Set up
Survey and measurement design
Surveys tend to get longer as each team adds its own questions. We review or design your surveys and the metrics behind them, so the results are reliable and someone can act on them.
When it helps
Signs you need this.
- Completion rates are low or falling.
- Results are challenged in every meeting.
- The survey has grown by a question a quarter for years.
- You're deciding between NPS, CSAT, effort or a mix.
- Scores can't be compared across markets or languages.
How we do it
How we design a survey.
Purpose
We confirm which decisions the survey must inform. Questions that inform none are removed.
Metric choice
We select the headline and supporting measures for each touchpoint and explain the trade-offs.
Questions and scales
We write clear, single-idea questions with consistent scales, and order them to avoid bias.
Logic and length
We use routing and operational data to keep each respondent's path short.
Testing
We test with people outside the project, then check early data for drop-off and patterned answers.
What you get
What you have at the end.
- A survey document with wording, scales, routing and build notes
- A measurement framework showing which metric is used where and why
- A marked-up audit of any existing survey
- Guidance on translation and comparability
- A change log, so trends stay interpretable
From our principal consultant's work at Qualtrics
Ran dozens of design workshops for a telecoms operator and led its move from NPS to a satisfaction measure. Aligned the core surveys across dozens of affiliates of a global pharmaceutical company.
Done as part of Qualtrics teams, before UnfoldCX was set up. Qualtrics was the contracted party. More experienceWhat we need from you.
- The surveys you run today, with recent results
- The decisions each survey is meant to inform
- Stakeholders for a short round of interviews
- A test audience from outside the project
Questions
Questions you may have.
Should we use NPS?
NPS suits relationship tracking. For a specific interaction, a satisfaction or effort measure usually tells teams more about what to fix. We help you choose, and manage the switch if you change.
Can you review a survey we already run?
Yes. A review is a small piece of work and often the best place to start.
How long should a survey be?
A transactional survey should take a couple of minutes at most. We set the length from what you'll act on and cut the rest.
Related
Related services.
Feedback capture
Surveys, digital intercepts and contact-centre feedback at the touchpoints you have prioritised.
Read moreText analytics
Topic models that turn open comments into themes you can count.
Read moreProgramme review and recovery
An audit of what you run today, with ranked recommendations and a plan to act on them.
Read more
Next step
Tell us where your programme is.
The first call is free and carries no obligation. You'll speak to our principal consultant, who would lead the work.