Case study · GetYourGuide
Setting a six-month design direction for the segment our previous redesign had deliberately left behind. Case study about using strategy to create executive trust and drive experimentation.
At GetYourGuide, we had recently completed a redesign of our homepage. Our next objective was to develop a homepage tailored for returning users.
The team faced challenges in formulating effective hypotheses and a clear vision for the next six months. As the UX Manager overseeing the responsible teams, I collaborated with our Product team to address this issue.
We established a sound 6-month product strategy with the team, got buy-in from our executive team, and used this to scope an MVP. Based on this, we conducted many A/B tests to create a “baseline” new UX for returning users.
Given the tight deadlines, we scoped 2 different use cases for a 3-day workshop, based on the size and impact of the target customer segment.
My role was to work with our Director of Product to define scope, prepare the workshop, and facilitate the event.
We answered 3 different questions:
The homepage design was a modular structure that adapts to the context and primary user needs of the returning user segment.
When a user was identified as a returning user, they would see a "for you" tab alongside other brand categories. We kept UX consistency between this tab and the other ones.
We identified two segments, each with distinct moments, intents, and measures of success along the travel journey.
Customers who have an upcoming booking with GetYourGuide, until the end of their trip (7 days after their activity date).
Goal for GetYourGuide: help these customers find additional activities for their upcoming trip, measured by conversion rate.
Customers who have no upcoming booking and have been on an activity between 8 days and 180 days ago.
Goal for GetYourGuide: increase engagement frequency of these customers to increase probability of future bookings.
For each of the segments, we defined user scenarios and needs, and wireframed modules for these specific needs. The modules would be turned on/off and reordered based on the prioritization of the needs. The engineering manager assisted us in understanding the technical feasibility of our plans, which helped break down the vision into smaller, actionable components.
Final design for MVP
We presented the plan to the CPO, then the CEO, CMO, and CPO to gain trust in the direction and secure approval.
The direction was positively received. We had discussions about scope, expectation setting (direction vs MVP & hypotheses testing), and agreed on monthly progress check-ins.
Decisions made for the MVP mainly relied on the PM and the EM, based on impact and effort. For example, we decided to descope complex personalization algorithms in some cases to be able to launch and learn quickly.
My role in this stage was to work with our content and product designer to get the high-fidelity designs ready and give feedback on the MVP.
Results in market
experiments on inspirational content. Engagement was there — but not enough operational justification. We eventually removed the module.
"Jump back in" history sections drove the largest metric uplift across every returning segment we tested.
Bounce rate dropped for next-trip users. Booked-activity modules engaged same-trip users but hurt conversion — so we pulled them too.
The workshop, while only taking 3 days, allowed us to formulate a sound strategy that we could confidently communicate to our executives. This gave us time and space to launch experiments and learn iteratively with exec buy-in.
During the initial strategy creation, the engineering manager helped us understand our capabilities. This both made the strategy feasible, but also allowed the team to get a head start on engineering. The workshop also allowed us to create context and bond as a team.
We assumed different visitor segments had distinct needs, but similar modules performed well across all of them. A leaner approach would have been to start with 1–2 similar experiences and diverge through testing; we did the opposite.
“Weekend getaways” was grounded in research showing travelers take shorter, more frequent trips. But making it successful requires a broader company strategy: different inventory and a repositioning of GetYourGuide as a go-to resource for this type of travel. So we omitted this as a module.