About:
Tropical Nomad is a coworking space in Bali built around one core promise: smooth, focused work. We mapped the end-to-end journey of digital nomads, validated friction points, and turned field insights into a Service Blueprint.
Category:
Service design
Year:
2025
Location:
Denpasar, Indonesia
Design framework:
Design thinking, service blueprinting
Research
Define
Ideate
Outcome
The problem:
Coworking experience depends on small moments. Fast onboarding, stable Wi-Fi across zones, clear rules for calls vs. quiet work, and an easy way to ask for help. In practice, members hit micro-frictions that break flow: Wi-Fi dead spots, noise, unclear zoning, limited call rooms. Passive feedback (a QR code on a wall) doesn't work unless people see real impact.
Goal:
Create a practical Service Blueprint based on real field data and prepare the next phase: a staff co-creation workshop to ideate solutions, define success metrics (ROI), and set up an ongoing discovery loop.
My role:
Service design intern and UX researcher responsible for end-to-end research, synthesis and service blueprinting. From field observation through stakeholder interviews to the final blueprint and workshop prep.
Methods:
•
Service Safari
•
Stakeholder mapping
•
Staff interviews
•
Member interviews
•
Survey (Typeform + QR test)
•
Review mining
•
Insights synthesis
•
Service Blueprint (Miro)
•
Workshop prep
•
Success metrics (ROI)
Discovery
I started in the field, not in slides. I observed how members arrive, set up, choose zones, and recover when something breaks. I then validated patterns through staff conversations, member interviews, a survey experiment, and review mining. The goal was to capture real service moments that affect deep work, comfort, and feedback behaviour.
/01
Service Safari
Before interviewing anyone, I mapped the service as it happens in real life: arrival, first-minute onboarding, Wi-Fi setup, zone selection, noise and calls, and how people ask for help. This created a grounded baseline and prevented me from designing solutions based on assumptions.
/02
Stakeholder mapping
Coworking is a service system, not a reception desk. Mapping stakeholders revealed who actually shapes the experience: Community Desk, Operations, Marketing, Security, Public Area staff, Maintenance, and F&B.
All service actors mapped (frontstage, backstage, user)
Decision makers vs active contributors vs support roles separated
Third parties captured (Google, comms channels, ISP, contractors)
/03
Stakeholder interviews
After mapping roles and influence, I ran short stakeholder interviews with Tropical Nomad staff to understand how the service works behind the scenes. The focus was on ownership: who handles onboarding, who resolves Wi-Fi and noise issues, how events are communicated, and where operational constraints appear in daily reality.
Role clarity. I documented what each role actually does in the customer journey. Community Desk vs Public Area vs Security vs Maintenance vs Ops.
Ownership mapping. For each recurring situation (setup, Wi-Fi, noise, calls, comfort), I captured who is responsible and who gets involved when it escalates.
Service reality check. I captured operational constraints and informal routines that aren't visible to members, but directly shape the experience.
Miro
/04
Survey and member interviews
I started with a simple idea: collect feedback through a QR survey placed on the internal wall. Typeform showed that people opened it, but almost nobody finished it. That moment was important. It reminded me that coworkings are deep-work environments, and "passive feedback" competes with calls, deadlines, and focus.
So I pivoted. Instead of waiting for responses, I approached members directly, ran short in-depth conversations, and in many cases filled the survey together with them. That shift immediately improved participation and gave me richer context behind the answers. Not just what people chose, but why.
"Listening properly, like even if they got out Google Translate. I've had to do that multiple times here. Sri and Pavel are great and super helpful, but the rest of the staff can be confused or dismissive."
Anonymous member
/05
AI powered sentiment analysis
Moderated research told me what happens inside the space. To see how Tropical Nomad is talked about outside it, I used AI models to analyse public reviews and online mentions across coworking directories, travel blogs and Reddit discussions.
I clustered recurring negative and neutral comments, filtered out external factors such as traffic, area crowding and weather, and benchmarked the rest against B Work and Zin Café. What remained were the most actionable, internally controllable friction points.
Combining both sources gave a comprehensive UX and operational audit: moderated research explains why something breaks, sentiment analysis shows how widely it is felt.
Define
After collecting evidence in the field and online, I distilled the work into a clear problem picture. This phase named what truly breaks the coworking experience for deep work nomads, what already works, and translated scattered observations into a shared service model the team can act on.
