Methodology

How we score

Score = 0.40×Quality + 0.25×Speed/time saved + 0.20×Price + 0.15×Ease of use
(each 0–10, total rounded half-up to one decimal)

40%Quality
Five task-specific checks, 2 points each: can the tool produce the thing the client needs?
25%Speed / time saved
Five task-specific checks, 2 points each: templates, automation, apps and integrations that cut the time to send.
20%Price
Monthly list price of the plan scored, mapped to fixed bands.
15%Ease of use
Free plan or trial, self-serve signup, nothing to install, single seat allowed.

Desk-research scores v1

Desk-research score v1 · checked 2026-10-22

Every score on the site today is a desk-research score. We checked each tool against the public checklist for its ranking using vendor pricing pages, help-center articles and feature pages, all on 2026-10-22. Each point links to its source on the ranking page, under "Why each tool scored what it did". We did not sign up for paid plans and have not run hands-on tests yet.

1. Which plan we score

2. Quality and Speed (0–10 each)

Each ranking publishes five Quality checks and five Speed checks. Each check scores:

The sub-score is the sum of the points. If we can't find proof, the tool gets 0 for that check, not the benefit of the doubt.

3. Price (0–10)

We use the plan's monthly list price in USD. Introductory discounts are ignored. We use the annual price per month only when the vendor publishes no monthly price, and we say so on the ranking page.

Monthly pricePrice score
$0 (free plan covers the task)10
up to $109
$10.01 – $158
$15.01 – $207
$20.01 – $256
$25.01 – $305
$30.01 – $404
$40.01 – $503
$50.01 – $752
over $751

4. Ease of use (0–10)

5. Review ratings

G2, Capterra and similar review sites blocked our automated checks, so we do not use their ratings. We use a Product Hunt rating only when every tool in a ranking has one with at least 20 reviews. Quality is then 0.8 × checklist + 0.2 × (2 × rating). In v1 that applies only to the freelance-proposals ranking. In other rankings, any rating we found is shown as context and does not change the score.

6. Ties

Equal totals are ordered by Quality, then Price, then tool name.

The raw evidence file is research/desk-research-v1.json in the site repo, and a flat export is in scores-v1.csv.

Hands-on tests (planned)

  1. Use the fixtures in /tests (brief + prompt per ranking).
  2. Use a fresh account on the plan scored above, with default settings.
  3. Run the exact prompt 3 times, keep the median output, and time each run.
  4. Score with the rubric, and record the tool version and date.
  5. Publish the outputs alongside the scores.

Comparisons and badges

Comparison pages (/compare) reuse the exact entries from the rankings both tools appear in; nothing is re-scored. Tools ranked #1 may embed a badge. Badges are free and never affect scores.

Independence

Sponsors never affect scores. Affiliate links are disclosed.