Key insights

1. Bridge the insights-to-actions gap with evidence-based Service Design

Many organisations collect data and many design services, but the space in between, turning raw data into actionable design decisions, is often neglected. Mauro's team addresses this by combining basic data science principles with the Double Diamond framework, creating a hybrid approach where data collection and modelling are embedded directly into both the problem and solution spaces. This forces teams to identify which data sources exist, which are missing, and how they connect to the service being designed.

2. Use a resource-based view to align diverse stakeholders

Instead of starting with abstract goals, Mauro's team asks each stakeholder what data and resources they can actually contribute, and what value they expect in return. This shifts the conversation from wish lists to realistic commitments, and helps cluster stakeholders around shared levels of ambition: strategic, operational, and guest-facing. By grounding alignment in concrete contributions rather than visions, fundamental disagreements surface early and common ground emerges naturally.

3. Pre-align individually before facilitating group workshops

In complex multi-stakeholder environments, one-on-one conversations before the main workshop are not a waste of time. They are essential. Meeting each party separately allows you to collect their perspectives in a uniform way, spot animosities, and synthesise common ground before anyone is in the same room. This means the group session can focus on building forward rather than re-litigating the overarching goal. Sleeping on the synthesis also gives facilitators time to reframe disagreements with fresh language that all parties can accept.

4. Design honeypots, not stop signs

When tackling behavioural challenges like overtourism, policing and prohibition might not work. Tourists simply ignore roadblocks. A far more effective Service Design approach is to create attractive alternatives: a scenic hiking route that bypasses the farmer's land, a viewing platform near endangered birds, or a nearby beach. Data-driven visitor flow models provide the evidence needed to design these interventions in the right places. The principle generalises well: redirect behaviour by designing better options rather than restricting existing ones.

5. Separate the designer and the data scientist roles when facing clients

Mauro recommends having at least two people in the team: one who speaks the creative design language and interfaces with stakeholders who respond to that, and one who operates more analytically in the background. The fusion of insights happens internally within the design team; clients are presented with synthesised solutions, not raw analysis. This avoids confusing stakeholders and keeps the conversation focused on outcomes rather than methodology.

Automated transcript

## Teaser

You have to do Service Design. You have to do it in this smart technologies. That means, data is always the answer.

But what was the question?

So we run into this issue that we have a strategy that orders us to get more data.

We have more and more projects that require more and more data, but a customer base that largely does not quite like us collecting this much data.

We need to be able to cross the insights to actions gap.

We call it evidence-based Service Design. Taking very basic data science principles and processes and combining them with a fairly well known Service Design approach.

This is like a hybrid between the classic design thinking approaches on something like design science research.

If you have a pre alignment with each party beforehand, then during the stakeholder meetings you don't have to litigate the overarching goal again.

If you want to stop over tourism at its root, you'd have to intercept their customer journey when they're booking online.

The way to get people away from this farmer's land is not to put up a stop sign.

It is to plan a hiking route nearby that walks around the same scenic route.

## Why Service Design Here

we start in the dark, not because I think this is a dark topic, but because, our entanglement with Service Design, as you will see started out of necessity.

So the way we practice Service Design. Is a hybrid of having to justify the work we do and of course, designing new services for tourism. But my personal entanglement with Service Design started earlier. I previously worked at the University of San where we did a lot of Service Design and also general design thinking for higher ideation workshops with tech firms, with banks, with insurances and stuff like that.

And later after my PhD, I thought I, I need to go back to the mountains. I need to go back to my hometown in Coor which is why I joined the FHGR, the University of Applied Science of the Griss. And there, of course, the biggest tourism fact at the biggest industry faction is tourism. And fortunately, tourism has a lot of demand for Service Design.

And that's where I've been for the past three years now. But for tonight, I mostly want to be talking about two things. The first is the one I teased. So what is the context in which we apply Service Design, and how does it shape how we interact with the method and how we interact with our clients? And then in the second half, I have a very concrete example for you, step by step.

How we developed a maybe unique service. We'll see. We'll see what you have to, what you think about it afterwards.

## Tourism Context in Grisons

So I saw there are a lot of Swiss people in the jet, but for those that are not, this map is the map of the lovely griss and just that you get an idea of the context where this story tonight is happening in, because of course, when we design services, we not only can.

