Help Center
Analytics & Insights
Turn raw responses into trends, segments, sentiment, themes, and decision-ready findings.
What this guide helps you do
Move from charts and raw answers to clear findings, evidence, and recommendations your team can trust.
Before you start
- Collect enough responses for meaningful comparison.
- Choose the segments you care about.
- Know the decision the analysis should support.
What you will do
- Start with the overview
- Segment the audience
- Read open text themes
- Monitor engagement health
- Activate analytical layers when needed
You are done when
- You can explain what changed.
- You know which segments or themes matter most.
- You have evidence to support the recommendation.
Overview
Analytics helps you understand what happened and why. Sentink combines dashboards (completion velocity, engagement curves, funnel drop-off partial vs complete), structured charts, segmentation, response quality overlays, demographics, labelled distribution sources, sentiment and theme clustering for open-ended text—with optional splits by gender and age when captured—scoped per open-ended field or pooled, narrative AI summaries grounded in filtered slices (see AI & intelligence), and conversational chat in natural language. Pair these views with Smart Data Layers in the Advanced Analytics article when you maintain curated analytical overlays without cloning surveys.
When to use this feature
Step-by-step guide
Start with the overview
Review response volume, completion rate, score distribution, and top-level trends.
Segment the audience
Compare departments, locations, channels, demographics, or time periods.
Read open text themes
Use sentiment and theme grouping—optionally split by gender and age—and scope per open-ended field or pooled sets before jumping to narratives.
Monitor engagement health
Review funnel drop-off curves, velocity cards, completion vs partial trends, duplicated device detection cues, geography overlays, labelled source mix, automated retarget prompts for stalled respondents.
Activate analytical layers when needed
Switch Smart Data Layers (cleaning overlays, segments, weights) documented in Advanced Analytics.
Interpret with sample size
Segmented analytics are most useful when each segment has enough responses to support a conclusion.
Screens / UI explanation
Charts explain structured answers; theme stacks explain qualitative evidence.
Layer switching should be explicit in stakeholder decks.
Exports respect filters and quotas.
Best practices
Do not over-read tiny segments.
Pair charts with verbatims sourced from scoped AI passes.
Log which distribution label produced each insight.
Tips
Use engagement curves immediately after UX changes.
When AI chat feels vague, constrain by question, timeframe, cohort, demographic cell, labelled source.
Common issues
AI summary feels too generic.
Apply a tighter filter or wait for more responses so the model has stronger evidence.
A chart looks different after filtering.
Check whether hidden filters or date ranges are still active.
FAQ
Where do NPS and CSAT product pages fit with analytics?
Use /features/nps for Net Promoter Score charts and KPIs, /features/csat for satisfaction ratings, means, and distributions, and /features/analytics-reporting for the broader results workspace.