6. Engagement & Culture

The say-do gap

Measure events, not opinions

Use when

Asked to refresh the values, or to measure culture at all

The say-do gap, shown as a sequence of 6 parts: The finding, 1. Reversal, 2. Candour, 3. Removal, 4. Contradiction, System traces.
  1. 1

    The finding

    MIT SMR and Glassdoor analysed 1.4m reviews across 500+ large employers. ~Three quarters publish core values. The correlation between the values emphasised and how employees rate the company on those same values is close to zero.

  2. 2

    1. Reversal

    Did you see a decision reversed after someone junior raised a concern?

  3. 3

    2. Candour

    Did you tell your manager something they would not want to hear, and what happened?

  4. 4

    3. Removal

    Was anything removed from your workload?

  5. 5

    4. Contradiction

    Name one recent decision that contradicted a stated value.

  6. 6

    System traces

    Meeting hours, after-hours volume, approval cycle times, time from red flag to resolution.

Agreement with a statement can only be spun. Events can be counted and disputed.

How to use it

Run the event instrument quarterly on a 30-day recall window. Rewriting the values is not a culture intervention. Watch-out: publish the counts, including the uncomfortable ones, or the instrument becomes another survey nobody believes.

Card details

Model type
Sequence
Origin status
Established
Content last reviewed
28 July 2026

Origin

Established

Donald Sull, Charles Sull and Andrew Chamberlain, MIT Sloan Management Review and Glassdoor, 2019. The four-question event instrument is ORIGINAL. Harsimran Kaur Kapoor, 2026.

Used in these workflows

Neighbouring models

NextThe psychological contract

Was this useful?