Just look at the market. In tourism, we oftentimes have to take into account how tourism is funded, how it is supported by local governance and so forth. And in the Griss, a lot of our GDPR at GDRA lot of our GDP is is due to tourism. We have. Roughly 11 tourism regions with roughly 40 destinations. It depends how you counts.

What is clear? We have four very large destination management organizations called D os. A lot of regional ones by my count, about 25 D os, which together market this mountainous holiday region with about 5.5 million overnight stay per year, and an alto untold number of daily guests. What is important for our story tonight is that each of the regions you see here, most of these Ds are funded either through regional funding trying to promote the region itself, or through a pri semi-private destination management organization that has, for example, some cable cards in their portfolio and so on.

So we have this unique mixture of public, private. And semi-private entities interacting with each other. And most of the time when we have a, as a school and this context we have to operate in one of those two worlds. So the.

## Smart Destination Strategy

Tourism strategy of the Griss explicitly calls out that it wants to be quality tourism, meaning more meaningful experiences over mass tourism, and it wants to do so through smart technologies.

What this means on the organizational levels is twofold. So we have the smart destination framework, which means we want to use data driven. Services decision making mechanisms, the data ecosystem to improve our guest experiences. But only if we can link it with this idea of a smart city or a smart regional government, which is the public administration, their data infrastructure, and the benefits that we can provide to the everyday lives of residents.

This is both. In terms of strategy, that politics, the, a lot of the money is funded from the people. So a lot of tourism, politics should give back to peoples, and this is the context where we have when we do new service designs or if a client approaches us with, Hey, we want to test, I dunno, a VR application, a new cable car design, something like that.

We have to always think in this tool, dual structure. So if the Canton says, okay, you have to do Service Design, you have to do it in this smart technologies smart destination framework, that means, okay, data both for accountability and the design is always the answer. But what was the question? So this is something I'm not saying anything new to you.

Any of our projects usually start with, a lot of data that we have available and a bigger list of data that we need. So two examples I have here from the GNS on the left. I like it. This is a really small example where on the left there you can see a sand bank. And this sand bank is special because there are endangered birds that only nests there.

But it is also a favorite tourist spot. And we already know from past experience that just blocking the road doesn't do anything. People just go over the roadblock and go there. So we needed some data to know why are the tourists coming there and how could we maybe create a coexisting coexistence between the two and the birds and tourists.

So the. The offering that we designed, there was some sort of event space on the right. You can see there is a beach so that the tourists that come there just for the beach are still there. And then we have a viewing platform where you can watch the birds from a safe distance to help them nest in peace where these points are and what information is required and how to make them visible was part of the data that we collected in this project.

Of course we also have bigger projects on, for example, how can we close the data gaps in the carbon footprint of an entire destination? If we talk scope two or three data all to say in all of our pro in all of our projects, whether Service Design or not, we require a ton of data and our clients expect we either can work with their data or that we collect additional data for them.

## Data Dilemma and Trust

The problem is. Oftentimes, either the data's available and we cannot use it, or the data does not exist at all. And this is because nowadays most of the data the basic data that most businesses has, email, travel routes, name orders, and stuff like that. You can use that to optimize maybe your website or your portfolio, but you cannot use it for deep Service Design.

For that, you would need more interesting data, behavioral data, consumption data, movement data, maybe even videos or psychological profiles of your guests. But of course, I don't need to tell you. There's, that's a problem with the privacy laws and it needs a deep trust. So we run into this issue that we have a strategy that orders us to get more data.

We have more and more projects that require more and more data, but a customer base that largely does not quite like us collecting this much data. We've seen this both in terms of the legal avenues that we have since the GDPR, which have dwindled, and of course also the willingness to share of our guests.

These are just two numbers from the pandemic because this is when this really skyrocketed along the lines of the ratification of the GDPR in the eu or the new data privacy laws in Switzerland. So this is our dilemma when service designing, we have either projects where we have. So minimized automized data that is not really interesting for Service Design or we have a whole trust of data that we first need to study and specify to even be able to use it for any kind of Service Design or to put it in a very obnoxious buzzword.

We need to be able to cross the insights to actions gap because we can do the analytics and we can afterwards. Design something, take an action. What was happening in between was the problem for us and this is where our brand of Service Design came in.

