Ralph Hinderberger
Senior Growth & UX Strategy Leader
Think Growth, End-to-End.
From AI visibility to measurable conversion: I bring design thinking into the AI era, agent-based processes included.
Experimentation · User Experience · Conceptual Design · GEO & SEO · Agent workflows
For more than 25 years I have been responsible for data-driven growth in e-commerce and digital products. I used user research to build experience quality along the customer journey and turned it into measurable conversion through experimentation programs. Today I use the same research for visibility in search and in AI systems. Both start from real user needs, and AI-assisted workflows make the whole thing scale.
Right now, links are giving way to AI answers.
Open to new roles · Based in Velbert, Western Germany · Open to moving back to the southern Rhineland
Approach
Growth is built along a chain. Silos break it.
Visibility without experience quality burns through traffic that is already shrinking. Experience quality without measurement is just opinion. And a test program that runs on gut feeling, with no rules for prioritizing ideas, ends up as a pile of one-off cases with random winners.
That is why I work in a chain of six stages: Understand, Enable, Materialize, Test, Convert, Monitor. Each stage loosely follows the field-tested design thinking approach and has its own methods. What makes it pay off is a well-planned handover to the next stage. Every experiment raises new questions, so after "Monitor" the cycle starts again at "Understand". I call this loop Continuous Growth Optimization (CGO), and every round is its next iteration.
01 · Understand
Building a solid foundation
Each CGO cycle starts with a deeper understanding of user needs: goals, tasks, desired outcomes and the emotional states connected with them give a detailed picture of an experience that reliably satisfies the target audiences. If these audiences have yet to be defined, I use proven sociological segmentation models to bring them into focus.
Jobs to be Done, user research, feedback and sentiment analysis, web analytics and a look at the competition show where growth potential is still untapped and which KPIs it can move. The ideas that come out of this are then ranked by potential, importance and effort, using a modified PIE approach. From that point on, the roadmap is backed by evidence.
Methods
02 · Enable
Laying the technical and organizational groundwork
Optimization only works once it is possible. That takes clean on-site code, a suitable setup for testing and personalization, and cross-functional communication. Together they provide reliable data, visibility in search and AI systems, and stakeholders who are able to make sound decisions. It also means equipping teams with AI-assisted workflows that add speed and keep quality intact.
Methods
03 · Materialize
Turning insights into design
Findings become tangible: information architecture, wireframes, prototypes, content and cognitive UI design.
This is where a good analysis either becomes a good digital experience or falls short, from the category page through PLP and PDP to checkout. AI-assisted code generation cuts the time from idea to clickable prototype to hours instead of weeks.
Methods
04 · Test
Measuring, not guessing
Either qualitative in a usability test (lab or remote), quantitative in an A/B test, or both combined: every variant runs against the original. Research and prioritization come first, in a process I have refined over decades. As a result, the win rate in the programs I led most recently was 30 to 40 percent at a probability to be best (P2BB) of at least 90 percent. In smaller shops with little traffic, it was 60 to 70 percent at a P2BB of 75 percent or more. And every losing test is a bad investment that never went live.
Methods
05 · Convert
Making winners permanent
Overall winners are rolled out to everyone. Segment winners, and variants that showed a consistent but not yet conclusive advantage over a long test period, go live as personalizations, monitored against a holdout group: only the audiences that preferred them see them, while everyone else keeps the default. This is an additional lever that sometimes gets lost in the overall results, and it is the reason effects add up instead of canceling each other out.
Methods
06 · Monitor
Proving the impact
After the rollout comes the proof: dashboards, follow-up measurement of the campaigns and AI citation performance, documentation. This keeps it clear what works, what wears off and where the next cycle should start. Over time, individual wins add up to a program.
Methods
Services
Where I can help right now.
Most projects start with one of five situations. Pick the one that sounds familiar.
When ad costs rise and conversion stalls
More revenue from the traffic you already pay for
Every month the clicks get more expensive, while the share of visitors who buy stays flat. I find out with data where people drop out of your shop or funnel, rank the fixes by impact and effort, and prove which changes actually pay off.
-
Typical situation
The monthly report shows higher cost per click or acquisition and the same or lower conversion rate. Management asks why more budget no longer brings more revenue, and the team keeps shifting money between campaigns instead of improving the shop itself.
