Transparency
A concise overview of how scores are built, what inputs we use today, and the guardrails that keep the index independent.
The AImpulse Index is a 0–100 momentum score for each tool. It is not a review score from our editors and not a paid rank. It summarizes how a tool is moving across six public signal families that we can observe without vendor cooperation.
Today, those six values are produced by our scoring pipeline from public web evidence (primarily open search results and public pages), then blended with the fixed weights below. We do not currently claim direct live feeds from GitHub, Reddit, Hacker News, or job boards for every tool. Where a per-signal row is missing on a profile, the UI marks it Derived from the overall score; where a row exists it is marked Stored.
Scores are recomputed on a scheduled pipeline (daily recalibration; fuller rescored runs when enabled). When a run is delayed, we prefer a short note on Latest News over silently inventing mid-week movement.
Weights below match the current scorer (they sum to 100%). Exact blending can still be tuned as we improve measurement quality.
Data sources
Public web presence and share-of-voice we can observe without vendor cooperation — search results, public posts, and similar open signals.
How it's weighted
20% of the composite.
Example
A tool that keeps showing up in public discussion and branded search results scores higher here than a one-day press blip with no follow-through.
Data sources
Public practitioner chatter visible on the open web — forums, comparative posts, tutorials, and review-style discussion where we can see it.
How it's weighted
15% of the composite.
Example
A tool with recurring “how do I…” threads and comparative debates across communities scores high on Community Discussion.
Data sources
Public ecosystem and integration strength — docs, repos, APIs, and developer-facing surfaces that show up in open search and directories.
How it's weighted
15% of the composite.
Example
A product with clear public developer surfaces and ecosystem mentions outscores a closed tool with little visible technical footprint.
Data sources
Public news articles, funding announcements, and reputable tech coverage we can find without paywalled private feeds.
How it's weighted
12% of the composite.
Example
A company with diversified coverage (product + market + funding) outscores one outlet repeating the same release.
Data sources
How central the tool looks inside its declared category — the largest single input to the composite today.
How it's weighted
28% of the composite. This is the heaviest signal by design.
Example
The #2 tool in a crowded category can still score strongly here if public evidence shows it is a category default, not a niche side project.
Data sources
Public proxies that adoption is moving now — hiring mentions, review velocity, marketplace signals, and similar open evidence where available.
How it's weighted
10% of the composite.
Example
A tool showing up in hiring templates and practitioner reviews at an increasing pace scores high on Adoption Velocity.
Tiers are labels derived from the headline score. They help readers scan the Top 100 and category pages quickly.
| Tier | Score range |
|---|---|
| Elite | 86–100 |
| Dominant | 71–85 |
| Rising | 46–70 |
| Established | 21–45 |
| Emerging | 0–20 |
The 12-week chart prefers recorded score updates when we have them. If we do not, it shows a straight projection from the latest score delta — and says so. We do not invent a wavy “history” to look more active.
Dollar amounts on tool profile pages appear only when hand-verified against the vendor's public pricing page or stored in our database as verified JSON. If we have not verified a product's plans, we show the index pricing model label (Free, Freemium, Paid, Enterprise) and a qualitative guide — never invented $15 / $40 tiers.
See something wrong? Email contact@aimpulse.app with the vendor URL and we will correct it.
For popular tools we publish canonical X vs Y pages with scores, signal breakdown, and an editorial verdict. These live at /compare and are included in our sitemap for search engines.
If you represent a tool and believe a public signal has been mis-attributed or a data source is stale in a way that materially affects the index, contact us with specifics (URLs, dates, and the signal family). We do not change scores based on opinion alone, but we do fix factual ingestion errors.
contact@aimpulse.app