## Evidence Based Service Design

We call it evidence-based Service Design, but what it really just is taking very basic data science principles and processes and combining them with a fairly well known Service Design approach, namely the double diamond from the design council.

For those that are not familiar, the idea is that we first define the problem space. What is it, what we even want to solve and what are sub problems there? So can we agree with our multiple stakeholders, for example, in a destination that this is the problem and not that I'll have a concrete example afterwards and only once this is done, we go to.

Classical iterative development. We develop a lot of ideas. We generate a lot of solutions, usually in co-creation with our clients. And then once we have a large list or a bunch of ideas, we combine, we reduce, we converge on one or two prototypes that we can test. And from there we run the ative cycles.

So the data science part came into it when we discovered that two things. First. Oftentimes both in the problem space and in the solution space. As I've told before, our clients did not have the data and we did not have the data. So not in the problem space. We don't just have to first discover what the problem is and define it down with our with our clients.

We also have to. Find out which data sources are there that can help us in defining the problem and which are actually available to us. So that is what's happening in the problem space. And that the same thing is happening again in the solution space where when once we have some data, we can either run some basic models or we can even create a data-driven service, or we have a concrete goal for the Prototype.

At the end this is like a hybrid between the classic design thinking approaches on something like design science research. We are not quite a scientific discipline. We are not quite just design thinking. We're somewhere in the middle. You get the idea it sounds much more fancy than it is.

It is still design thinking. At its core, the design thinking principle are the important part of it, but we need to emphasize the data collection, processing, and modeling, both for accountability and because we believe it creates better services. And for those of you that are academically minded, if you want a formal definition evidence-based Service Design is a step-by-step approach based on data, scientific principles and customer data to either improve existing services, create new services that effect effectively address the needs and preferences of our targeted stakeholder groups.

So now that we. Got the very high minded ideals out of the way. Let's make it with a concrete example. Hopefully much more clear.

## Case Setup Visitor Flows

So the problem that we'll be talking about, interestingly, also started in the pandemic. During the pandemic, we've been approached by a lot of destinations that wanted to track their visitor flows, both because they were required to do for example, to not allow too many people into the destinations to be able that what do you call 'em? The safety message measures in terms of distance are upheld and so on and so forth. And of course, the organizations that approach us were the destination management organizations, the d os. In this case specifically we're talking about devo.

And the problem with most demos in Switzerland is they are kings without the realm. Meaning they are large organizations and they oftentimes have sway over marketing, branding, advertising, and allocation of funds, but they don't have usually a lot of guest data of their own. They only have a very indirect view of visitor flows, maybe through booking system or guest cards.

They themselves have no offers, and therefore they cannot impact visitor peaks all that much, and therefore most of their services have low predictability, especially during. An extraordinary time like the COVID pandemic. So this was the broad problem in phase one, where we came together with Devo and it told us we need to be able to track our guests.

How can we do that?

## Stakeholder Goals and Data

So phase two then was asking the question. Is it really tracking the guests or are we talking about resource allocation and so forth? And there we have a very basic questionnaire. We see it on the right there. It also has an overly fancy name, the Applied Tourism Intelligence Management Framework.

What it is another canvas, but a useful one because, first step in this phase two is get as many of the stakeholders in a room as possible. So not only the DMO, but also the municipality, the government the cable cars, the hotels, the restaurants. We put them all in the room. We post the original problem.

Hey, we would like to take track our guests. Is this something that interests you? And then quickly, we also talk about data. So we do start. With an overarching goal, but then we quickly bring it down to a resource-based view, meaning which available data sources can we use in collaboration with which stakeholder to generate what kind of added value?

In the case of davo we were not able to get complete agreement, but at the end we agreed on three levels of problems. And what's the specific. Possible solutions could be on these levels. So on the very top, you have the big companies and the destination management organization on the strategic level that wanted to be able to track enough data points and enough guests to create forecast and simulations of visitor flows through davo one level down.

You have the operational levels that want a daily or medium term forecasts and specifically link them to their systems so they can use it for resource management and capacity management. And then for the even smaller ones and also the municipality and for the local residents, they wanted a live and direct information system, either over a screen, over the website or something that can tell you, Hey.

Maybe don't leave with your car yet. There will be a blockage leave at eight o'clock. Then you will have free roads, stuff like that. So these are the three levels of goals that we had. Still no solutions. So still based on what data we have, we then first try to create a first model of can we model Vista Row flows with what we have.

