Home › How we score
How we score and rank tools
Every established tool on AgentAICompare gets five scores from 1 to 10, one per factor below. The overall score is the plain, unweighted average of the criteria we could score, rounded to one decimal (for example 7.4/10). Lists are ranked by that average, highest first, with ties broken by starting price; tools without enough independent reviews are ranked after reviewed tools (see below). We score 51 tools across 13 categories; latest scoring 2026-11-08.
The rubric
| Factor | Score | What we look at |
|---|---|---|
| Price value | 1–10 | What you get for the starting price, compared with other tools in the same category. Free plans and flat, unlimited plans score well. Quote-only pricing isn't scored (Not enough info). |
| Features for agents | 1–10 | How much of the job an agent actually needs the tool does well, and whether it is built for real estate agents rather than a general audience. |
| User feedback | 1–10 | Independent reviews and discussion we find ourselves: G2, Capterra, Trustpilot, app stores, Reddit and the BBB. We weigh the rating and the number of reviews, and recurring complaints about billing, refunds or support. When there are too few independent reviews to judge, this shows as Not enough info instead of a low score. We never invent ratings or review counts, and vendor testimonials don't count. |
| Terms & refund policy | 1–10 | Published terms of service, who owns what the tool creates, and a clear refund or cancellation policy. Long contracts, cancellation by phone only or early-cancellation fees lower this score. |
| Ease of sign-up / free trial | 1–10 | How easy it is to actually try the product. A free plan or free trial scores highest, a self-serve paid plan scores well, a demo plus a sales-led trial or sandbox is moderate, and only genuinely hard access (a long contract with no trial) scores low. |
| Overall | Average | The scored criteria added up and divided by how many there are, rounded to one decimal. Every factor counts equally; criteria marked Not enough info are left out. |
When there isn't enough information
We don't guess. If a criterion can't be judged fairly, it shows as Not enough info and is left out of the average instead of being scored low.
- Hidden pricing isn't counted against a tool. If a vendor only quotes prices or requires a demo, we can't judge price value, so we skip that criterion. We still note it: every tool page shows its pricing status (for example "Custom / contact sales").
- Thin reviews aren't counted against a tool's score. If we can't find enough independent reviews to judge, user feedback shows Not enough info and is left out of the average (it does affect rank; see below). Where real reviews exist, good, mixed or bad, we score them.
- Ease of sign-up means how easy it is to try the product. A free plan or free trial scores highest, a self-serve paid plan scores well, and a demo with a sales-led trial or sandbox scores in the middle. Only genuinely hard access, like a long contract with no trial, scores low. Hidden pricing is never penalized twice.
- Reviews still matter. If we can't find enough independent reviews to score user feedback, the tool keeps its score, but it is ranked below every scored tool in its category that does have a user feedback score, in every list on the site. Among themselves, these tools are ranked by score as usual. They carry a note: "Not enough independent reviews yet, ranked below reviewed tools." When enough reviews appear, we score them and the tool is ranked normally.
- A #1 has to earn it. A tool only gets the #1 spot (our top pick) in a category if it has independent reviews and scores at least 7.0. If no reviewed tool in a category reaches 7.0, that category has no top pick: we say "No top pick yet" and list its tools in score order without rank numbers. Scores themselves don't change.
- Adoption figures don't change scores. Some tool pages show an adoption badge (for example how many agents use a tool, or how many reviews it has on G2). It is for context only, always links to its source, and never affects a score or rank.
- At least 3 of 5. A tool needs at least three scored criteria to get an overall score and a rank. The overall is the average of the criteria we could score, and the score card says "Based on N of 5 criteria" when any are missing. Tools with fewer than three show "Not enough info to score yet" and are listed after the ranked tools.
Scores compare tools within their own category. A 7.5 in virtual staging and a 7.5 in CRM both mean "strong for its job", not that the two tools are interchangeable.
What never affects a score
Payment of any kind. Featured listings, sponsor tiers, fast-track review fees and affiliate commissions never change a score, a rank or a verdict. Whether a tool has an affiliate program plays no part in the rubric. Sponsored placements are always labeled. See our disclosure.
No fake star ratings
Our score is an editorial score, not a customer rating. We don't show it as stars, and we don't mark it up as a review rating for search engines. When we mention a G2, Capterra or Trustpilot rating, we give the source and the number of reviews behind it.
New tools
Tools that are too new for our full listing requirements appear in their categories with a New badge and no score, so promising launches aren't missed. Once a tool qualifies (for example, 6+ months live, published terms of service and a refund or cancellation policy), we research it and score it like every other tool. See New & Emerging and who qualifies.
Because virtual staging is a crowded category, new virtual staging tools are only considered after at least one year live, with legitimate independent reviews and proof of real customers. Existing staging listings are not affected.
How often we re-score
We re-score every tool at least every 90 days, and sooner when a vendor changes its pricing or terms or when new review evidence appears. Each tool page shows the date its score was set. Vendors who think a fact is wrong can contact us with a source; we correct facts, but we don't negotiate scores.