Satisfaction audit: tracking what is actually good

Regularly inventory what genuinely works in the relationship — not as gratitude practice, but as an honest baseline.

Why it works

Rusbult’s model shows satisfaction is the strongest predictor of commitment, but negativity bias means dissatisfaction is easier to remember than satisfaction. An active satisfaction audit counteracts this by making the positive visible as data, not just as an abstract impression. When satisfaction is low, the audit identifies specific deficits rather than a global verdict, which makes them actionable.

How to do it

1. Once a month, list five to ten things that are genuinely working in the relationship — specific behaviors, not general attributes.
2. For each, estimate how much it contributes to your overall satisfaction (low/medium/high).
3. Compare this month’s list with last month’s: is the total pool growing, shrinking, or stable?
4. Identify which category of need (companionship, growth, desire, security, support) is most and least met.
5. Use the least-met category as a topic for one direct conversation this month.

Evidence

Satisfaction is the most robustly studied predictor of relationship commitment and stability across Rusbult’s and others’ research. Longitudinal data show that satisfaction measured at one time point predicts commitment and relationship continuity months and years later. (observational)

Correlational; the model predicts commitment from satisfaction but not all dissatisfied people leave, and not all satisfied people stay. Investment and alternatives moderate the relationship.

Sources

• Rusbult (1983), A longitudinal test of the investment model, Journal of Personality and Social Psychology
• Le & Agnew (2003), Commitment and its theorized determinants — meta-analysis, Personal Relationships

Common mistake

Treating the satisfaction audit as a gratitude exercise rather than an honest assessment — inflated satisfaction ratings obscure the actual picture and prevent targeted improvement.