Webinar: Inside Swiss Federal Railways’ 2026 Customer-Centricity Journey: The Rise of AI Personas

Hello, everyone, and welcome to today’s webinar insights with Federal Railways twenty twenty six customer centricity journey, the rise of AI personas. My name is Maria, and I’ll be moderating today’s session. We’ll start with a brief introduction to Market Logic and Persona agents that move into Swiss Federal Railways journey from solving their knowledge management challenge to using personas as a customer centricity tool. And I’m very happy to be joined today by doctor Alexandra Daniela Zaug from Swiss Federal Railways, who leads customer insights initiatives at SBB. From Market Logic, we’re joined by Michael Vincent, who works closely with the SBB team, and Joseph Rini, Director of Product Management. At the end of the session, we’ll open things up for the audience Q and A. And with that, I’ll hand over to Michael and Joe to briefly introduce Market Logic and Persona’s agents. Thanks Maria. Many people will already be familiar with Market Logic, but for those who aren’t, Market Logic are here to help the world’s biggest organisations and brands anticipate their consumers faster. We do that by centralising all of their market insight data, primary, secondary, their structured, non structured data sets, so that they can understand and know what they know already in the organisation. We then use the latest technologies and capabilities within AI to build tools such as personas that allow and help them to bring the consumer into their everyday decisions, and then onto developing monitoring tools, trends monitoring tools such as consumer trans agents and competitive trans agents that allow them to stay in front and ahead of any changes in that market space, ultimately turning insights into innovative ideas faster, testing those and bringing them to market at speed. The journey that we started with Doctor. Alexandra today was a typical one where Doctor. Alexandra, although was incredibly AI savvy, but was looking for a solution that would allow the company, SBB, to have wide access of those valuable insights using a trusted AI platform, and really to start to drive into one of the initiatives, which is around customer centricity. I’m very pleased to hand over to Joe in a second to take us through the discussion today on personas, then Doctor. Alexandra will take us through the story. Over to you, Joe. Yeah, great. Thanks, Mike. So quickly, before we go to SBB to really talk about how they have been using and are using the personas in their organization, I wanted to spend a few minutes talking about really what these personas agents are so it’s clear to everyone before we get going and a little bit around typical use cases that customers are using and how they’re set up and so on. So what are these persona agents? Just to be really clear, these are profiles brought to life by the power of large language models, so Gen AI, but based on our customers understanding of their either consumer segments or could be B2B expert groupings and so on. And we use our technology to bring those to life in the backend so that there’s a chat based experience that users can go to to speak directly to these synthetic or synthesize personas, understand the lifestyle and personas sale. And maybe not doing everything that I’m mentioning here, and I’ll cover further, but this is just a kind of general overview of how our customers across the board are using this. The real strength of these persona agents is you’re then able to speak to your consumers, so to speak, much earlier and at any time, anywhere. So bring them much more into your ways of working and flow and really accelerate time to insights. So again, just some, let’s say general or across the board ways that our customers are using the personas at this point, and I’ll kind of go left to right, top to bottom. So first of all, campaign testing, we are seeing a wide swath of our customers use the personas to go in and expose them to campaign ideas, product concept reviews, especially in early stage, you can go to one or more personas and run concepts by the personas, get their feedback, tweak and ideate. Of course, general lifestyle exploration. So you wanna know more about this persona or segment rather than trying to look at that in a static deck somewhere, you can now go and in this really intuitive way chat with the persona and learn more about them. And you’ll see all this in in a live demo of it, of course. Also, packaging feedback and so on. Another interesting area that especially in the healthcare, pharma, and other b to b focused customer groupings we have is all around things like adoption journey mapping or other ways to interact with experts and so on. And we have nice sets of personas that can be set up to handle that use case as well. And then of course, sort of across any of the persona groupings, because there’s this group chatting capability, we facilitate a comparison between the segments, which is really rich. Don’t wanna spend too much time on this, but what sets us apart? First of all, these are easy to set up. They’re highly tailored to each of our customer groupings. Of course, easy to access in our one centralized Market Logic DeepSights platform, as Mike mentioned. And then we have a