Know exactly how strong each relationship is.
REI is a single 0 to 100 score for every relationship in your Nynch network, grounded in four bodies of network and trust research that have been validated across decades of social science. Most AI CRMs pick a number out of the air. We picked four foundations and named them.
The Activity Planner forecasts the next 30 days of touchpoints. Each rhythm group has its own warmth signal. Decay arriving is visible before it matters.
One score per relationship. Four foundations behind it.
REI replaces the patchwork of separate social, authority, and warmth scores you find in other tools. It is composite, grounded, and outcome-aware: composite because it combines tie strength, layer placement, commitment history, and decay; grounded because each component has a published source; outcome-aware because closed deals and stalled deals feed back into the score, so it learns which engagement patterns predict revenue for your specific business.
Granovetter tie strength
Strong, weak, and absent ties are different objects with different decay curves. A quarterly check-in with a strong tie is more valuable than five pings with a weak one.
What it gives REI: different cadences and different alert thresholds for different tie strengths. Your strong ties never get the same nudge as your cold ones.
Dunbar social layers
There is a hard cap on how many close, intimate, and meaningful relationships any person can sustain. The classic Dunbar layers are 5, 15, 50, and 150.
What it gives REI: bucket sizes that respect human limits. Nynch will not pretend you have 500 close relationships. Focus 10 sits inside layer 1 by design.
Morgan and Hunt commitment-trust
Trust is computed from a track record of kept commitments, not declared in a CRM field. Make a promise, keep it, the trust score moves. Miss it, the score moves the other way.
What it gives REI: the Say/Do Ratio as a real metric, with overdue-promise alerts and commitment-history tracking baked into every contact card.
Hawkes processes
The mathematics of self-exciting events. Each interaction lifts the score by a decaying amount. Silence lets the score cool predictably along a known curve.
What it gives REI: the temporal decay model, including the 8.33x Rhythm-Break Rule. Going 8.33 times longer than the natural cadence means the relationship is past its half-life. That is a Hawkes-process implication, not a guess.
What that means in practice.
The foundations are the credibility anchor. The behaviours are what you actually feel.
- 1 Every contact has one number, and it makes sense. No more juggling a separate authority score, social score, and warmth score. One 0 to 100, with the components visible if you want to drill in.
- 2 Cadence is differentiated. Strong ties get fewer, longer nudges. Weak ties get faster, shallower ones. The system never recommends the same touch sequence to a 25-year mentor and a conference contact.
- 3 Buckets respect Dunbar. Focus 10 lives in your inner layer. The Bucket Game daily routine flips through buckets sized to human limits, not to whatever the database can hold.
- 4 Trust is earned, visibly. Every commitment you make to a client is tracked. The Say/Do Ratio is on the contact card. Overdue promises trigger alerts before they become reasons not to refer you.
- 5 Decay is predictable. The 8.33x Rhythm-Break Rule tells you the moment a relationship is mathematically past its half-life. Nynch surfaces it before the silence becomes terminal.
- 6 Outcomes feed back in. Every closed deal, every stalled deal, every won renewal updates the weights. By month three, REI is calibrated to your actual revenue motion, not to consultants in general.
How REI compares to a generic warmth score.
Most CRMs that claim to score relationships use a single proxy: time since last contact, response rate, or a vibe metric. REI is composite, grounded, and outcome-aware. The difference shows up in which questions each system can actually answer.
| Capability | Generic warmth score | Nynch REI |
|---|---|---|
| Treats strong and weak ties differently | No | Yes (Granovetter) |
| Caps close-relationship counts to human limits | No | Yes (Dunbar) |
| Computes trust from kept commitments | No | Yes (Morgan and Hunt) |
| Models temporal decay with a published curve | No | Yes (Hawkes) |
| Updates from your real deal outcomes | No | Yes |
| Day-1 score and day-365 score are different | No | Yes |
| Auditable: every component is inspectable | No | Yes |
Nynch borrows from each foundation. We do not claim to implement Granovetter or Dunbar or Morgan and Hunt or Hawkes in pure form. We use the strongest, most-cited insight from each body of research and combine them into one score that holds up against the messy data consultants actually have. Grounded in, not implementing.
This is the framing every founder, sales rep, and product page should keep when talking about REI. It is the reason a sceptical buyer will believe you, instead of bouncing on the academic name-drops.
See REI on a live account.
Book a 30-minute demo and we will walk you through REI on a real contact list, including the Say/Do Ratio for one of our customer accounts and a Hawkes-process decay curve in action.