And what we had of course are public data, like from. The Swiss department of, I don't know how it's called in English, it's called Astra. It's the one that surveys all the motorways weather data and of course booking and holiday apartment renting data that we have and the guest card data.

This gave us a live system that was not quite well, but it also showed us based on the available data points, okay, we cannot reach the free goals that we just decided on. This is not enough. We can give it a live ticker, but it is. Not that details. The predictions are not that great. We need something more.

## Sensors Models and Dashboards

And the something more was we identified again in co-creation with our stakeholders. What are specific measuring points that are important and quickly we noticed? Yeah, the parking spaces are quite important and specific points of interest. Now, point of interest here can be anything. It can be the entrance.

To a local hotel. It is the lo entrance to the swimming pool. It is the entrance to the skating arena and stuff like that. And there now we are finally in the solution space. Still on the data that we have, we know quite exactly what density, what kind of data we need to design something on top. So we looked for additional data sources that fit.

We mostly used 3D sensors and video analytics. I will have some examples shortly that we installed in these points of interests. Of course, in accordance with the municipality, the DMO, and all the stakeholders. Oftentimes, especially the 3D sensors, these are infrastructure that most of these stores already had in inventory.

So they were all already, they had this data, they just had not linked it up with other users yet. So phase five, now we have. The goals, we have an idea of what we need, how this model needs to look like on all the three levels. Now it's time to create some first prototypes. So here what you can see, I don't know.

I'm not sure if the video will play. Yes, it plays. This is a test for the algorithm behind the camera. What you see here, the boxes are all people, and of course they are boxes because we are not allowed to send out or use personal identifiable data. So the camera. You can imagine, like it has an AI behind its lens.

It only sends the information on movement vectors and counts to us. So it does not send, even if it dep pixelates here, it does not send, this is a man and blah, blah, blah. So it sends both the raw video feed and then the movement vectors from this. We then could recalculate our models and create the first dashboards as prototypes also to see what is actually used for the four goals that we defined beforehand.

So we had some basic predictions on infrastructure, usage, infrastructure in these cases. Roads, cable cars parking spaces, and well other things. Visitor hours, of course. When are the peaks, when are. When will be the peaks? More importantly, traffic statistics quite important for a valley that only has one exit, one, one entry, and overall just guess numbers, which then again could pluck into a more high level predictive algorithm.

## Rolling Out and Monetizing

Now, this is of course not yet a Prototype. This is just a dashboard, and a dashboard does nothing if it doesn't reach the right people. And so we started to test again across three levels across with all our stakeholders. The first was on the guest level. This is, I have to admit fully, this is still a bit tricky.

We do have this information public on the website and then the guest app, and now they was included it in their AI guide, but we also know. Mostly the locals look at it and not that much. So this is still very much underutilized and we're still looking for digital touchpoints where we can reach our customers much better on the organizational level we had to do quite a lot to get a functioning data governments framework, because Of course, yes, all of the.

Entities that you see on the right here. They share data and they agree to the free goals, but a lot of them have different needs. So each of them has a different dashboard. Each of them has a different configuration of the model. Each of them has different data sets in the backend. So on the organizational level, we had to figure out who even receives this information.

We had to also do a little bit of, training with them on what can you now do with this model? How can you use it, for example, to plan your resources and so forth. And finally, of course, on the strategic level here, again, we are still working on that quite a bit. We currently are working on a follow-up project where we especially want to work on the monetization.

Because currently it's the destination paying for all of this apart from the touch points that are owned by other stakeholders. But there is a lot of interest, for example, from bu so the DMO one level, higher on the cantonal level to use this with their own data sets to create something new, maybe create other services and so forth.

And of course we are now also looking at. Smaller mini services on how we can integrate that, for example, on optimizing the vista flows through the cross country skiing loin. I have no idea what loin means in English. I'm sorry. The cross skiing lanes. And stuff like that. So on top of the data that we collected and on top of the goal and the prototypes that we decided on, we are still working on refining those into services that are used in the backend and in the front end with our guests.

Quick check because yes, this is still a very much ongoing thing, but following the six phases, we were able to develop a sustainable data strategy. Now, probably the best outcome of this was that we have now a platform as a service where our stakeholders can also deposit, for example, future data.