host of advanced capabilities. I think you’ll see some of this in the demo, like history, project management capabilities, group chatting, summarization, the ability to ingest and give feedback on images, soon to come the ability to generate images, and a robust road map around the entire offering that, of course, Market Logic brings to the table. And this is all within our scalable and trusted environment, of course. So last slide on me before I pass it over to Alexandra, how are these personas set up? So there’s a couple different ways, and I think you’ll hear about how we did this together with SPD. The first is to really take predefined personas or segmentations from the customer. So perhaps longstanding understandings of given segmentation groups that the customer would have, they’re probably living in PowerPoints, they’re disseminated within the organization. We take those and bring those to life in the backend and turn them into these chatable personas. More and more, I should say, we’re starting to also do this, or we are successfully doing this now with customers from raw transcripts. So if you have groups of transcript interviews with various segmentations or survey data, we can take those and turn them into personas as well, along with working with several trusted partners like Mindline and Next Atlas to bring their personas to life in our environment. And happy to talk more about that in the Q and A as well. And just starting out is something that we’re really working on this quarter together with some lead customers. We call it the persona builder. It’s a little bit of a different approach. We’re not gonna focus on it today, but it’s the ability to rather on the fly generate personas from either the DeepSights repository or an onboarded, let’s say, survey database and so on. So that that’s another neat area that we’re going in this persona space. Great. And with that, I’d like to turn it over to you, Alexandra. Thank you. Thank you, Joe. I’m very happy to be here and to show you how we use DeepSights and DeepSights personas in particular. For those, who do not know, the SBB, maybe a few words, we’re the National Railway of Switzerland. Our headquarters in Bern and we’re founded in around nineteen oh two. Obviously, we transport passengers, both for long distance and regional distances, but we also have freight services, major portfolio of real estate around stations and mobility hubs. And we do the operating and the maintenance of railway infrastructure, so trains can actually run. A few facts and figures. We have around one point four million passenger each day. And on our network, there are around eleven thousand five hundred train each days. And in twenty four, the punctuality was around ninety three percent. I think in twenty five, it was even ninety four percent. So let’s start with the things you’re really interested in. That is how we use DeepSights and DeepSights personas. There are two parts. The first part is our knowledge management challenge. The second one will be the personal best. So when someone in within our organization needed insights, They typically came to us before DeepSights asked us, what do you have about the topic? Can you tell me more? We would search on our SharePoint, not always find on our search SharePoint, and then ask other colleagues. Typically, this would take about a couple of days up to a week, and then we send back, hopefully, a lot of files. So the person who needs customer insights can go through all the files and check whether there is actually something in it that helps or not. In theory, they could also access the SharePoint. But as we do not find our insights as customer insights team, it is all the more difficult for those who do not know customers insights at all. So, obviously, we have a couple of pain points. Our insights are very hard to find. It’s time consuming to find the insights. And what’s a real pain? The answer depends on who you ask. When you ask someone senior, you get more information. When you ask someone junior, you get less information. And there is even a bigger problem once a person leaves the SBB, the knowledge is lost as well. And yes, self-service would be useful or could be done in theory, but actually it isn’t possible. We’ve tried several times to amend this and to introduce repositories, databases, but it was always a failure because it took too much time to introduce metadata, to add metadata to each study. And the problem was also you have ten to fifteen minutes for the metadata, and still it doesn’t cover everything that’s in your study. And you can be sure the next time someone asks for a study or needs insights, it isn’t the same thing as you covered with metadata. So, with the rise of AI, we thought, well, that’s our opportunity to solve this problem. And we started to write a long list of requirements. And you see here our three key requirements for the solution we had. First of all, we wanted the intelligence search. So no need for syntax for querying the database or something complicated, just natural language. And as we are a company speaking German, French and Italian, multilingual support was important as well for us. For us as researcher, we wanted low maintenance, implying that we don’t have to spend ten to thirty minutes for metadata, that