Great products start with the user, not the feature.

There can be many reasons. Perhaps search engines are bringing in different visitors than before, or it’s just normal seasonal fluctuation. Analytics, heatmaps, session recordings and A/B tests show you where users drop off and which variant converts better. Why users drop off, they don’t show. And not every drop-off is a lost customer: many people go far into a process and then stop because they want to look at it again, discuss it or think it over.

Only qualitative methods reveal the reasons. That might be a drop-off analysis, i.e. short surveys exactly where users leave. Moderated or unmoderated usability tests work just as well. Unmoderated tests also let you reach larger samples: participants complete realistic tasks such as “Buy product XY”, go almost all the way to checkout and think aloud as they do. You see what they notice, what they overlook and how your copy lands. On most websites, the basics of conversion optimisation have long been in place. The decisive hurdles often lie in the details, and they only become visible when you watch real people.

Yes. With a live product, talking to real users shows where things get stuck. In the discovery and ideation phase, it shows what’s still missing. Both are often surprising and almost always valuable.

Market research asks how a product is received in the market: who buys, why, at what price? Its typical tool is the survey. UX research looks more closely, namely at how people actually use your product. I observe what users do, not just what they say, because the two often differ. The result is concrete pointers on what your team can improve.

Yes, at any stage: simple scribbles, paper prototypes, clicked-together slides, designs or interactive prototypes. With AI, even elaborate prototypes can now be built very quickly, so interactions can be tested early on.

That’s exactly why a step before is worth it. Because building has become so easy, things quickly get made that nobody needs. In an early discovery phase, I therefore talk to your users first: what do they really need, and is there a use case for your idea at all? Only once that’s clear is an MVP worthwhile, and we then test it specifically with prototypes.

Jakob Nielsen’s rule of thumb says that five to six test users uncover around 80 percent of usability problems. However, that applies per target group. Most websites and apps have several user groups. With two large groups, you already need 10 to 12 participants to reach the same level.

Then there’s complexity: is it a simple website or a long, multi-step process? Besides the main purpose, such as placing an order, there are often other usage scenarios that need covering too. On a complex website, six people won’t find 80 percent of the problems. The rule is a good starting point, but I define the right sample with you based on your target groups and use cases.

I still offer expert evaluations, guideline checks and competitor analyses. But AI now handles much of this quickly and well, right up to agents that regularly check your website or monitor competitors. My time is best invested where AI can’t go: in conversation with real people.

Then I’ll tell you anyway. I’m happy to phrase criticism diplomatically, but I don’t sugar-coat. When your success and your users’ satisfaction are at stake, you have to speak plainly, even if the bearer of bad news isn’t always welcome. My clients appreciate that. It’s exactly what research is for: finding out what isn’t working well yet.

At short notice. Just write to me, it’s straightforward. In a first conversation, we’ll clarify what you need and where the greatest need lies.

Little during fieldwork, but definitely some at the start. When I work out the research questions and hypotheses, I need a contact person to answer my questions. Otherwise you end up with results that miss what you actually wanted to know.

In B2B, often from your own customer base, provided consent has been given. In B2C, via recruitment agencies. There I take particular care not to invite seasoned “professional testers” who skew the results. What matters is defining clearly in advance who we want to talk to.

Yes, always. For moderated remote tests I plan for it from the start, and it’s of course also possible for lab tests, even though these are less common today.

Happily your own, such as Teams or Zoom. Participants share their screen, and you can see what they’re doing at all times, including in apps. For tests with broad target groups, I use specialised platforms such as Rapid User Tests. They also handle recruitment, and if you wish, your own customers from your database can be included. The automated analyses on these platforms have become quite good. For robust results, however, you need an experienced researcher who moderates the sessions or reviews the recordings, because the most important problems often only show up on close inspection.

We decide that together with your team, depending on how you want to work with the results. Options include:

• Presentation: the classic slide deck for the results meeting.
• Written report: for example a Word document that the team reads together in the meeting and then discusses.
• Results workshop: together with your team, with a Miro board if you like.
• Video clips: anonymised excerpts if you want to hear your users in their own words.
• Research dashboard: a permanently usable interface where you can dig into subtopics, search the transcripts and ask the data set new questions later on.

I’m happy to show examples and case studies, anonymised on request.

No. I’m purely a UX researcher and don’t design or develop websites or apps. Instead, I give your design and development team a solid basis for their decisions. I’m happy to help build prototypes for testing, but I don’t create designs in Figma or similar tools.

I work at a fixed hourly rate, not with flat-rate packages. That way, you pay exactly for the effort your question requires. A small start is always possible.

That depends on the question, method and sample, and it can be quick: we’ve done 100 interviews in one week. Many clients also book me as a permanent research partner who handles requests on an ongoing basis, including in busy periods such as before Christmas or around a product launch.

Honestly: if management is fundamentally not open to user research, it will be difficult. Often, though, all that’s missing is a convincing first piece of evidence. A small start is enough, such as a quick usability test at a critical point. Nothing convinces decision-makers like short video clips of real customers struggling with their product. And a list of quick wins shows that the effort pays off quickly.

Remotely and on site in and around Munich, in German and English. International agencies also regularly commission me for studies in other countries.

A non-disclosure agreement is standard in my projects. I handle recordings and personal data in compliance with the GDPR:

• All participants are informed in advance and give written consent.
• Recordings, transcripts and data are stored on servers in the EU.
• Reports and transcripts contain no real names.
• Only project members have access to the raw data.
• Recordings are deleted at the end of the project or after an agreed period.
• On request, we conclude a data processing agreement (DPA).