-
What you get
The biggest drop-off points, backed by analytics, session data and user feedback. A ranked list of measures with expected effect and effort that your team can implement. A test plan that shows which changes work before they go live for everyone, and a short summary you can take to management.
-
Evidence
In programs I led: more than 200 A/B tests in three years, more than 30 percent direct revenue uplift and more than 10 percent more purchases per user, each measured for the tested variant against the original.
When your shop keeps changing, or a relaunch is coming
Change your shop without losing what works
Most shops no longer relaunch in one go; they change all the time. Every release can quietly cost revenue, and when the numbers slip, nobody can say which change it was. I put a safety net under every change, so losses show up early and get fixed fast. The same applies if you do go for a full relaunch.
-
Typical situation
Conversion has been slipping for weeks, several releases have gone live in the meantime, and the monthly report is the only early warning. Or the date for a relaunch or platform switch is set, the agency delivers technology and design, but nobody checks independently whether the new shop sells at least as well as the old one.
-
What you get
Measure
Guardrail metrics and a baseline for your shop, such as conversion rate, revenue per session, drop-off rate and tracking completeness. Every release gets a before-and-after check, with an alert when a value falls below its threshold. If the numbers have already dropped, I trace the loss back to individual releases.
Observe
Session recordings, heatmaps and journey analyses in tools such as Contentsquare, Hotjar or Mouseflow show where users now hesitate or drop out.
Listen
Online surveys and feedback portals capture what customers say about the change in their own words. Sentiment analysis of that feedback shows how they feel about it.
Fix
With tools such as Contentsquare and zenloop, a problem that suddenly appears on site triggers an alert and goes to IT as a ticket, together with the session data behind it. You see what broke and why, and I contribute proposals on how to fix it. A developer looks for a technical fix, a designer for a visual one. I work between both and suggest solutions that hold up for each: on-brand, technically feasible and easy for users to grasp.
Relaunch
User tests of the key journeys before and after go-live, and close monitoring in the first weeks. The technology stays with your agency or IT team.
-
Evidence
Overall functional responsibility for an online shop relaunch, from concept through go-live and follow-up optimization. The lessons from that project shape how I protect shops through change today.
When testing delivers little or the lead is missing
A test program that produces results again
You have paid for a testing tool and a team, but there are few results you can show. Or the person who ran the program has left. I get the program back on track so it delivers measurable effects on a regular basis, and I hand it over in a way that keeps it running.
-
Typical situation
Tests are chosen by whoever argues loudest, many end without a clear result, and the tool is increasingly used to roll out features. Management starts asking what the investment is actually bringing.
-
What you get
A clear prioritization process for test ideas, sound statistical evaluation and results translated into revenue impact. As an interim lead or alongside your team, with documentation and handover from the start, so the knowledge stays in-house.
-
Evidence
Built and ran an experimentation and personalization program in-house. Win rate of 30 to 40 percent at a probability to be best of at least 90 percent.
When organic traffic drops despite stable rankings
Visibility in AI answers
More and more customers get their answers from ChatGPT, Perplexity or Google's AI Overviews and never click through. I show you whether and how your brand and products appear there, and what you can do to be named when people are deciding what to buy.
-
Typical situation
Rankings look fine, but organic clicks keep falling. Nobody can say how often your brand comes up in AI answers, and the usual reaction is to keep publishing blog posts and wait for things to recover.
-
What you get
A repeatable measurement of how often and in what context AI systems name your brand for relevant buying questions. A plan for content, product data and structured markup that makes you easier to cite. Reporting that links this visibility to traffic and conversion, so it can be tracked internally.
-
Evidence
A young discipline, so I say it plainly: my current reference is ongoing GEO work for an online marketing agency (pro bono). The measurement method is the same one I would use for you.
When your agency's clients want more than traffic
Senior CRO and GEO capacity for your agency
Your clients ask why the traffic you deliver doesn't turn into more revenue, or why they don't show up in AI answers. A senior hire for that doesn't pay off yet. I step in on a white-label basis with everything on this page, and you decide what your client needs.
-
Typical situation
A key client is unhappy with conversion or with its visibility in AI answers and starts talking to full-service agencies. Your team is strong on campaigns and SEO, but nobody has the senior experience to run research, testing or GEO at the depth the client expects.
-
What you get
Scope
Everything on this page, from conversion analysis and release safeguards to test programs and AI visibility. You decide what your client needs, project by project.