So if they enter a collaboration with, let's say, MasterCard or Swisscom to. Maybe refine our models or just refine their business. They have an infrastructure where they can share it and they have a common data strategy on the destination level that tells them this is something all the others want, and this is something maybe only you want.

Which led to quite a close collaboration with local companies. We're still working on working with the local cable cars and with other destinations, but also there. The platform in the middle. This predictive algorithm helped to bring a lot of people on board. This sort of repetition of the first point, the data strategy not only works, it is also in accordance with the destination strategy.

The. Local governance. So with the destination strategy of both the Griss and Davos, and I have to be honest, we're still working on monetizing the data processing. So the service itself, we're still a bit in the prototyping phase, but this is the nice thing. Iteration is not just an excuse, it is also the step into a next project.

And with that hopeful thoughts that our work never ends. I ends the reporting part of this presentations, and I look forward to your questions.

Thank you so much RO and I love the fact that there is even a thank you word on on here, which is written in Ruman, which is one of the four, local languages of Switzerland that not many people know.

## Q&A Stakeholder Workshops

So that's a beautiful thing to see to see also that, so now we're moving to the q and a time so let's start right away with the few first questions. One person asks who you talk more about the meeting with the DA diverse stakeholders.

How did you facilitate the workshops in order to get to those levels? For example you had to map each stakeholder to their requirements and it boiled down to three levels. Can you give maybe a few specifics, because you said like we arrived to that common ground, but I imagine that in between the line there was a bit of a complex process behind that.

Maybe can you share a few behind the scenes elements around that? So this is also probably the most important step because afterwards, once you're in agreement, the actual work. Not gets, not trivial, but gets easy. So yeah, this was a big fight, especially because you might also see it a bit in the goals, the original goals that the destination have are quite a lot more specific, which I'm not at liberty.

I tell what they were. So first step we have to figure out, and this sounds a bit dumb, but what are the animosities between the involved players? Who are the players that maybe we should not put together in a room, but interview separately? Who holds what power in terms both of data and money? And then see who's aligned how, and then afterwards, it's not simple, but first step, you go to each of them and you record in a uniform manner how they think about this problem, what they want out of it.

Especially what they can deliver. So this is also where it helped us a lot to just focus. Yes, there is a use case, but we're doing this quite resource oriented. What could you even contribute? What would you like to contribute? And is this enough to get the things out of it that you want? So this is first step.

Just collect all of them. Then we compiled possible solutions and possible recommendations. So the ones that you see here, there was a pre level where we oriented it up invited them all to a final workshop. We told them, Hey, we see you have a cluster of. Ideas here on the strategic level, you have the cluster of ideas here on the operational level and ones of the guest levels.

Which ones do you like? The answer, of course, was all of them. So in the with all the stakeholders, we then just discussed this step, what do you want in each of the clusters? But the clusters is something we created by going to each of them in turn and asking them, what is your use case? What are, what can you contribute and who do you see responsible for what?

Thanks so much. I, there is a follow up question around this question around the stakeholders, which is someone saying, I'm curious about the type of questions and discussions you had with the stakeholders when you presented the problems to find some solutions. Do you have two, three model questions to break the ice with the stakeholders?

Maybe can you share some specific tips and tricks that you use to maybe spot these special dynamics? What are questions that you bring into place that get people to talk and feel that they can share some valuable data, as you say so brilliantly? Very good question. And not to plug myself, we also have.

It's called edix. It's online documentation on how we do ideation workshops, but that's a separate part. To give some quick examples, if we are in a really, but we hope to avoid this situation, but if we are in a really adversarial conversation, which here was not the case we usually start by defining common ground.

Just for example. Very basic icebreaker questions. Why do you like the destinations and why do you think the guests like the destinations? And then the overlap is something that we can work with. Another thing that helps a lot is starting with very broad goals, not processes. Usually you are approached with very concrete ideas, either already for a service or for something that should happen.

So we want to track visitor flows. Yeah. But why? Because then we know where everyone is. Yeah. But why? So that we can make the guest experience better. Yeah. What does this mean? So that you try to drill all the way down to, hey, if this works exactly the way that you imagine it, what has happened?

Do we have 10,000 more guests? Are the guests that are here more satisfied? What are you playing at? And if you break it down to this level, you see quite quickly where there are fundamental disagreements. So a classic fundamental disagreement is you have the people that thrive on huge masses of tourists, and then you have the businesses that thrive on quality tourists and longer stays.

