Based on real professional adoption data, not opinions or reviews.

How MindovAI measures AI adoption

MindovAI is a behavioral data platform, not a review site. Every score is calculated from real usage signals submitted by professionals, weighted, checked for integrity, and versioned like any serious index. This page documents exactly how.

01 — Core Principles

Four rules the whole methodology follows.

Before any formula, these are the commitments everything else on this page is built to honor.

01

Adoption measures usage, not popularity. A score reflects what people actually do, not what they say or vote for.

02

Vendors cannot pay to improve their score. No subscription tier or sponsored placement ever enters the calculation.

03

Scores change because usage changes, not opinions. No manual override, no editorial adjustment.

04

Every methodology change is publicly versioned. Dated, explained, and never applied silently.

02 — Adoption Score

How much is a product actually used?

The Adoption Score is a single number from 0 to 100 that reflects real, current usage. It is calculated entirely from a product's own signals. It never depends on how other products perform, and it is never affected by what a vendor pays. It is built from four independent dimensions.

V
30% Weight

Volume

The number of active usage signals a product has received, weighted by how recent they are.

F
25% Weight

Frequency

How often people use the product, measured in weekly usage instances from occasional to multiple times a day. More frequent use scores higher.

D
25% Weight

Dependency

How critical the product is to a professional's workflow, from barely noticeable if it disappeared to impossible to work without.

B
20% Weight

Adoption Breadth

How many distinct industries use the product. Broad adoption signals wider utility than a single niche.

Formula
Adoption Score = Volume × 0.30 + Frequency × 0.25 + Dependency × 0.25 + Breadth × 0.20 Volume and Breadth grow on a diminishing-returns scale: each additional signal or industry adds less than the one before it, so the score never hits a hard ceiling. This scale is entirely self-contained. It is never compared to, or affected by, any other product.
NOTE

Rank is relative. The score behind it is not. A product's position within its category can move as other products gain or lose signals, since Rank is a comparison. But the Adoption Score itself changes only when the product's own usage changes. See Ranking below.

03 — Signal Collection

What a professional actually submits.

A signal is not a click or a star rating. It is a short, structured declaration of real product usage, collected directly from the person using it.

01

Role & industry

Job function and sector, used for Adoption Breadth.

02

Product & use case

Which product, and what it is used for.

03

Frequency

How many times per week, from occasional to multiple times a day.

04

Dependency

How critical the product is to the workflow.

8 steps total, including company size and paid status
~90 seconds average completion time
No account required to submit
04 — Signal Weighting

Not every signal counts the same.

A signal submitted yesterday carries more weight than one submitted two years ago. Instead of a hard cutoff window, each signal loses weight gradually over time, so scores stay current without abrupt jumps.

Today
100%
30 days old
91%
90 days old
75%
6 months old
56%
1 year old
31%
2 years old
10%

This is how much of a signal's original weight remains in the Adoption Score, based on its age. A signal never drops to zero. Even old usage still counts a little, since a product once trusted rarely loses all its relevance overnight.

Underlying formula: weight = e^(−0.0032 × days), floor 1%
NOTE

Current limitation, stated plainly. MindovAI does not yet verify signals through email or account login. Every real signal today is collected anonymously, with IP and device level deduplication to block spam and repeat submissions. All real signals are weighted identically for verification purposes in v1.0. Email and account based verification, with a corresponding weight, is planned for v1.1. See Versioning below.

05 — Worked Example

See the calculation, not just the result.

Every score on MindovAI can be broken down like this. Figures below are illustrative.

C

ChatGPT

Adoption Score 91.4 / 100 · Category: Chatbots
Volume (30%)27.9
Frequency (25%)22.6
Dependency (25%)21.8
Breadth (20%)19.1
Volume and Breadth are shown as their weighted contribution on the diminishing-returns scale, not a raw signal count. This score would be identical even if every other product in the Chatbots category doubled its signals overnight.
06 — Confidence

How much can you trust this score?

