The Evidence: the network's numbers, window and method

The rule

We chose the lowest, most defensible truth over the prettiest lie

This paragraph isn't a polite preamble: it's the part of the page that makes every figure that follows credible. Without it, the numbers below are exactly the same as anyone else's.

We could have published that nine out of ten users find the right person on the first match. The raw data would have allowed it: we'd just have had to skip counting a few real deliveries. We counted every one, and the number went down.

In a market where everyone inflates their numbers, honesty isn't a cost — it's the weapon. Whoever massages their own numbers isn't founding a discipline, they're doing advertising — and the difference only shows up on one number: the one that hurts.

From this follows the publication rule, and it has no exceptions: no threshold, no condition, no deferral. A figure is published even when it's low, even when it's gotten worse, even when it lands in the wrong month. It's the Fifth Law applied to ourselves, and a discipline that measures everyone else and absolves itself isn't a discipline: it's a marketing department.

There's exactly one check before publishing, and it isn't a threshold

The platform can raise percentages up to marketing floors configured inside it, and flags when it does so with a dedicated field. When that field is active, the value the interface returns isn't the real one: it's the raised one.

The consequence is the opposite of a block. The number still gets published — but it's recalculated from the raw data, and we publish the true figure instead of the one inflated by the system.

Checked on 16/08/2026: no floor applied. Every percentage on this page is the real value.

An example of what this means in practice: the floor configured for the "right person on the first match" figure is 50%. The real value is 68%. If it ever dropped below 50, the platform would show 50 — and we'd publish the true number.

The network

How big it is, and how it was counted

Only qualified profiles: people who filled in a bio and reciprocity goals. Unsubscribed profiles are excluded. It's a stricter count than would be convenient, and it's why the number is lower than what you read a year ago.

2,141

Qualified Connectors

As of 13/08/2026 · profiles with bio and reciprocity goals filled in only, unsubscribed profiles excluded

286,365

Companies reachable through the network

As of 13/08/2026 · sum of the B2B contacts Connectors declared during profiling, not an address book in our possession

811

New qualified Connectors in the last twelve months

August 2025 – August 2026 · gross intake; net of churn the network grew only a little — see the note below

20 of 20

Italian regions with at least one Connector

As of 13/08/2026 · 124 provinces and 29 sectors represented

The number that matters more than the total

Two thousand people isn't a big network. 286,365 reachable companies is — and that's the figure that really describes the pool, because every Connector brings their own sphere of influence with them: an average of 134 declared B2B contacts and 8 out of 19 sectors.

It's worth saying what that number isn't: it isn't a list we own. It's the sum of what people declared they can reach, during profiling. None of those names are in our database, and the method explains why.

Effectiveness

What happens to a match after it's delivered

All the metrics in this section exclude Scouting_Match entries — matches that come from a commissioned search rather than the ordinary calculation: mixing them in would inflate the results.

68%

Finds the right person on the very first match

All-time · rating ≥ 4.5 on the first attempt, Scouting_Match excluded

85%

Finds it by the second match

All-time · same criterion, rating ≥ 4.5

91%

Finds it by the third match

All-time · same criterion, rating ≥ 4.5

2.6

Matches needed on average before the right one

All-time · this is the system's real friction, not a theoretical average

94%

Of delivered matches receive a signal

Last 30 days as of 16/08/2026 · a signal is a yes, a no, or a not now

86%

Of meetings held are rated 4.5 or higher

Last 30 days as of 16/08/2026 · rating given by whoever experienced the meeting

4.2

Average rating of a meeting, out of 5

All-time · average of every rating collected, not only the successful meetings

76.7%

Average affinity of delivered matches

As of 13/08/2026 · harmonic mean with relational multiplier — the platform also has an arithmetic mean, which is higher, that we don't use here

Why the windows aren't all the same. The response rate and meeting ratings are over thirty days; "by the first, second, third match" are all-time, because over thirty days the sample would be too small and the number would swing from week to week. It's a methodological choice of the platform, and it has to be stated whenever the two families of numbers sit on the same page — like here.

