Trivago
Deal Hunter
A mobile web app and notification system that curates unique hotel experiences
Deal Hunter was an experiment I developed in the Trivago Innovation Lab that aimed to explore new approaches for boosting organic traffic to trivago.com. The objectives were to reduce reliance on acquisition marketing and paid traffic while enhancing the value provided to travellers beyond the metasearch service.
The MVP
My role and
competencies:
I played a central role in the project, starting with assistance to the board of directors in selecting a key business challenge for the lab. My responsibilities encompassed framing the problem space, establishing the product vision and goals, conducting extensive research, leading the procurement process for innovation partners, and collaborating closely with Dentsu's ECD on design direction and creative direction.
Chapter 1:
Problem
statement
Our challenge is to create a product and strategy that not only significantly increases the long-term organic traffic to trivago.com but also encourages travellers to embark on shorter trips more frequently, resulting in increased bookings for hotel nights. The solution must reduce trivago's reliance on paid marketing and SEO, creating a self-sustaining flywheel effect for increased website traffic and click-through rates, ultimately contributing to the reduction of marketing expenditures.
Marketing channels performance according to Trivago's ICPs: 'Flexible Deal Hunter' offered the biggest impact opportunity
Our hypothetical flywheel outcome
Trivago spent 20% more than the industry's baseline on Search Engine Marketing (SEM) and obtained 25% less penetration according to Digital Experience Analytics
The hypothesis
We believe that we can create a flywheel effect for web traffic if we design a mechanism to inspire an increasing number of people to search for hotel accommodation more frequently.
We will know that to be true when marketing expenditure remains constant while traffic growth to trivago.com is sustained over time.
Chapter 2:
Roadmap & WoW
We overdelivered the project as a final product within the same timeframe
Ways of working, method, and rationale
I employed a customised 'Discover, Define, Develop, Deliver' framework to suit our 12-week timeframe for release and testing. The collaborative cross-functional approach, involving the Lab, Dentsu, and 3rd party contractors, enabled a continuous flow of information and data-informed decisions. The rituals and cadence that I implemented also helped maintain velocity throughout the process.
Chapter 3:
My research
Picking our Ideal Customer Profile
Using circumstance-based segmentation, I identified our potential target audience by reviewing studies from various Trivago verticals, collecting third-party reports, and analysing trend forecasting articles and macro-trends. This approach allowed me to define the 'Flexible Deal Hunter' profile, and two sub-categories: 'secure planners' and 'mind travellers,' which proved crucial during the ideation phase.
User journeys: framing pains and gains
I analysed UX and marketing documentation to identify common patterns, pain points, goals, and frames of mind in the user journey. This served as a foundation for improving the overall experience.
I synthesised a user journey model based on the identified patterns and aligned it with the three domains outlined in our strategy (Growth Hacking, Digital Tools, and Content Strategies). This approach helped prioritise areas of opportunity, creating a value-based experience that aligned with users' goals and jobs-to-be-done.
Choosing the right domain to operate
I also conducted desk research to identify areas of opportunity for sustained growth, which led to a focus on the 'Organic Growth' domain. Simultaneously, the engineering team explored various technology stacks within this domain. Using the space-saturation technique, I mapped out the organic growth marketing domain, extracted common themes, and identified key attributes to inform decisions and set success metrics.
Chapter 4:
Ideation sessions
1-Day design sprint
In a one-day workshop with cross-disciplinary participants, we refined the problem space, using techniques like space saturation, clustering, and HMW statements to expand possibilities while focusing on the business challenge.
Next, we ran Crazy 8s and generated numerous sketch ideas, leading to eight product concepts. The team further developed these concepts and, in the following sessions, evaluated them against strategy and business objectives, resulting in four final viable product ideas.
Sprint workshops
-
Revisiting the Problem Space
-
Space saturation
-
HMW statements and filtering
-
Crazy 8s - Keep / Build / Kill
-
1 Pager sketches
-
Concept filtering
Sprint 1
Lo-Fi one-pager iterated concepts
Sprint 2
8x top one-pager concepts in higher definition
Sprint 3
Top 4x Hi-Fi concepts including strategy and success metrics
Chapter 5:
Winning concept
Trivago Deal Hunter (formerly Trivago Tracker) is a mobile-first web application that tracks deals in the best hotel experiences, and notifies users about unique yet affordable “staycations”
Product mechanics
Riskiest assumptions
We believe that
To test this, we will
And measure the
We'll succeed if
Chapter 6:
Validate & build
Keeping velocity during the building stage
To keep the design and technical development agile, we worked in parallel streams: user experience, user interface design, marketing and creative, back-end infrastructure, and interoperability with Trivago's platform, back-end, and CRM.
Along with the UX lead, I created a feature backlog, translated it into wireframes, UI designs, and interactive prototypes. We validated every decision through rapid usability testing (dogfooding), including branding, naming, terminology, and copy tonality.
Mobile web app and
email notifications
1
2
3
1 - Deal Hunter mobile web app
2 - Personalised email notification
3 - Trivago platform
Instagram test
campaign
Work-in-Progress
snapshot
Chapter 7:
Experiment
results
The mobile web app results
The results reveal both positive indicators and areas for improvement. On the positive side, engagement events within the app, like selecting a hotel destination type, show an impressive 80% user retention rate. Additionally, 50% of users subscribed to Deal Hunter email notifications upon viewing hotel deal prices. Furthermore, the average dwell time per session was a considerable median of 55 seconds.
Chapter 8:
Conclusions
In summary, the product displayed promising results in user conversion, aligning with our initial vision. However, the persistently high bounce rate on the home screen necessitated additional testing and optimization. Nonetheless, a key validation of the product's potential came when it was integrated into the main platform as a toggle feature once lockdown measures were eased.
Next steps
Social Campaign
-
Test other messages based on the 2 remaining categories (eco-friendly, historic)
-
Test different messages and creative relevant to each new geographic region where Hunter is launched
App -
Review landing screen to improve dropout bounce rate
-
Revisit headline and UX copy on interactive elements (buttons)
-
Refine UI designs further on filters screen (optimise elements in the viewport)
Notification emails -
Refine UI design and test iterations for optimum click outs to trivago.com
-
Highlight price/discount on fetched items