we just can upload study and everything is covered. And it was very important to have high quality. When we started, this was around the time where LLMs grew popular, but didn’t display any sources except perplexity. So it was very important for us to have these sources displayed together with the answer. We wanted really control and transparency. We’re very happy that DeepSights just meets these requirements. And at the business level, the main advantage is really that we have instant access to our studies and that we can reuse our studies. So we don’t need to do everything again just because we don’t find it. On the feature level, we’ve got synthesized summaries. That’s, to be honest, even better than we expected or we had on our requirements list because we thought we would get a list of files with the exact page referenced so we can check it. But to be honest, it’s so much better to have just a ready to use answer. And I think, in particular for our users, that’s a huge, huge advantage. We have source control. In DeepSights, there is only selected content and only a few members of our organization are allowed to upload content. Yes, we can ask in natural language in multiple languages as we wanted to. I’m not the next part, but I think market research specific regs provide much better quality of the matches between answers and sources than just an ordinary LLM. And for us as researchers, it was also too important to have the software as a service, implying we don’t need internal resources, be that for the setup, be that for maintenance or for development. We just can plug in once we’ve done the security and data protection process. And I think our key to success is focus. We are focused in scope, meaning we only use selected features of DeepSights. There are a lot of ways you can use DeepSights. There are a lot of features. We focus on search for the moment. And last December, we started with the personas as well. And we want to use them really actively and not just them on a superficial level. We also have limited access for users. This is because the SBB has a lot of people working in the field. The majority of our employees works in the field, for instance, a train conductor, as a shop assistant, as a maintenance worker for rail tracks. So there is no need for direct access to customer insights. And this would make the costs explode if just everyone had access. So we’re focused on really those who need customer insights for their work. And as I already said, we have high quality content. This was a long discussion when we set up DeepSights. Do we upload each and everything we have? What is good quality? We have market research back to nineteen ninety eight, so we would have quite a lot of studies we could upload, but then decided there’s no point in uploading all the material. We focus on studies that are younger than twenty fifteen and are good quality. And there are a few selected ones that are older, but not much. So, the value we get from, this focus is insight based work. It’s really great that users just can ask questions. They can ask follow-up questions. That’s also very important. Previously, when you think of the process, they have to go through whenever they had a new question, they had to come back to us and ask, have you something about that and about this? And like this, they can just ask the questions they want to. And I’m really excited. I think soon we will be able to chat with our insights, and that will be even greater. So, this is a huge advantage and it’s also the reuse of studies and reports. The first step now when we do market research is to check whether we actually have insights about this topic. Also, if there is partly something that would help us to shape the study smaller, like this, we can save money. And of course, as it’s so easily available and we can share it much broader, it deepens customer centricity. It’s a first step to a data driven company and the customer focused company. Of course, as always, there are challenges. The main challenge here is the quality assessment because with AI, thinking is never optional. So you always have to assess the answers. But the better the answer is, the less you see sense in checking each and every answer. So you get kind of lazy, but really to keep in mind that not always not a hundred percent security is given with AI. It’s very difficult to keep that in mind and tell people you have to think yourself. In particular, as data literacy is a bit unevenly spread around within our organization, I think that’s like everywhere. And as a market researcher, you put much more importance on data literacy than others do. And sometimes it’s also a bit difficult to explain when they come and say, hey, a general purpose LLM has provided an excellent answer, much better than DeepSights. There is more and I like that more. Why do I have to use DeepSights? So to explain where the difference is, and there is really a huge difference in quality, isn’t always easy when people aren’t really data literate. But I think that’s a challenge for another day. And I would like to start with personas for customer centricity. For those of you who haven’t worked with personas yet, feel a few words what a persona is. It’s a very detailed, clearly defined person who doesn’t exist in