Of course, at the end of the day, all of them would like to make some money. But not all have the same idea on how to get there. So a lot of fundamental disagreements are mostly along those lines on, let's call it broadly quality tourism versus mass tourism. And when we see these fundamental disagreements, we usually try to find a common ground ourselves.

So we ask them first separately, and then we try to find a common ground by analyzing both answers of both entities. Maybe one thing that I'd like to point out here, which is maybe a thing that is very specific to Swiss culture is in Swiss culture we have the thing where we have conversations first in one-to-one before then having the big workshops.

Because there is this idea of have helping everybody keep face which is a very specific Swiss way of disagreeing. First we disagree in private. So that we then can meet together to build the common ground. And so that's maybe something that I think that you do in a very beautiful way. But that is maybe for people coming outside from Switzerland, maybe a bit of a long process, but that is also culturally relevant to, to Switzerland.

True, especially all the pre alignments. It sometimes can be, feel like a waste of time, but. If you have a pre alignment with each party beforehand, then of course during the stakeholder meetings where everyone is involved, you can focus on the really relevant parts because you don't have to litigate the overarching goal again.

But yeah, you're of course this is very much a Swiss thing to find a compromise first. Yeah. And I believe it, it has some beautiful qualities that I think other practitioners from other countries can also benefit from. 'cause there is one quality of this. One-to-one thing is that it forces to sleep on it for the synthesis of, oh, in fact, the common ground is around here, which you couldn't do live in a workshop maybe because you don't have the.

Time to sleep on it. Where by doing it in this slower way, maybe you have a bit more time to say Oh, but now if we look at it differently, in fact there is an agreement, but the words are different. The framing is different. And maybe if we use. A new language to speak about that agreement, that common ground, then suddenly we can get everybody on board.

I think that's a beautiful way. Obviously we are the Swiss Service Design Network, so therefore we also believe that there is a cultural aspect that is also interesting in the way Service Design is. Practiced. I'd like to go on the next question.

## Overtourism and AI Personalization

And so this is now a bit of a zoom out to the case in general.

Do you use this model to tackle over tourism? So you said Davos is this very specific place. And do you use that model to tackle overt tourism or any conflicts between. Tourists and local citizens. That's the first part of the question. And the second part of the question is, do you see any potential in using AI solutions in order to personalize the experience?

So maybe let's first look at this question of overt tourism conflicts, tribulation between tourists and citizens. Yeah, over tourism was on their mind after the project was finished. Because there was quite a directly after the pandemic with a lot of revenge tourism, and the answer is yes and no.

Yes, we'd like to but no currently because just being able to predict visitor flows doesn't help much. And the huge problem that we have is that once. The guests are here, it's already too late. So if you want to stop over tourism at its root, you'd have to intercept their customer journey when they're booking online or they're booking their tools and stuff like that.

So what we can do, and this is why this also, yes it froze into sharp relief where the problems actually are and what the problems are. So oftentimes overt tourism, of course, has an objective dimension, how many people are walking through here, but there's a huge subjective dimension especially in nont tourist regions.

Over tourism usually is just four or five really bad weekends where a lot of people arrive, and that's the overt tourism. And having a data ecosystem that is able to capture that. Helps a lot. So we have a similar project with Vimal tourism and in this region there's also National Park which has, there, we have a measurement system for low frequency because there's just some hikers walking through.

But there are some weekends where there's a hundred guys walking past the station, which for a national park is quite a lot and fields for a lot of guests. Like this park is overrun and there we can work with suggestions because they enter at the park at a specific place and then you can just cut off trails that are overused here in Davos.

We mostly use it for, to manage parking, to manage resources. We tested a recommender system, for example, through the hotels that you say, Hey, you're my guest, please only leave at five o'clock. Then you don't get in a traffic jam. That kind of doesn't work all that well because the tourists usually have their plans and they want to keep their plans.

So intercepting them in the destination itself is quite difficult. I think it gives us a lot more tools to plan with and maybe plan alternatives. So if we, for example, seen davo a lot of people, this is not the case, but as an example, a lot of people are walking illegally through a farmer's land. The way to get people away from this farmer's land is not to put up a stop shield a stop sign, sorry.