Confidence is entirely separate from the Adoption Score. It does not measure how good a product is. It measures how statistically solid the data behind its score is, based on sample size and data integrity.

01
Very Low
Not enough signals to draw a conclusion. Excluded from category ranking.
0–9 signals
02
Low
Early signal. Enough to appear in ranking, not yet statistically strong.
10–49 signals
03
Medium
A credible, defensible sample size for category level comparison.
50–199 signals
04
High
A strong, well distributed sample. Reliable for business decisions.
200–999 signals
05
Very High
Large scale, statistically robust adoption data.
1,000+ signals
NOTE

Confidence can be downgraded independently of volume. If more than 40% of a product's signals are flagged as duplicate, suspicious, or rejected, or if less than 20% of its data comes from the past 12 months, its Confidence level drops by one tier, regardless of total signal count. A large but stale or low integrity sample is never treated as equal to a fresh, clean one. In practice, a high Adoption Score is very hard to reach without a meaningful base of real, verified signals: the score itself scales down sharply whenever real signal volume is low, so Score and Confidence tend to move together.

07 — Ranking

Ranked within category. Never across.

Comparing a chatbot to a video generator on a single global scale is not meaningful. They do not compete for the same adoption. MindovAI ranks products only against others in the same category, by ordering their independently calculated Adoption Scores.

R

Category ranking

Products are ordered by Adoption Score within their category. The score behind each position is calculated independently. Rank changes when the order changes, not when any individual score does.

E

Emerging section

Products below Low confidence are not included in the numbered ranking. They appear in a separate, clearly labeled Emerging section, visible but never ranked ahead of what the data can support.

08 — Momentum

The score is a snapshot. Momentum is the trend.

Momentum tracks how a product's Adoption Score has moved over the past 7 and 30 days. It is shown separately and never folded into the score itself. One metric shows where a product stands, the other shows where it is heading.

09 — Recalculation Cadence

When a score actually updates.

The Adoption Score updates the moment new usage is submitted. Only metrics that require comparing products against each other run on a daily cycle.

REAL TIME

Adoption Score

Every dimension, Volume, Frequency, Dependency, and Breadth, recalculates immediately when a product receives a new signal. Nothing about it waits for another product's activity.

DAILY

Rank

Category position is a comparison between products, so it is captured once per day to keep it stable and avoid rankings shifting mid-comparison for reasons unrelated to a given product.

DAILY

Momentum

The 7 and 30 day trend is calculated from daily snapshots of the Adoption Score, so a consistent history is always available.

10 — Data Integrity and Anti-Fraud

What stops someone from gaming a score.

MindovAI does not accept votes, star ratings, or reviews. Only declared usage signals, filtered through several integrity checks before they count toward any score.

01

Deduplication

Each signal is checked against IP hash and device fingerprint. Repeat submissions for the same product within 24 hours are rejected automatically.

02

Status filtering

Every signal carries a status: Active, Duplicate, Suspicious, or Rejected. Only Active signals contribute to a product's score.

03

Score integrity is not for sale

Vendor subscription tier has zero influence on Adoption Score, Confidence, or Rank. Paid plans unlock analytics dashboards, never better data or higher placement.

04

Synthetic data, transparently

MindovAI currently supplements early-stage products with a small amount of synthetic usage data, derived from public indicators of professional usage such as job listing requirements and public developer activity. It is always flagged internally, weighted far below real submissions, and never counted toward Confidence. This is temporary and will be phased out as real usage signals accumulate.

11 — Data Privacy

What happens to the data behind a signal.

Signal submission is anonymous by design. Here is exactly what is collected, and what is not.

IP address

Never stored in raw form. Converted to a one-way cryptographic hash used only for duplicate detection, then discarded from readable form.

Device fingerprint

An anonymous technical identifier used solely to detect repeat submissions. Not linked to any personal identity.