reality. So this means it’s really, really specific. You have to be able to think like her to see her or his perspective. And it’s not just average, it’s very specific. So it’s not in an urban area. Christina here lives in Bern. She is not in a range of sixty to seventy. She’s sixty two years old. So you really can step into the shoes of this persona. And it’s not just any person randomly chosen. This persona represents a group of people who share similar needs, preferences, and behaviors. Typically, you have a name, you have a category like, this regular leisure leisure commuter, you have a picture, a few key characteristics as well as a slogan so you can place the persona easily in a category and a description. The SBB has ten ad hoc personas. For us, it was very important to have a broad range of personas to have edge cases, minorities, as well as the majority of customers depicted. This is the reason why we chose to implement ad hoc personas. I will talk about that in a few minutes. Let’s shortly go through our personas. We have Christina, which uses public transport mainly for leisure purposes. She needs it to be simple, easy and wants to travel hassle free. Kenji, our tourist wants to travel hassle free as well, but he doesn’t know the system. It’s everything new as he is an overseas tourist. So he needs much more guidance, much more explanation. He doesn’t speak either German in Italian or French. So we somehow need to communicate in English, which isn’t always standard, on the national level, but very important for for him, to get along. Then we have Luca, a young adult who is very price sensitive because he doesn’t, earn much money. There we need to offer, inexpensive, possibilities to use the train and, also to provide for his flexibility. Sina, on the other hand, doesn’t really like flexibility as, her mobility is reduced. And for her, it’s more important to really reliably plan a journey and to be safe that everything works like she planned because, just switching at the platform doesn’t work for her. It’s probably also a bit difficult for Elias, who’s our car driver. If he can avoid it, he doesn’t use public transport at all. So he very rarely uses public transport. In contrast to Leah, who is a public transport user first, so whenever she can she will use public transport and places a lot of importance to sustainability. Then number six is Omar. He has recently arrived in Switzerland, so he doesn’t know the system. He doesn’t speak very well German, French or Italian. So it’s a bit complicated for him and we need to make it as easy as possible that he can use public transport. Andreas is focused on smooth, efficient journeys. He is the business traveler. He chooses the mode of transport that’s most efficient and when he can work on the train journey and save time, then it’s fine for him, otherwise he takes his car. Our child Mila has also different needs from the other personas. She is twelve year old, has started to use the train by herself And so she needs to feel safe. And in a crowded station, for instance, or in evenings, it doesn’t feel very good for her to use trains, but that’s completely different to an adult. And finally, for our real estate colleagues, we have Eleanor, she represents the shop owners and the restaurant owners within a station. So you see a very large, range of different personas of different needs. And I said these are ad hoc personas not only representing, the majority of customers, but also edge cases. And therefore, we used more or less brainstorming existing insights to build these personas. You can traditionally build personas based on insights. So you do a quantitative research, you add a qualitative part, and then you have insights based personas, which are very high quality. And you can use them differently than our ad hoc personas, where we have a lower quality of insights. We have just what we had a lot of brain work from customer insights, managers, but not real research. That’s a bit different when you use personas. Let’s have a look at the use cases you have. On the left, you have the persona information, either ad hoc or insight based. And on the top you have the answers foundation, again, either based on customer insights or synthetically generated. So when you have insight based personas and answers based on customer insights, You can use personas for research light or digital twins even. This is ideal for pre testing ideas. Get the first idea which visuals work, which don’t, though, or for messaging, for slogans. It’s an early signal. It’s not certainly not the final evidence. It’s not something I would make a million decision based upon the this discussion, but it really helps you to shape your work and to get a first step. And when you have insight based personas and synthetically generated answers, you have more degrees of freedom. So the answer are a bit more creative and you can use it for innovation and creation. You can generate ideas, you can explore different concepts, different ideas, see how different personas with different needs react to it. And you can also discuss direction or take it as a conversation starter. And finally, the way we use personas at the moment, we have ad hoc personas with the synthetically generated answers. So we use them for culture and empathy. For