim of universal quality. Category weights can differ, but a review should explain the factors that moved the verdict.

RangeMeaningDecision signal
9.0–10Exceptional in its stated classA leading choice with limited, clearly stated compromises
8.0–8.9ExcellentStrong recommendation for the intended audience
7.0–7.9GoodWorth considering; trade-offs matter by use case
6.0–6.9MixedCapable, but notable limitations or weak value
Below 6.0Material concernsAlternatives are likely preferable for most intended users

Price and competition change. The same underlying product can reasonably receive a different recommendation at a materially different price or after major updates.

Evidence and confidence

Where useful, DopeTech may show a confidence label alongside a verdict. High confidence indicates direct, repeatable evidence across the main use cases. Moderate confidence indicates adequate evidence with named limitations. Preliminary means the assessment is based on short access, incomplete features or primarily attributable research.

Confidence is not the same as score. An excellent first look can still carry preliminary confidence.

Comparisons, awards and community data

“Best” lists define who the recommendation is for and when they were checked. Eligibility may consider availability, review evidence and meaningful category fit. Commercial relationships are disclosed and must not purchase ranking position.

Community ratings and votes are separate signals. They may be filtered or withheld when abuse, coordinated manipulation, conflicts or insufficient context make them unreliable. Editorial awards remain editorial decisions.

Common questions

Does every review involve hands-on testing?

No. The page should identify whether coverage is tested, a first look, research-based or a community signal. Research-based coverage must not imply hands-on use.

Can a company pay for a higher score?

No. Commercial arrangements must not buy a score or favourable verdict. Sponsored or commissioned material is labelled and separated from independent editorial conclusions.

Why can a review score change?

A score may change after material pricing, software, security or performance changes. Material updates should be dated and explained.

See something we missed?

Evidence-based methodology feedback is welcome at editorial@dopetech.au. For factual corrections, include the page URL, relevant version and a primary source where possible.