Crewhu Surveys

Crewhu CSAT/NPS survey data: the list/search/get tools and the detractor/promoter sentiment slices, the trend, detractor-follow-up, promoter-recognition and per-tech roll-up workflows, and the edge ca

SKILL.md

GitHub

Crewhu Surveys (CSAT / NPS)

Surveys are Crewhu's primary product — short CSAT/NPS responses tied to a ticket close, a specific tech, or a project milestone. This skill covers listing surveys, finding a specific one, and isolating the two ends of the curve (detractors and promoters) so MSP managers can react.

Anti-triggers

  • "User" here is an MSP technician, not an end user. crewhu_users_* returns the staff being scored. A customer's staff account in a tenant is m365-users or cipp-users; the customer's billing or ticket contact is a PSA contact (connectwise, halopsa, kaseya/autotask).
  • Acting on a detractor — Crewhu records the feedback and stops there. Raising the follow-up ticket, logging the call-back, or crediting the account happens in the PSA.
  • Compliance or insurance questionnaires — those are also "surveys" in MSP conversation but share nothing with CSAT; use the compliance-pack or scalepad (scalepad-controlmap).

API Tools

List & Search

Tool Purpose
crewhu_surveys_list Paginated list of recent survey responses
crewhu_surveys_search Search surveys by keyword / metadata
crewhu_surveys_get Pull full detail for one survey response

Sentiment Slices

Tool Purpose
crewhu_surveys_detractors Detractor responses (low scores / negative comments)
crewhu_surveys_promoters Promoter responses (high scores / positive comments)

Common Workflows

Pull recent survey trend

  1. Call crewhu_surveys_list with a wide enough window (90 days minimum).
  2. Bucket results by week and compute average score and response count.
  3. Surface the trend line, not just the latest week — single weeks are noisy.

Detractor follow-up queue

  1. Call crewhu_surveys_detractors to get the negative responses.
  2. For each, call crewhu_surveys_get to retrieve the full comment and the responsible technician (the responses include user attribution — cross-reference with crewhu_users_get).
  3. Produce a follow-up list: customer, ticket, tech, score, comment, suggested action (call-back, account-manager hand-off, escalation).

Promoter recognition loop

  1. Call crewhu_surveys_promoters to find positive responses.
  2. Use the responsible tech to drive a recognition workflow:
    • crewhu_badges_user_recognition to see whether the tech is already trending in recognition.
    • Consider awarding a badge via the badges domain (see crewhu_badges_history_list).

Per-user roll-up

  1. Use crewhu_users_list to enumerate techs.
  2. For each tech, use crewhu_surveys_search (keyed on the user) to pull their responses.
  3. Compute average score, response count, and detractor rate per tech for a manager scorecard.

Edge Cases

  • Sparse responses — Some techs have few surveys; flag any per-user metrics with N < 10 as low-confidence rather than ranking.
  • Comment-only feedback — Some responses have a comment without a score. Surface those separately; do not coerce them into the score average.
  • Time zones — Survey timestamps are in the tenant's configured zone; normalize before comparing across tenants.

Best Practices

  • Always show denominators (response count) alongside averages.
  • Pair detractor lists with promoter lists in a manager dashboard so the picture is balanced.
  • Capture the tech for every response in your output so action items are unambiguous.

Related Skills