our customer centricity program, so we can build awareness around customer centricity. We can help people practice empathic listening and just simply chat with a customer, not a real one, but at least with a persona. Customer centricity as a mindset is really important for us as SBB, and we’re trying to create a customer centric mindset all over the company. We if we’ve identified four crucial elements. You need to have a human centric mindset. As a railway company, we are very engineering based. We we have a lot of engineers. We’re have quite a technical mindset. So we need to change this to a human centered mindset. And we’ve identified two capabilities you need to have to be human centered. You need to be able to be, to listen actively, and you need to have empathy for your customers, for your colleagues, so you really can change something. Because it it’s not possible to do everything by the rule. You cannot have rules for everything. You need to be able to be empathic and then search for a solution. And in order to do that, you need scope for action. It doesn’t help if you’re really customer centric. You see the solution, you see the problem, but you’re not allowed to act. And so we work both bottom up and top down. And we’ve included DeepSights as well for our customer centricity initiatives. Bottom up, people can use customer insights, and they can chat with personas in order to have a first idea to explore first ideas. Top down, we use DeepSights for the leadership conference as well as for the leadership discussion. And how we use it, I will present to you right now. We have an annual leadership conference. There are about fifteen hundred management level participant each year, and each year has a focus topic. We’re very happy to say we have customer centricity for twenty six. That’s really, really great. So, we have a lot of attention on this topic. And traditionally, the topic is kicked off at the Leadership Conference. The program has revolved all around customer centricity. And we also had an exhibition booth, where management and participants could join us and ask questions on the terminal with DeepSights. They could also chat with DeepSights during the break. So, it was really a great, possibility for them to get to know DeepSights as a tool. Why do we do this at the management conference? Well, it creates a lot of visibility. It’s great when your top management, when your board members say how important or stress how important customer centricity is. And it’s it has really a strong signaling effect. We’re very well aware this doesn’t change mindsets. You have to do more. And in order to do more, we have the leadership discussion. Again, this is something we do every year, but with the focus topic, which is, again, customer centricity. So we have this discussion. Each one lasts about three hours. We have up to twenty sessions with twenty people and target group are line and specialist management. We want them to understand how important it is to shift perspective. We want them to give a possibility to practice active listening, to learn how to ask relevant question. It isn’t so easy to ask the right questions, to customers, in particular when you have real life customers. So persona is a great way just to start to learn how to ask questions and see what happens and how you maybe have to reshape your question. And we would like to reflect each and everyone on their own attitudes toward human and customer centricity. So in the end, they need to understand each and every decision counts and everything we do at SPB needs to improve the customer experience, be that in a direct way or in an indirect way. And with the leadership discussion, we can really help to achieve these goals. We use DeepSights in the prep stage, so before the leaders come to the leadership discussion, they have a task of discussing different customer experiences with different DeepSights personas. Maybe remember I’ve said we have a whole range of personas and that was important here for the leadership discussion. So you really have this feeling of different perspective, different needs. The personas give different answers, which is really important for our leaders to understand that there is not just one customer and one fits it all. After the check-in, will share their experience with the personas chat and once again reflect on the different perspectives on the different mindsets and needs of the personas. Because we still have a few misconceptions misperceptions. Yes. A lot of us are customer centric and a lot of us do want to work with the customer centric mind, but sometimes it’s a bit difficult. Yeah. And a lot of us think, well, I use trains on a daily basis. My friends use it. My family uses it. So yeah. I know what customers need. What customer needs? Yeah. True. But only for one customer, only from one perspective. And persona helps you easily to change that. We often hear frontline teams will tell us if there’s something wrong. Yes, they will, but that’s far too late. Then we really have a problem if frontline staff tells us we have a problem. So we need to be earlier. We need even when we design products, when we design services, we need to think about the customer, about the different kind of customers and needs they have. And personas help