Country

Derived from network location at submission time. Used in aggregate for Adoption Breadth, never displayed per individual signal.

What is never collected

Name, email, or company identity are not required to submit a signal in v1.0. No signal is ever linked back to an individual person publicly.

12 — Versioning

If a score changes, there is a reason on record.

MindovAI's methodology is versioned like any serious index. A score can move for exactly two reasons: real usage changed, or the methodology itself was refined. Refinements are always dated and explained here, never applied silently.

CURRENT — v1.0
Live since [deployment date]
  • Adoption Score built from Volume, Frequency, Dependency, and Adoption Breadth (industry), calculated independently for each product
  • Volume and Breadth use a diminishing-returns scale with no hard cap, and no dependency on any other product
  • Signal weighting by freshness, continuous decay. Verification based weighting reserved for v1.1
  • Confidence based on sample size and signal integrity, independent of the Adoption Score
  • Rank is category relative and recalculated daily. The Adoption Score behind it is absolute and recalculated in real time
  • No single global cross category ranking
Planned — v1.1

Email and account based signal verification, with a corresponding weight in the Adoption Score.

13 — Dispute Process

If a score looks wrong, here is what happens.

Vendors can request a review of their product's score at any time. Disputes never result in a manual score override. They result in a check of the data behind the number.

01

Vendor flags a concern

Through the vendor dashboard or direct contact, citing the specific metric in question.

02

Signal audit

The underlying signals are reviewed for status, duplication, and integrity flags.

03

Correction if warranted

If faulty signals or a calculation error are found, the score is recalculated and the fix is logged.

04

Response to vendor

The vendor receives an explanation either way, including cases where no change is made.

14 — Citing MindovAI

How to reference a score.

MindovAI data is free to cite in research, journalism, and analyst reports, with attribution and a link back to the product page or this methodology.

"According to MindovAI's Adoption Score, [Product] holds a score of [X] in the [Category] category as of [Date], based on [N] usage signals (source: mindovai.com/methodology)."
15 — Frequently Asked Questions

The questions we expect you to ask.

No. Subscription tier and any sponsored placement affect only visibility and analytics access, never the Adoption Score, Confidence level, or Rank. Sponsored placements, where used, are always clearly labeled as such.
No. The Adoption Score is calculated entirely from a product's own signals. It can only move if that product's own usage changes. What can move is Rank, since it reflects a product's position relative to others in the same category.
A signal is a declared instance of real usage: a professional confirming they use a specific product, how often, how critical it is to their workflow, and their role and industry.
Adoption Score is comparable in principle across categories, since it is absolute, but ranking is still kept within category because usage patterns and typical signal volume differ too much between, say, chatbots and niche developer products for a single ordered list to be meaningful. A separate Trending view highlights recent activity instead.
Signal weight decays continuously with age. A score can move slightly over time even without new activity, reflecting that older usage carries less certainty than recent usage.
Ranking is not shown as a percentile position for categories with fewer than 5 tracked products, since a relative comparison is not statistically meaningful at that sample size. Adoption Scores are still displayed normally, since they do not depend on category size.
Every signal is checked for IP and device duplication and assigned a status. Only signals marked Active contribute to a score. Duplicates, suspicious entries, and rejected submissions are excluded entirely.
Adoption Score answers how much a product is used. Confidence answers how much you can trust that number. A product can have a high score and low confidence if its usage is strong but based on a small sample.
It appears in the Emerging section with no score displayed until it receives its first signals, rather than showing a misleading zero.
16 — Governance

What MindovAI will never do.

No pay to rank. Vendor payment status never enters any scoring calculation.

No reviews or star ratings. Only declared, behavioral usage signals count.

No silent methodology changes. Every revision is dated and documented here.

No hidden thresholds. Every number on this page is the real number used in production.

Find your product

Search for your product below. Once selected, we'll verify you represent it before granting access to your profile.