It is to plan a hiking route nearby that walks around the same scenic route. And the data to plan these kinds of interventions comes from these kinds of systems. So yes, but with a lot of work to the overt tourism question. Thanks so much. I really appreciate the quality that you bring in this. Just saying stop, don't, do is not the best way.

But there is like having honeypots somewhere else saying Hey, there is another interesting thing that doesn't disturb anybody, but it is way better. I really like this approach also, which is, also, again, very sweet in not having to police, but rather in suggesting and and helping people maybe on, on the technological side, we have another question which asks, do you see any potential in using AI solutions in order to personalize the experience?

So you've been showing that it's quite data rich and that you have this approach. Has. All of this stuff that is happening with AI changed your approach around personalization of experiences. Do you have an opinion on that? Yes. And there are of course a lot of pilot projects being tested. So generally there are two ways that we currently think about it.

The first is, again, a classical service, so griz in tourism. Has a collaboration called joa. There's an AI in there, and what it is personalized travel guides that are fed with some kind of specialized knowledge. For example, for K, which is, or claims to be the oldest city in Switzerland, they have an AI Roman that can help you walk through the city and give you fun facts about ancient Rome and where the historical sites were and so forth.

So there are these kind of small services, i'd say gimmick, not in a derogatory way, but this is a small part of a larger experience. Probably the larger impact is figuring out how can you get this type of data and especially the kind of specialized knowledge, for example, that tour guides have or that local hoteliers have into.

Chatbot because yes, you can currently plan your vacation in the instance with chat GBT or Claude or whoever, and it will give you a lot of popular spots. But what it is really bad at currently, for example, is being accurate on what's open when, what's the price, what are the booking conditions, how does it look like?

It tends to be a bit too nice. It tends to still hallucinate a lot of stuff that is not there if you pressure it enough and. More importantly, for a lot of our partners, it is not really good at personalizing, interestingly enough. So if you give it, for example, your psychological profile and then you feed it all the information that BU has available for all the guests in the griss, it'll still give you a fairly generic, fairly stereotyped recommendation.

Which is a problem because of course that is not true personalization. That's just stereotyping. The question becomes, what are maybe, can we solve this by either training our own systems? Can we feed the data in some ingenious way into existing chatbots? Do we need answer engine optimization?

It'll be somewhere around there. But currently yeah a lot of the s I have the feeling a bit disillusioned with the possibility to it. We at this moment have with this wave of LLM chatbots and what it can do. Yeah, the use case is not quite there yet, I believe, and the combination with these kinds of large data ecosystems is still far away also legally.

## Service Design Roles and Sustainability

And now if we zoom out even further and we think about the whole practice of Service Design, we have a few questions that I'll try to merge live for you. One is around this. First the A context, which is that yes, we know that destinations are complex systems with many different actors, with different powers, with different interests, with different cultures, et cetera.

And so the question is, what do you think is the role? Service Design practitioners what's their role in facilitating these collaborations between those actors? And not only for one specific project, but for their long term. And the second element to that is do you see that Service Design could be something interesting to help?

D mos rethink or think differently or think even further their tourism system? Very good question. So I think interestingly enough, the d os themselves oftentimes, especially the big ones, they use Service Design and they use it quite effectively and. The purpose they use it for is the same as we do.

You said it in the question itself. The facilitation of different stakeholders is very difficult and often challenging. And it sounds quite simple, but just having a process that you can point at and say, Hey, look, we're in phase one. That usually helps a lot for example, in a workshop to get PE people together.

But that's just a very base level. It's also the philosophy of we are not dictating. From top down what services we want and what we want to do. It forces us to take accountability if we have to create our services in tandem with local stakeholders, with local groups. And this is also why we are emphasizing this data layer as much.

For example, if we also have a lot of projects around over tourism that are specifically launched by the municipality because they have complaints from their populace. Our first step there is quite rigorously defining how does the population understand the problem of our tourism? And then suggesting solution and then checking with them again, are these popular?

And sometimes they can diverge from what the municipality envisioned, what their solution would've been and what they would've designed. So philosophically the benefit is really. In the spirit of co-creation that you're forced to take multiple perspectives into account. And on the micro level, it is really just, you shouldn't use these approaches as checklists.

