Editorial standards
Our job is to help a specific buyer make a better decision—not to turn a feature list into a recommendation. This page explains the evidence we use, how we label limitations and how affiliate relationships are kept separate from editorial judgment.
Our review process
We identify the buyer, job, budget, alternatives and failure conditions before researching the product.
We use current official documentation, paid or free product access where available, dated screenshots, named inputs and credible independent sources.
Every conclusion should state who benefits, who should skip and what the evidence cannot prove.
Evidence labels
| Evidence type | What it means | What it does not mean |
|---|---|---|
| Hands-on access | We accessed the interface or workflow and can document the input and visible output. | It does not automatically prove long-term reliability or business ROI. |
| Official research | A feature, limit, price or policy was checked against a current first-party source. | Vendor documentation is not independent proof of performance. |
| Independent evidence | A credible third party adds context such as user experience, methodology or market data. | Anecdotes are not universal results. |
| Editorial inference | We interpret the evidence for a defined buyer or workflow. | It is judgment, not a guaranteed outcome. |
How software reviews differ from physical-product guides
For software, we examine workflow coverage, learning curve, plan restrictions, data interpretation, integration requirements and ongoing cost. A screenshot confirms that a feature or result was visible at a given time; it does not confirm the accuracy of every estimate behind it.
For physical products, useful evidence includes official specifications, form factor, compatibility, warranty, current availability and clearly licensed product media. We do not claim ownership, a purchase or a long-term field test unless that happened and can be supported.
Scoring and recommendations
We avoid universal scores that hide buyer differences. A product can be excellent for a growing private-label seller and poor value for a beginner with one simple listing. Recommendations are therefore tied to a use case and include the strongest reason to buy, the most important limitation and at least one realistic alternative.
Pricing, promotions and affiliate links
Pricing and discount claims are dated when timing matters. Coupon pages should show the code, duration, eligibility language and checkout evidence where available. We may earn a commission from qualifying links, at no extra cost to the reader. Commission availability does not guarantee inclusion, ranking or positive coverage.
Commercial pages include a visible disclosure. See our Affiliate Disclosure for program-specific details.
Updates and corrections
High-intent pages are revisited when vendors change prices, plans, interfaces or material features. Older pages may be rewritten, merged, redirected or removed from search if they no longer meet the standard. To report an error, send the page URL, disputed statement and supporting source through our contact page.
Detailed workflow for a commercial guide
1. Write the buyer brief
Before research begins, we state the intended reader, the task, likely budget constraints and what would make the product a poor fit. This prevents a review from becoming a summary of whatever the vendor emphasizes.
2. Create an evidence plan
We list the claims that affect the decision—pricing, plan limits, compatibility, workflow coverage, data source, physical dimensions or subscription requirements—and decide what evidence can support each one. A product screenshot supports a different claim from an official specification or an independent user report.
3. Capture reproducible inputs and outputs
When account access exists, useful screenshots include the input, selected marketplace, filters, date and visible result. A cropped interface image without those details may look persuasive but cannot be reproduced. Sensitive account information is removed or excluded.
4. Challenge the preferred conclusion
We look for limitations that could reverse the recommendation: a missing workflow, an expensive plan requirement, variation-data problems, ongoing subscription costs, portability issues or a strong lower-cost substitute. A review is incomplete until the main reason to skip is clear.
5. Edit for decision usefulness
The final page should answer the question near the top, include a comparison or decision table where useful, explain evidence and limitations, link to related guides and carry a visible disclosure when commercial links appear.
How we handle AI-assisted editorial work
AI tools may assist with outlining, language editing, structured comparison and repetitive production tasks. They are not treated as a factual source. Material claims require verification, and generated text must be reviewed for fabricated features, dates, prices, statistics and experience claims. Pages containing obvious placeholders, generator metadata or unsupported assertions are removed from public view until they can be rebuilt.
Image and screenshot standards
- Official or affiliate-program assets are identified as promotional material and used according to available program permissions.
- Hands-on screenshots should document a real interface state and avoid exposing account or customer information.
- Original diagrams explain a workflow but are not labeled as product screenshots.
- Images receive descriptive alt text and responsive sizing; decorative text is not a substitute for the written explanation.
- A screenshot is dated evidence, not a permanent statement about a changing product.
Update priority
We prioritize updates when a page has commercial intent, receives search visibility, contains time-sensitive pricing or documents software that changed materially. Minor wording changes receive a lower priority than a correction that could change a purchase decision. Pages outside the site’s current focus may remain unpublished rather than consuming crawl and editorial resources.