The reading becomes automatic

The numbers in real time

The panel below reads MigaMATCH live, as you look at it, instead of carrying the date of the last manual reading.

If the panel doesn't load, the numbers with their date and method are still the ones in the sections above.

The archive

The people who did it, filmed telling the story

Real footage, in real rooms, with people speaking in their own name. It's the material that goes into the discipline's productions — and the part that can be called evidence, because it happened.

Ital Tetti · Rimini Valerio Annichiarico, after the first month

The cooperative's president talks about the first contracts won through referrals, just as the programme was getting started. It's a snapshot of a beginning, filmed before anyone knew how it would turn out.

Filmed in 2021 · open on YouTube (in Italian)
Ital Tetti · Rimini The same company, eight months later

The third interview, with the results grown. Note the figure you hear in the video: the complete, closed count says €3,086,300 in revenue generated through referrals between September 2021 and September 2022, notarised one by one. "Eight months" is the date of the interview, not the measurement window.

The case, with the measurement method
Forum Connections · Florence, 1 March 2025 What changes when the tables aren't left to chance

Two testimonials recorded live in the room, about the work of the algorithm that decides who sits next to whom. It's the first Forum where the tables were arranged by AIRA© instead of by hand.
Forum Connections · a Connector Sergio Pezzotti talks about his own results

A Connector from the network, not a hired spokesperson: he talks about what happened to him in the meetings, with his own numbers.

The video only starts on click, from youtube-nocookie.com: before that moment, none of the four clips opens a connection to YouTube.

The founder in a public setting, on the same subject: Relational Engineering and innovation (in Italian).

The boundary

What isn't evidence, and why we say so here

A page called "the evidence" is also judged by what it leaves out. These three things exist, some of them are even good, and none of the three gets into this page.

TG29 and its scenes

It's a narrative piece set in 2029: a fictional newscast telling where the discipline will have gotten to three years from now. The figures spoken in those scenes — companies involved, revenue growth, numbers for a company that shares its name with no real company — are written for the story. They are not measured and are not attributable to anyone.

On screen the distinction is clear, because the on-screen caption states it's a dramatisation and the real footage is marked as such. A web page can't make that distinction: someone who reads a story inside an evidence page reads it as true. That's why what's here is the name, not the link.
Example calculations

"Forty agents at three meetings a month makes a hundred and twenty meetings" is correct arithmetic, and it isn't anyone's result. Where a calculation like that helps explain a mechanism, it's labelled as an example calculation — and it never appears lined up with the measured numbers.

The difference is simple: a measured number has a window and a method. An example calculation only has assumptions.
Projections on potential

With two hundred Connectors you indirectly reach fifty thousand people, and at second degree more than a million. It's a multiplication, not a measurement, and as such it remains a theoretical potential: it describes how big the space is, not how much of it has been used.

The same goes for any sentence that starts with "even if just ten per cent…". There isn't one on this page.

The rule, in one line

A strong number only gets published with its measurement method right next to it. A number with no method is exposed; a number written for a story isn't a number. It's the same rule that kept our best-documented case out of the spotlight for months: until the measurement method was tracked down, that figure wasn't used.

The network, at a glance

The active clubs

Each tile is the APP PRM of an active relational club: a client company that built its own network of Connectors with us. It isn't a measured number like the others on this page — it's just the network, as it is. Drag to rotate it, bring a club to the front, click to open its landing page.

Drag to rotate the network

The first step

Your numbers, with the same method

AIRA-SCAN© applies the same rule to your company that this page follows: it measures, states its method, and doesn't project anything it can't prove. Ten questions, three minutes, and you can redo the score by hand.

Then the extended report arrives by email, and within one working day someone calls you: you look at the result together and assess whether the AIRA-DX© Stoic Analysis makes sense for your company. It's free and limited to twelve a week, so it isn't owed to whoever does the check-up: it's decided together.

Want to understand first how these numbers come about? There's the method, and there are the six laws it descends from.

Measure your Relational Capital — ten questions, three minutes