How SoftPilot Researches Software

This page explains how SoftPilot researches software, what our labels mean, and how we decide what to recommend. It applies to every category guide, software profile and comparison we publish.

Our research process

1. Define the buyer’s question. Who is buying, for what job, at what team size and budget. Each category guide is built around those questions, not around a fixed list of features.
2. Check fit before research. We confirm a product belongs in the category by reading how the vendor positions it. Products that do a different job are moved to a better-fitting category, and the guide says why.
3. Read primary sources. Official pricing pages, plan comparison tables, help-centre documentation and product pages. Where a pricing page renders prices with JavaScript, we read it in a browser rather than relying on cached or third-party copies.
4. Record and date every fact. Each profile lists its sources with the date we accessed them, and shows when pricing was checked and when facts were last verified.
5. Compare on stated criteria. Comparison tables and “shortlist if” recommendations explain which criteria they rest on: price basis, free options, team controls, channels, limits and so on.
6. Review before publishing. Drafts are checked against the recorded sources, and unverifiable claims are removed or marked as not verified.

Which sources we trust

Source	How we use it
Vendor pricing pages and plan tables	Primary source for prices, plans, limits and trials
Vendor documentation and help centres	Primary source for how features, seats and credits work
Vendor product pages	Product positioning and feature names; marketing claims are attributed as the vendor’s
Third-party reviews and blogs	Background only. Never used as the source for a fact we publish, because they are often out of date or commercially motivated.

What our labels mean

• Desk research: the profile is based on official sources. We have not used the product ourselves.
• Hands-on tested: we have used the product, and the page explains what we tested.
• Not verified: the vendor does not publish the information, or it did not display during our check. We do not substitute estimates.
• Vendor-stated: a claim (such as accuracy rates, customer counts or compliance certifications) that we report as the vendor’s own statement.
• Promotional: a price that the vendor shows as a temporary discount. We show the regular price alongside it where the vendor displays one.

How we report prices

• We quote prices exactly as displayed, with the currency, billing period (monthly or annual) and date.
• Some vendors show prices for the visitor’s country. When we could only see regional prices, we say so and do not convert them.
• Where it helps comparison, we calculate a cost per contact, per verification or per seat from published figures, and show the calculation.
• We do not publish ratings, scores or “best overall” badges, and we do not mark up reviews or ratings in structured data.

Keeping pages current

Software pricing changes often. Every profile shows when its pricing was checked and when its facts were last verified. We re-check pages when we learn of a change and as part of a regular review cycle, and we correct errors under our corrections policy.

Use of AI

We use AI tools to help with research and drafting. AI output is never published as fact without being checked against the recorded sources. Where an AI tool we used is itself one of the products being compared, the page includes a disclosure.