How We Rank Tools

Who Is Behind This

The Hub is built and maintained by Aydın Nasuh, founder of ShubHQ and an independent growth specialist with 19 years in the IT industry. The rankings are produced by an automated pipeline (described below) and reviewed by a human editor. We refresh each category every 3 to 6 months and track how tools move between updates, so you can see the trend, not just a one-time snapshot.

Our rankings reflect how leading AI assistants respond when asked to recommend tools in a given category. We do not fabricate scores or inject editorial bias. Every data point comes from a real LLM query.

Our Process

1. Prompt design — For each category we craft a specific prompt asking for the top recommended tools.
2. Multi-LLM querying — The same prompt goes to 5 AI assistants (see below). Each returns its own independent top-10 list.
3. Entity resolution — Tool names are normalized to canonical entities, so "UptimeRobot" and "Uptime Robot" count as one.
4. Independent judge — A separate judge model compares the 5 lists and produces a single consensus ranking, weighing how many assistants list each tool and how highly.
5. Transparent publication — Every category page shows the prompt, the assistants and versions used, the collection date, and per-assistant placement for each tool.

AI Assistants We Query

The Judge

After the 5 assistants return their lists, an independent judge model consolidates them into one consensus ranking. It weighs how many assistants include each tool and how highly each ranks it. The judge is deliberately a different model from the ones being aggregated, to avoid any single assistant's bias dominating the result.

• mimo-v2.5-pro — used in 2 runs

How to Read a Tool Row

• In AI lists (X/5) — how many of the 5 assistants included this tool at all.
• Ranked #1 (Y×) — how many assistants put it in first place.
• Per-AI ranking — each assistant's exact placement, or ✗ if it did not list the tool.

Tool Metrics (DR, Backlinks, Activity)

Alongside the AI ranking, each tool shows three objective signals about its website. These are supplementary context, not part of the AI ranking itself. They are refreshed on a roughly 30-day cycle, so they reflect a recent snapshot rather than a live figure.

• DR (Domain Rating) — a 0–100 score estimating the strength of the domain's backlink profile, sourced from established third-party SEO databases. A higher number means more, and more authoritative, sites link to it. It is a relative authority signal, not a quality verdict.
• Backlinks — the approximate number of distinct external domains that link to the tool's website, from the same third-party SEO databases. We show referring domains rather than raw link counts because it is the more meaningful measure of link diversity.
• Activity (Quiet / Active / Very active) — how frequently the tool's site has published or updated public pages recently. We count new pages from the last ~30 days: from the site's own sitemap (using each page's last-modified date) when one is available, otherwise estimated from how many pages the public search index has indexed for the domain in the past month. We then bucket that count: Quiet (under 50), Active (50–499), and Very active (500 or more). It is a freshness/publishing-cadence signal, not a measure of product quality or user activity.

Where a metric is unavailable for a site (for example, a domain with no crawlable dated sitemap and no reliable index estimate), we simply omit that badge rather than show a guess.

What Our Rankings Are NOT

• They are not editorial opinions. We report what LLMs say, not what we think.
• They are not influenced by sponsorship. Featured badges do not affect AI ranking.
• They are not deterministic. LLM outputs vary by prompt, date, and personalization. Our methodology reduces this noise but cannot eliminate it.

Overall Stats

• 19 AI list queries
• 13 tools ranked
• 1 categories covered
• 5 AI assistants used
• Data since June 2026

Prompts Used

What are the best project management tools? List your top recommendations.