Customer Research
What customers actually value, switch for, and pay for, measured, segmented, and sized.
Customers are polite in interviews and inconsistent in surveys; good research designs around both. We combine qualitative depth (jobs-to-be-done interviews) with quantitative structure (conjoint, MaxDiff, statistically clustered segmentation) so you learn what customers do and trade off, not just what they say.
Who this is for
Product and growth leaders choosing what to build, for whom, and how to position it.
How the work is done
Qualitative depth
Jobs-to-be-done interviews with laddering to surface the progress customers are hiring for, switching triggers, and anxieties. Sampling designed for disconfirmation, recent switchers and churned customers included, not just fans.
Quantitative structure
Survey instruments with bias controls (sampling frames, attention checks, social-desirability mitigation). MaxDiff for feature priorities; choice-based conjoint for willingness-to-pay and packaging trade-offs, analyzed with hierarchical Bayes where sample size allows.
Segment and size
Segmentation via k-means or latent-class clustering on need and behavior variables, not demographics for their own sake, validated for stability, then sized and mapped to acquisition channels.
Translate to decisions
Findings land as decisions, not posters: which segment to lead with, which features move choice, what price structure the trade-offs support, and the message hierarchy the evidence justifies.
How the Customer Research engagement runs
We begin with the decision, use the evidence that can genuinely change it, and make the reasoning reviewable from first input to final handover.
What we need to begin
- A customer list with segment tags, and explicit permission to contact.
- Churned customers and lost-deal contacts, which usually teach more than satisfied ones.
- Existing research: surveys, interview notes, support tickets, and sales call records.
- Product analytics showing what customers do, to set against what they say.
If an input is unavailable, we state the gap, its effect on confidence, and the agreed workaround. It is never quietly ignored.
Your four-phase engagement map
-
Phase 1
Qualitative depth
Jobs-to-be-done interviews with laddering, sampled for disconfirmation, including recent switchers and churned customers.
-
Phase 2
Quantitative structure
Surveys with bias controls; MaxDiff for feature priorities and choice-based conjoint for willingness to pay, analyzed with hierarchical Bayes where sample allows.
-
Phase 3
Segment and size
Cluster on need and behavior via k-means or latent class, validate for stability, then size and map to acquisition channels.
-
Phase 4
Translate to decisions
Land findings as decisions: which segment to lead with, which features move choice, what price structure the trade-offs support.
Methods and models we draw on
- Jobs-to-be-done interviewing
- Laddering
- MaxDiff analysis
- Choice-based conjoint (hierarchical Bayes)
- K-means & latent-class segmentation
- Survey bias controls
- Switching analysis
Methods are chosen for the problem, not the brochure, expect a subset of these, applied properly, plus whatever the evidence demands.
The decision this enables
Build/position/price decisions grounded in measured preference, not the loudest customer’s opinion.