Net Promoter Score

NPS uses three response groups

The official Bain Net Promoter System calculation page classifies 9 and 10 responses as promoters, 7 and 8 as passives, and 0 through 6 as detractors.

The formula is:

percentage of promoters − percentage of detractors = NPS.

Passives remain in the total respondent denominator even though they are not added or subtracted. The score ranges from negative 100 to positive 100.

Define the survey event

Record who was invited, which experience the question refers to, invitation date, response window, channel, respondent role, and whether one response per person, household, or episode counts.

A parent, client, former client, referral source, and employee answer from different perspectives. Report them in separate cohorts rather than combining roles into one score.

Build a fair invitation cohort

Create the eligible list before responses arrive. Deduplicate people or households under the chosen unit, remove only predeclared ineligible records, and preserve the reason for every exclusion. Avoid surveying only families with recent positive interactions or staff-selected recipients.

Track whether each invitation was sent, delivered, opened when that evidence is available, completed, declined, or inaccessible. One household should not receive repeated invitations because several services or staff touched the same episode unless the survey design explicitly calls for event-level sampling. Set a quiet period so frequent service does not create survey burden or greater weight.

A fictional NPS calculation

Pineglow Services invites 100 fictional eligible households after a quarterly service review. Forty submit one valid response during the stated window, so response rate is 40 of 100, or 40%.

Among respondents, 18 are promoters, 10 are passives, and 12 are detractors. Promoters are 18 of 40, or 45%. Detractors are 12 of 40, or 30%. NPS is 45 − 30 = 15.

The report shows NPS 15 and the 40% response rate together. It does not imply that the 60 nonrespondents share the same views.

Make the survey accessible

Offer usable channels, language access, disability accommodations, AAC-compatible options, plain wording, and enough time. Explain who will see responses and whether they are identifiable.

Avoid requiring speech, phone use, literacy, or one motor response. A caregiver can support access without authoring the person’s answer. When one household includes several perspectives, define whose response the metric captures.

Protect voluntariness and privacy

Survey choice should not affect scheduling, staff behavior, benefits, or continued care. Avoid having the treating clinician pressure someone for a high score. Use purpose-limited access and a clear route for urgent safety or privacy concerns.

If the survey asks for details, collect only what is needed for the stated follow-up. Separate public testimonial or marketing permission from service feedback.

Examine response bias

People with strong positive or negative experiences may respond at different rates. Digital-only surveys can underrepresent people with access barriers. Staff timing or wording can influence the result.

Track invitations delivered, responses, partial responses, declined participation, and access failures. Compare trends only when cohort, question, timing, and channel remain reasonably consistent.

Follow the score with a reason

An optional open question can ask what shaped the rating and what would improve the experience. Code themes such as communication, scheduling, fit, access, staff consistency, cost clarity, respect, or outcomes while preserving the original text for authorized review.

Close the loop when someone requests contact. Safety, privacy, discrimination, or care concerns need the appropriate escalation path regardless of the score group.

Keep NPS away from clinical inference

A recommendation rating can reflect location, parking, cost, expectations, staff warmth, outcome, or a single recent event. It cannot diagnose a clinical problem or establish treatment effectiveness.

Pair NPS with direct client feedback, complaints, access, continuity, safety, clinical outcomes, and burden. Never use a high score to dismiss a serious concern or a low score to label a family difficult.

Use trends with context

Report the counts in each group, NPS, response rate, period, and cohort. Small samples can move sharply after one response. Confidence intervals or a longer reporting window may help when volume is low.

Avoid ranking clinicians or teams on NPS alone. Case mix, invitation practices, response access, and small denominators can distort comparisons and create pressure for selective surveying.

Keep sources in scope

The CASP resources page links organizational and ABA practice materials, some requiring separate access or licensing. It does not establish an ABA-specific NPS standard. Bain supplies the calculation; the ABA safeguards and workflow above are editorial controls.

Before acting on a change in NPS, compare invitation delivery, response access, respondent mix, question wording, timing, and sample size. Route individual concerns through their proper safety, privacy, access, or service-recovery pathway rather than waiting for the aggregate score.