Visibility
Behind the scenes, as part of your team under your brand, or openly as your specialist. Your needs decide how I appear.
Loyalty
Your client stays your client. If a client ever wants to work with me directly, that only happens with your consent.
-
Evidence
I know both sides of this setup: as managing director of two UX agencies, and on the client side, where I steered agencies as the manager of an online shop. I currently support an online marketing agency with its own visibility in AI systems (pro bono).
Expertise
Six fields in depth.
The six stages above show how I work. The six fields below show what I bring to them: approach, tools, references and the metrics I am willing to be measured by.
-
AI visibility & GEO
Applies to stages: Enable · Monitor
Rules that held for years are now changing, fast and in a radical way. Classic search engines make users search; AI answers make them find what they need. Classic SEO is the baseline. I add the layer that turns AI visibility into business results.
Approach
Citation analysis: for which questions does ChatGPT, Perplexity or Google AI mention a company today, and for which does it not? Next, I build clearly assigned entities, verifiable statements drawn from reviews and mentions, structured markup, and topic clusters organized around Jobs to be Done rather than keyword lists.
Tools
Prompt sets for repeatable citation tests across several AI systems, sentiment and review analysis, schema validation, agent-based research workflows.
References
Ongoing GEO work for an online marketing agency, based on JTBD (pro bono).
Metrics
Citation rate per question, share of correctly assigned entities, sentiment of the sources, referral traffic from AI systems.
-
SEO & content structure
Applies to stages: Enable · Monitor
Classic search is still the largest single channel. It is also what allows AI systems to find content in the first place.
Approach
Technical audit of crawling, indexing and page load; a sound information architecture and internal linking; content with real substance instead of sheer volume of text and images; priorities set by revenue contribution, not by search volume.
Tools
Crawlers and indexing data, Search Console, web analytics, ranking and log data, AI-assisted content briefs.
References
Online shop and portal projects in e-commerce, as well as directory and service portals.
Metrics
Organic revenue contribution, visibility per topic cluster, index coverage, landing page entry quality.
-
Experience & UX research
Applies to stages: Understand · Materialize · Test
Qualitative methods explain the why, quantitative methods prove the how much. Together they turn opinions into decisions.
Approach
Eye tracking, usability labs, moderated and unmoderated remote tests, online surveys, persona and JTBD work. The findings feed into concepts, wireframes and prototypes that go straight into development.
Tools
Eye-tracking setups, remote testing platforms, survey tools, AI-assisted prototyping.
References
More than 100 usability lab tests in eight countries and 1,020 participants in remote studies, for companies in telecommunications, software and IT security.
Metrics
Task success, time on task, drop-off points in the journey, standardized satisfaction scores.
-
Conversion optimization & experimentation
Applies to stages: Understand · Test
A test program only deserves the name when it runs on a clearly defined process.
Approach
Hypotheses built from research and analytics, every idea rated on a PIE scorecard, sample size planned before launch, results analyzed by probability to be best and by segment, and the losing tests documented as well.
Tools
Experimentation platform with multivariate and personalization features, web analytics, session recording, statistical evaluation.
References
More than 200 A/B tests in three years in programs I led, plus a shop relaunch where I held overall functional responsibility.
Metrics
Win rate of 30 to 40 percent at a P2BB of at least 90 percent, more than 30 percent direct revenue uplift, more than 10 percent more purchases per user.
-
Personalization & segmentation
Applies to stages: Convert · Monitor
Many tests win for particular audience segments rather than for everyone, and that is where the lever is. Overall winners are rare, so an experimentation program pays for itself through the sum of its partial wins.
Approach
A segment model based on behavior, traffic source and need; segment analysis for every single test; segment winners turned into permanent rules; ongoing checks against the default.
Tools
Personalization and multivariate features of the experimentation platform, segment and audience data from analytics and CRM.
References
Set up the personalization process inside a running experimentation program.
Metrics
Uplift per segment, share of personalized sessions, cumulative effect compared with the original.
-
AI-assisted workflows & enablement
Applies to stages: Enable, and across all stages
Speed must not cost substance. That is why my workflows have quality gates built in, and nobody gets to skip them out of enthusiasm.
Approach
Agent and prompt systems for research, analysis, content production and reporting, with defined quality gates and a model for roles and approvals. On top of that, team training, so everyone knows where a model can be trusted and where it can't.