/01
Service Blueprint
To turn fragmented insights into a shared, actionable view, I created a Service Blueprint of the Tropical Nomad experience. From "looking for a coworking" to onboarding and daily deep work. The blueprint connects customer actions with frontstage staff actions, backstage operations, and third-party dependencies (Wi-Fi provider, comms channels, Google Reviews). It became the single source of truth for where breakdowns happen and who is involved at each moment.
End-to-end service view. What happens from first Google search, through onboarding and daily deep work, all the way to exit.
Frontstage and backstage clarity. What members see vs. what needs to happen internally to make it work.
Failure points identified. Wi-Fi, call room capacity, noise and maintenance, and communication channels.
Miro
/02
Onboarding and wayfinding
The current onboarding is incomplete and non intuitive. During the Safari I had to repeatedly ask staff for basic information: where the toilets are, how to book a Skype booth, which zones are for calls and which are quiet.
The same gap shows up across the whole stay. There are no clear rules on where members can eat, so people eat at work desks. Community information such as WhatsApp groups is discovered by chance on a physical notice board, because Nexodus offers no room booking or social features. Partner discounts arrive by email and stay there, and redeeming one means logging into email at the counter.
“A few times I didn’t know where to sit, and later I stumbled upon extra toilets by accident. I had to keep asking for everything, even how to book a booth.”
“I had no idea there was a community group chat until I saw it pinned on the board.”
/03
Wi-Fi, screens and heat
The basics of deep work are not reliable. 62% of respondents complained in writing about Wi-Fi drops, dead spots or forced manual resets caused by DHCP and IP conflicts.
Hot desks have no external screens, so the space is perceived more as a café than a professional workspace for complex projects. On the ground floor, high temperatures make it hard to work effectively, and in the non AC area reviewers mention cigarette smoke, humidity and dim light on cloudy days.
“Sometimes I have to reset my devices to reconnect to the network, even though I’m using the same device every time.”
“When I need to do deep work or complex multitasking, the lack of monitors at hot desks forces me to look for other spaces where a professional setup is the standard.”
/04
Calls and noise
Call booths are physically too tight, barely possible to enter, and members avoid them because of the heat and lack of fresh air. Calls then move into open work zones.
Staff do not consistently enforce quiet rules, especially with familiar long term members, which creates a double standard that frustrates paying professionals. Online reviews confirm the pattern: rooms advertised as quiet pick up noise from the neighbouring gym, and the open ground floor is full of calls at peak hours.
“In one room, one of the members was speaking at a very high volume, and the staff know him.”
“The quiet rooms still pick up gym noise, and the downstairs open area is full of calls. It’s hard to get true silence.”
/05
The pricing paradox
Pricing data revealed a logical flaw. The mid tier Power plan at 1.9M IDR includes 12 hours of booth credit, while the premium Unlimited plan at 2.9M IDR includes only 7.
At the same time, travel blogs and Canggu guides position Tropical Nomad at the higher end of the local price spectrum, next to cafés such as Zin that offer free coworking with consumption. For price sensitive nomads the membership has to clearly outperform free alternatives to feel worth it.
Reviews also point to a fragmented all in one promise: guests expect coliving on site, then discover it is a ride away, and evening focused workers miss a flexible night owl pass.
“It’s one of the nicest spaces in Canggu, but when Zin lets you sit for free, the membership here feels expensive.”
/06
What already works
Not everything was friction. Members explicitly praised the core team in open text feedback. The Community Desk acts as an essential community anchor and makes members feel genuinely welcomed and supported.
The environment is a strong initial hook: 7 out of 8 members rated the lighting and visual environment 4 or 5 out of 5, confirming that the space looks inspiring and fits the Bali dream. The recommendations build on these strengths instead of replacing them.
“Your employees are truly wonderful.”
“Everything looks good overall, keep it up.”
/07
Findings overview
Sixteen findings from three sources: moderated research, quantitative pricing data and AI sentiment analysis. Fourteen frictions and two strengths, mapped onto the Service Blueprint so every finding has a place in the journey and an owner in the team.
Ideate
Research told us where the service was breaking. The next step was to turn those findings into ideas the team could actually run with. I facilitated a co-creation workshop with Tropical Nomad staff, Community Desk, Operations, and management in the same room, using the Service Blueprint as a shared reference. The goal wasn't to pitch solutions. It was to let the people who run the service every day shape what happens next.
/01
Multi-model ideation
Instead of running a single brainstorm, I used three AI models. ChatGPT, Claude, and Gemini as independent ideation partners. I fed each one the same brief: the Service Blueprint, the validated pain points, and the raw research data (interviews, reviews, survey results). Asking the same questions to three different models gave me three separate, unbiased takes on the same problem.