But sometime it helps to pretend like it is a checklist and say, we are at phase one. This is what you can do now and we'll get to phase four where we come to concrete solutions. That also helps us helps a lot. And I'm personally curious about one thing in this view that you're showing right now, there is this powerful double roles that that you're living you're living the role of.

The design thinker, the service designer, the innovator, the idea rich experimenter. And then on the other side there is the data science person, the numbers guy, the business guy, the. The engineer and all of that. And so one thing that I have seen is that sometimes people struggle in the world of sage Service Design because it's this match of two worlds.

There is the whole business world, there is the whole design world, and sometimes people say. How do I work with these two identities? And I'd be curious to hear from you how do you play with that? Are you always presenting yourself as Maro, data science, Service Design guy? Are you sometimes changing hats?

How do you manage that? So we usually do it with. A dual role. So I don't think it's a good idea to present yourself as, yes, I'm both the data scientist and the designer. We usually it, it depends on what our client wants from us. So in very design heavy creative spaces, we don't need to over emphasize that we are still doing some analysis in the background is, for example, doing.

AB test with the population, asking for their opinion, collecting additional data and stuff like that. Then that's just a work package in what we do in the overall design. So to answer your question concretely, what helps is have at least two people, one that speaks the designer language and interfaces with the parts of your clients or your stakeholder network that respond well to that and to have.

The data scientist, just the analysis guy that has the license to operate a bit more not scientific, but a bit more nerdy to approach approach a topic. And then what We usually don't do this inside to action gap. You don't talk about raw data with people that don't want to. So you're always in the background.

The fusion, the. Putting things together as we did, for example with the three levels. The levels is our doing. We put these three things together. We could show you who said what behind that, what the support is behind each of those. You don't discuss the analysis with the client unless of course they're demand it or unless it is part of the design itself.

Yeah I'd say. Have two roles, separate them and only discuss them within the design team, and then come to the client with the solution or what you've come up with at least. Thank you so much. I'm mindful of time, so I will ask a very philosophical question. But we'll try to just give a few pointers.

And one thing that I want to put here is we have a lot of questions around stakeholders management, so I think we will have to rein invite you just on that because I think people are quite excited about obviously the approach that you have with evidence-based services line. But they have a lot of questions on how you manage all of that.

And and I'm sure we should see you again once around that. But now for the very philosophical question that I'm taking out from the chat long question, I'm gonna make it short, which is how. Do you align the business interests with ecological ones to achieve the improvement of existence of existing systems through Service Design?

So tourism needs obviously that the mountains stay beautiful mountains but still tourism needs people to come and visit. How do you balance that? How do you manage these very different needs, wishes, dreams, and visions? So I'll first give you the cowardly answer and then I'll give you my opinion.

The cowardly answer is, thankfully, in the gns at least, and I think in Switzerland in general, we don't have to worry about this too much because we're seldom approached with. Something completely unsustainable or something completely out of the blue that is just directed towards profit. So fortunately we don't have to balance that too much.

But to take the question seriously we try to insert our opinion in some places. So for example, that we. Operate with as many stakeholders as possible in the destination. That is both a value statement. One that values democracy and diversing opinions. But it is also, to be perfectly honest, just good design practice.

Because in tourism, if the local populist doesn't support what you're doing, and the DMO loses the moral license to operate. Then that's bad design. So this humanistic focus is baked into the methodology and therefore most people, at least it's I think, worldwide, they care about sustainability.

They don't want to destroy their environment. And if you work with the local populace, they can also tell you where the destination is maybe quite vulnerable and where it isn't. Of course, they are not arbiter of objective truths. And of course there are a lot of moneyed interest in tourism. The fact is a lot of our jobs and a lot of our villages exist because of tourism.

And this is not something that we can just ignore. The best we can do is just work within the guidelines, work within the sustainability goals, both of the gris tourism strategy or the Swiss tourism strategy. And if you align a lot with the local populace as well, I personally believe egregious I don't know very dangerous or ecological disasters will not happen on purpose.

At least they have not happened yet to my knowledge, fortunately. In my project. Thank you. You managed very well to take one of the most ful questions of the night and answering it quite brilliantly in two minutes, so congrats for that. I really appreciate that.

## Closing Thanks

Thank you Mauro. I really appreciated your approach your talk and all the questions that you answered. A big thank you to all of you for your smart questions, your participation your attention. A big thank you to everyone who has come tonight. I wish you all a lovely rest of the day.