Outcomes · ROI

The economics of loneliness, priced honestly.

We publish the factors we model from, with their primary sources. We do not publish vendor-estimated savings numbers. Your CFO can rebuild our math; your analytics team can stress-test our assumptions.

The factors

What we model from, and where each number comes from.

Each factor below is a published finding from a primary source. No adjustments. No weighted vendor blends.

5 sourced factors
01Annual turnover cost (per exit)
0.5x to 2x annual salary
SHRM, 2023
02Lost productivity from disengagement
$322B per year (US)
Gallup, 2023
03Mortality risk of social isolation
Equivalent to 15 cigarettes/day
Holt-Lunstad et al., 2010
04Absenteeism linked to loneliness
+5.7 additional days/year
Cigna Loneliness Index, 2020
05Retention lift from belonging
+56% less likely to leave
BetterUp Member Study, 2022

Factors are published findings. Sources are listed inline so your analytics team can pull the primary paper and verify the range we use. When a source updates, we update.

How we model

Three honest steps, not a savings calculator.

Our calculator is really a methodology. Inputs you bring, math we publish, outputs your finance team can defend.

Step 01

Inputs you control

  • Headcount and participating cohort size
  • Baseline belonging and connection signals
  • Current turnover, absenteeism, and EAP utilization
  • Deployment window (typically 6, 12, 24 months)
Step 02

Model we publish

  • Published ROI factors, with primary sources
  • Regression on correlated outcomes, not causal claims
  • Confidence intervals on every output, not point estimates
  • Methodology memo attached to every model run
Step 03

Outputs we report

  • Annualized savings range (low / base / high)
  • Attribution boundary: what we can and cannot claim
  • Sensitivity table across headcount, cohort, baseline
  • Summary page your finance team can countersign
Illustrative

A model session, rendered flat.

This is not a live calculator. It is a flat display of what the inputs and outputs look like when we sit with your finance team. The real session produces a memo, not a page.

Illustrative only
Model session · previewFlat display
Inputs
Participating headcount2,500
seats
Baseline belonging score42
cohort median
Cohort size250
per program
Deployment window12 months
typical
Output · annualized savings range
Low$0.9MConservative. Only reduced turnover counted.
Base$1.7MReduced turnover plus absenteeism signal.
High$2.8MAbove factored at the upper end of the sourced range.

Illustrative. Real sessions ship a PDF memo with assumptions, sensitivity table, and methodology footnotes. Confidence ranges are what count, not point estimates.

What we won't model for you.

Four refusals that keep this model honest. If a procurement team asks for any of them, we decline and explain why.

  • 01

    We will not invent a causal claim where the data supports only a correlation.

  • 02

    We will not present vendor-commissioned ROI numbers as independent evidence.

  • 03

    We will not inflate the sample by counting non-participants as benefiting.

  • 04

    We will not bury assumptions in a footnote. They lead every output.

Foundation-backed

Elitesgen, Inc. is wholly owned by Elites Generation Foundation, a 501(c)(3) whose charter legally forbids behavioral advertising and the sale of individual user data.

Visit elitesgen.org
Modeling for your organization

We will run a model with your numbers. Not for you. With you.

Bring your headcount, baseline, and target cohort. We will sit with your finance team, rebuild the math on a shared screen, and send the workings in a document you can store.