Customer feedback surveys

How to design, deploy and leverage surveys that drive real improvements in your business

9 min

Feedback surveys are one of the most direct ways to understand your customers’ perception. Well designed, they reveal friction points, improvement opportunities and satisfaction drivers that behavioural data alone can’t capture.

However, most surveys fail due to poor design, wrong timing or lack of action on results. This guide covers the full cycle: from choosing the right metric to turning feedback into tangible changes.

NPS, CSAT and CES: what each one measures

The three most widely used feedback metrics measure different dimensions of the customer experience. Choosing the right one depends on what you want to know and at which point of the customer journey you collect it.

  • NPS (Net Promoter Score): measures overall loyalty by asking likelihood to recommend. Scale 0-10, classifying promoters (9-10), passives (7-8) and detractors (0-6)
  • CSAT (Customer Satisfaction Score): measures satisfaction at a specific moment. Ideal after a purchase, support interaction or feature use
  • CES (Customer Effort Score): measures the perceived effort to resolve something. Lower effort correlates with higher retention

How to design effective surveys

Survey design determines response quality. Ambiguous questions, confusing scales or overly long surveys generate drop-offs and unreliable data.

An effective feedback survey has 3 to 7 questions, combines a quantitative metric (the score) with at least one open-ended question (the why), and adapts to the user’s context.

  • Always start with the main question (NPS, CSAT or CES) to maximise response rate
  • Include an immediate open-ended follow-up: "What could we improve?" captures qualitative insights
  • Avoid double-barrelled questions or biased wording that influences the response
  • Adapt language to the user profile: avoid technical jargon with end customers

When to send surveys

Timing determines both response rate and feedback relevance. Sending an NPS survey to a user who registered two days ago provides little useful information; doing so after three months of active use does.

  • Post-purchase: 24-48 hours later to capture the fresh experience without being intrusive
  • Post-support: immediately after closing a ticket to measure resolution
  • Relational NPS: every 90 days to active customers to track loyalty trends
  • In-app or in-page: contextual surveys triggered after key actions (completing checkout, using a new feature)

How to analyse results

Survey analysis goes beyond calculating averages. Segment results by customer type, channel, product or period to find actionable patterns.

Open-ended responses are a goldmine of qualitative information. Techniques like sentiment analysis, thematic categorisation and word clouds help extract the most recurring and priority themes.

  • Segment by cohort: do new customers face different issues than returning ones?
  • Cross-reference NPS with behavioural data: do detractors show a distinct usage pattern?
  • Categorise open responses by theme: product, pricing, support, experience

From feedback to action plans

Feedback without action erodes customer trust. If you ask for opinions and change nothing, they won’t respond next time. Each survey round should produce at least one concrete, visible action.

Create a closed-loop process: detractors receive personalised follow-up, the most mentioned themes get prioritised in the product backlog, and implemented improvements are communicated to the customers who suggested them.

Recommended tools

Tool selection depends on your volume, channels and level of integration with your existing stack. From simple solutions to enterprise platforms with advanced automation.

  • Typeform: conversational surveys with high response rates and excellent design
  • Hotjar: in-page surveys integrated with heatmaps and session recordings
  • Delighted: specialised in NPS, CSAT and CES with send automation and closed-loop workflows
  • Qualtrics: enterprise platform with advanced analysis, segmentation and workflows

Key Takeaways

  • NPS measures overall loyalty, CSAT point-in-time satisfaction and CES perceived effort
  • A good survey has 3-7 questions combining a score with an open-ended question
  • Send timing directly affects response rate and feedback quality
  • Segment results by cohort, channel and product to find patterns
  • Feedback without action erodes trust: always close the loop

Want to implement a customer feedback programme?

We design your survey strategy, choose the right tools and help you turn feedback into measurable improvements.