Hurry up⚠️ Only 20 tickets left in this poolWaysAwardsSubmit your WaysAwards project — enter the 2026 nominations
UX Research

The Voice of the Customer - about the shortcomings of declarative research in product discovery

September 5, 2023 9:00 AM
Long Lecture
🇬🇧 English
Wisła

About

Voice of the Customer - easy to get and analyze, hard to conclude from. Organizations have fallen in love with automatic techniques for tracking feedback from their customers. They eagerly reach for reviews, opinions, helpline reports and ratings on scales. It gives them a conviction that their organization is customer-centric. However, it is not the most accurate source of users' needs in product discovery. In the presentation, I will focus on what are the limitations of declarative methods and what are their alternatives.

Talk by

Photo of Iga Mościchowska, UX consultant and trainer at WITFLOW & [ation] center
Iga Mościchowska
UX consultant and trainer at WITFLOW & [ation] center

Watch the full talk

Watch this WaysConf session, then continue with related talks or explore the current programme.

Explore the 2026 agendaMore WaysConf recordings
From the recording

Talk in brief

A critical guide to using voice-of-customer data without mistaking declarations for complete evidence. Reviews, surveys, support messages, and satisfaction scores are easy to collect continuously, but they overrepresent extreme reactions, attract a small and self-selecting group, and reveal only what people can consciously describe. The talk contrasts this explicit knowledge with behavioral observation and generative research. Product analytics and usability tests show what people do, while creative tasks, cultural probes, and co-design can surface tacit needs and unexpected ideas. Declarative feedback remains useful for monitoring and exploration, but stronger discovery combines it with methods that validate behavior and deepen understanding.

Key takeaways

  1. 01

    Organize customer feedback by source and purpose

    Distinguishing who initiated the feedback and whether it describes context or evaluates a product clarifies what each source can support.

    Watch from 4:53
  2. 02

    Do not equate statements with behavior

    Reviews and recommendations can contradict actual usage, so product decisions need evidence of what customers really do.

    Watch from 9:16
  3. 03

    Expect self-selection and extreme opinions

    People with strong positive or negative reactions are more likely to comment, while moderate experiences and silent users remain underrepresented.

    Watch from 10:53
  4. 04

    Observe needs customers cannot articulate

    People rarely request alternatives outside their experience, making observation and product behavior essential for identifying hidden friction.

    Watch from 16:07
  5. 05

    Use generative tasks to uncover tacit knowledge

    Creative activities help participants discover motivations, preferences, and possibilities that direct questions may never reveal.

    Watch from 18:42
Read edited transcript highlights

These concise notes were edited from automatic captions and checked against the talk structure. They are not a verbatim transcript.

Voice-of-customer data is accessible but partial

Support conversations, surveys, reviews, social posts, and satisfaction measures can be gathered continuously and compared over time. That makes them valuable for monitoring a product and spotting possible changes. Their convenience can also create false confidence. Each source reflects a particular trigger, audience, and question, so teams should identify what generated the feedback before treating it as representative evidence.

Watch from 4:53

A recommendation may conceal the actual pattern of use

Someone can publicly reject a game after spending hundreds or thousands of hours with it. The opinion and the behavior answer different questions: one expresses a current judgment, while the other records sustained engagement. If a team reads only the statement or only the usage count, it can reach opposite conclusions. Product discovery needs both the declared meaning and the observed action.

Watch from 9:16

Survey scale can create an illusion of representativeness

A message may reach a very large customer base while only a small fraction begin the survey, fewer complete it, and still fewer provide a substantive open response. The resulting answers can be useful, but they represent people willing and able to participate in that format. A large invitation count must not be presented as if the entire population had supplied comparable evidence.

Watch from 13:25

Observation reveals problems outside the customer vocabulary

People adapt to inefficient routines and may not recognize that a task could work differently. Specialists can suggest local improvements because they know the existing process well, yet the same expertise may anchor them to it. Watching real behavior helps researchers identify workarounds, delays, and assumptions that no participant would think to list as a product requirement.

Watch from 16:07

Creative tasks make hidden understanding discussable

Generative research asks participants to make, sort, retell, imagine, or respond to carefully designed prompts. During that activity, people can notice their own motivations and express relationships that direct interviewing did not reach. The output is not valuable because it is artistic; it gives participants another route into tacit knowledge and gives the product team new material for responsible interpretation.

Watch from 18:42
Explore WaysConf 2026

View the agenda

Get tickets