Same input, separate runs. Each model worked from the identical blueprint and pain points, with no knowledge of the others' answers. This kept the outputs genuinely independent rather than three variations of one prompt.
Breadth over depth. Three models surfaced a far wider solution space than I would have reached alone, including angles I hadn't considered around onboarding and zoning.
A check on my own bias. Running the ideation outside my own head meant the first ideas on the table weren't just mine. The models challenged my assumptions before I committed to a direction.
/02
Cross-model comparison
I then mapped all three outputs side by side and compared them recommendation by recommendation. The logic was simple: where independent models converge, the signal is strong. Recommendations that appeared across all three or that two models framed in the same way rose to the top. One-off ideas weren't discarded, but consensus became my primary prioritisation filter.
Convergence as priority. The more models independently recommended a fix, the higher it ranked. Repeated agreement across separate systems pointed to the most defensible, highest-confidence moves.
Overlap mapping. I tagged each recommendation by how many models proposed it, turning three messy lists into one ranked comparison of shared findings.
Outliers kept in view. Unique suggestions were logged separately as exploratory options, so nothing valuable was lost just because it only appeared once.
/03
Handover: recommendations
At the end of the internship, I delivered a prioritised list of recommendations built directly from the comparison and the Service Blueprint. The document wasn't a design deck. It was an action list, mapped to owners and sequenced by impact vs effort, with each item backed by how many models had independently arrived at it. The team could start on Monday.
The recommendations covered:
Onboarding flow (first-minute experience for new members)
Wi-Fi and zoning clarity across the space
Call room capacity and booking logic
Internal staff communication and escalation paths
Moving feedback from passive (QR code) to active (conversations + regular check-ins)
Deliver
At the end of the internship, I handed over three concrete outputs to the Tropical Nomad team: a final presentation, a Service Blueprint, and a prioritised list of recommendations for the stakeholders to decide on.
/01
Final presentation
A structured walkthrough of the full project. Context, research findings, the Service Blueprint, the key friction points, and a prioritised set of improvement proposals. I presented the recommendations with the rationale and trade-offs behind each one: impact vs effort laid out clearly so stakeholders could decide which to take forward based on their own priorities and budget. Built for both staff and management, so everyone left with the same picture.
/02
Service Blueprint
The end-to-end service view, from first Google search to daily deep work. Frontstage, backstage, and third-party dependencies mapped in one place. Adopted as the internal reference for how the service works and where it breaks.
/03
Recommendations document
An action-oriented list, not a design deck. Each recommendation mapped to an owner, sequenced by impact vs effort. Covered onboarding flow, Wi-Fi and zoning clarity, call room capacity, internal staff communication, and moving feedback from passive QR to active conversations. Handed over as a decision-ready basis for the stakeholders to prioritise against budget.
Outcome
Three things came out of this project. A validated Service Blueprint, a set of prioritised recommendations the team is already acting on, and an ongoing relationship that keeps the work alive beyond the internship.
/01
Impact
The Service Blueprint became the internal reference for how the service actually works. Stakeholder ownership, once a grey area, is now clarified across Community Desk, Operations, Maintenance, and F&B. Feedback collection shifted from a passive QR code nobody finished to active member conversations that produced richer, more actionable insight from day one. The HMW workshop was the first time Community Desk, Operations, and management solved the same problems in the same room.
/02
What I learned
Service design in a live environment means respecting the context you're researching. In a coworking full of deep-work nomads, every interruption has a cost. The method has to match the environment, otherwise you're not researching, you're just annoying people. I also learned that the most valuable output of a workshop isn't the ideas on the board. It's the disagreements that finally get said out loud.
/03
Next steps
The project continues. I stay in touch with the Tropical Nomad team to track how the recommendations are landing in practice, what's working, and where new friction appears as changes roll out.
Follow-up review. Revisit the Service Blueprint with the team to see which friction points have shifted, which new ones emerged, and what the implemented changes did to daily experience.
Success metrics. Define ROI indicators against the recommendations (onboarding time, Wi-Fi incidents, call room availability, feedback volume and quality) so impact can be measured, not just assumed.
Ongoing discovery loop. Establish a lightweight recurring cadence so feedback keeps flowing and the service keeps improving long after the internship ends.
© More works
(WDX®)
digital design
© help centre
(WDX)
clarification
Let’s talk