Tools
Multi-agent setups, prompt libraries, automation with n8n and Zapier, choosing the model to fit the task.
References
My own agent and skill systems for research, analysis and content; AI-assisted content creation for the newsletter of my local Protestant parish (pro bono).
Metrics
Turnaround time per content or analysis unit, error and correction rate, share of team members who work independently.
Results
Expertise in numbers.
25+
Years of hands-on work in UX, e-commerce and growth
200+
A/B tests in three years
1,020
Participants in remote studies
100+
Usability lab tests in eight countries
30–40%
Win rate in the test program at 90%+ P2BB
30%+
Direct revenue uplift, online shop
10%+
Purchases per user, online shop
1
Shop relaunch with overall functional responsibility
All figures come from projects I was responsible for, and I can walk you through how each one was derived. Uplifts refer to the tested variant compared with the original.
Career
Customer focus in e-commerce, in practice since 1995.
MEDION, Essen2017–2026
Senior UX Consultant E-Commerce · Growth & Conversion Strategy
Joined as Online Shop Manager, User Experience // Led data-driven shop optimization // Overall functional responsibility for the 2023 shop relaunch // Built and ran the experimentation and personalization program // Lead of the UX consultants // Quantitative UX research lead // Digital concept lead.
Usability People International GmbH, Cologne2011–2017
Managing Director
International UX project management // Qualitative UX research lead // Key account management for international clients // Head of international sales.
UX Management GmbH, Düsseldorf2007–2011
Managing Director
UX project management in Germany // Qualitative UX research lead // Key account management for German clients // Head of sales, Germany.
From 2003 to 2007 I served on the board of the German UPA, the German chapter of the Usability Professionals’ Association. I also co-wrote the book „Usability praktisch umsetzen“ (Hanser, 2003), contributing the chapter „Usability als Investition“. Details on the years 1995 to 2007 and on my education are available on request.
Pro bono
Two projects where I pass on my methods.
Together with an online marketing agency run by friends, I am improving the agency's own visibility in AI systems, working from Jobs to be Done instead of keyword lists. The aim is for the agency to be picked up as a source when its B2B clients ask their questions.
I also help edit the newsletter of my local Protestant parish: content, events, and setting up AI-assisted content creation for a team that has plenty of passion and faith, but works as volunteers with little time to spare. It gives me a chance to put my experience to good use close to home.
Skills
Three clusters instead of a word cloud.
01 · Growth & Conversion
02 · Experience & Research
03 · AI, Data & Tools
Ways of working
Permanent position, interim assignment or project ownership: I have worked in all three and am glad to continue in any of them.
There is also my coaching practice, which I have built up over decades alongside my digital projects and am still developing. It helps me mentor younger colleagues and improves self-reflection, for me and for the people I work with. Through energy work it keeps my body and mind in balance, which makes me more resilient under stress. In times of deep and far-reaching change like the present, that is a great help in coping with the effects.
Clients
Employers and clients from projects I was responsible for.
About me
I translate between users, data and management.
A good online business works like your favorite café. They recognize you, you find your seat without looking for it, you get what you want without having to explain, and so you keep coming back. Online, that translates into relevance, clear orientation and an experience without friction. I measure these things instead of simply claiming them.
In projects I play three roles at once. I understand users and have the expertise to validate assumptions about them reliably. I analyze the data, because opinion without numbers doesn't hold up. And I translate for management, because an insight only has an effect once it lands as a business case.
I am currently based in Velbert, on the edge of the Ruhr area, and work across the Rhineland, the Ruhr area and remotely. I am open to moving back to the southern Rhineland, to the Cologne/Bonn area. A role or a longer project there would be a good reason to make that move soon. I recharge through exercise, gardening and developing my own coaching formats, which, to be honest, also makes me more efficient and effective at work.
Profile
Senior Growth & UX Strategy Leader, end to end
Focus markets
E-commerce, digital products, SaaS
Ways of working
Permanent, interim, project ownership
Base
Velbert, working across the Rhineland, the Ruhr area and remotely. Open to moving back to the southern Rhineland (Cologne/Bonn area); a role there would speed it up.
Languages
German (native); English, fluent; Spanish, fluent in conversation, written with some limitations
Contact
Let's talk about your growth problem.
Tell me where things are stuck: visibility, conversion, a test program that won't get going. I reply within one business day. And if I'm not the right person for the job, I'll tell you so.
Form
Deutsch