exactly to do that. So our colleagues can just chat with our personas without fear of making mistakes. There is no need for organizational effort to organize real customers, And we want the leaders to shift their mindset and to enable customer centric ways of working for their entire team. As I said earlier, it’s very important to have the scope for action. It doesn’t help to be empathic when you can’t really change anything. And we hope we will have a trickle down effect when we teach leaders to be customer centric so they again can teach it to their teams. And, well, the question is how does it work? Joe will show you how it works on the platform in particular. now let’s take a look at the personas application in the SBB platform. On the left hand side, you can see a number of different applications which are available, but let’s come into the personas environment where we have configured a bunch of personas on behalf of FTP based on their data. Let’s take a look at Leah, and we can see a quick overview of who Leah is and some of her key characteristics to kick off a chat with her. Instead, I’m gonna come back out and actually kick off a group chat here with a number of personas. So let’s click on Andreas, Leah, and Luca and kick off that chat. And we’re gonna start with a very simple question trying to understand a little bit more about how they’ve been using SPP dining cars in the past couple months. And now what’s gonna happen is for each of these personas, they’re gonna reflect upon this question and based on the system’s understanding of the persona and all of the thinking and data that’s gone into them, answer these questions in a realistic and human like way. And if we start to hone in on the various answers we can see here, well, let’s let’s highlight that Leah, yet while she has used a dining car, she looks at it through a sustainability lens, and she wants more local and seasoning sourcing. Andreas actually hasn’t used a dining car. He dislikes the food. It also breaks his productivity control. He stays in first class. And Luca, he also hasn’t used it. It’s mainly a price thing in his case, and he doesn’t feel like it’s designed for younger people of his budget. So let’s do a follow-up question to understand that better. And, the prices cannot change, we’re gonna say, but what else could be done to encourage you to use the dining car more often? Same idea now. Each of these personas is not only aware of what the others have said in the chat so far, but they have all of this understanding of the given persona which is being factored in here to the the answers that they’re each giving one by one. So Leah suggests that, you know, sustainability upgrades might help her. Andreas is looking for efficiency and environmental control, and Luca would like more of a youth friendly approach in order to start using. We can summarize that to get a better understanding to quickly take away and download with key quotes and copy, and then we’ve got that now packaged up and and ready to move along to stakeholders downstream, really taking full advantage of the persona environment in our workstream. So thank you for showing us how it’s done. We haven’t used yet personas for the leadership discussion as the leadership discussion will start soon. Our pilot team was very happy to use personas and it’s they gave us exact sort of feedback we wanted, which is very nice. We would have accepted other feedbacks, yes, but it’s really nice to see that it works. So you get a different perspective. You can clarify your questions, you can challenge your own preformed ideas and opinions. And you can really look at the question at the product from different angles. So that’s really great and encouraging. And we’re very thrilled to introduce it into our leadership discussions. So, you can use personas as catalyst for customer centric thinking. It’s very easy. It’s very hands on. It shifts the mindset. You can explore different perspectives. You can really practice active listening, you can create conversation starters. That’s also very important. Usually, you start discussing about the feature, about the product. And instead of then discussing a feature and how to implement the feature, you could talk about how personas think about these features, what benefits the different personas have. And so the conversation goes in entirely a different way and is much more customer centric. And it’s a first step towards engaging them real customers. And, yes, it’s important really to keep in mind the limitation of the personas. In particular, as we have ad hoc personas, which are not based on research, We have to keep in mind, it’s not the final truth. It’s plausible. Yes. And it seems to be quite realistic, but it’s not based on insights. So it’s a first step, but it doesn’t replace insights. Nevertheless, it’s really fun to use personas. You can be creative, you can explore different perspective. So just let’s have fun. And I would like to thank you for your attention. And over to Joe. Great. Thank you so much for that, Alessandra. Super exciting to see how you at SDV are working with the personas offering. So in conclusion, before we kind of go to the question and answer period, we just wanted to cover a little bit of what’s to come or kind of a roadmap look at what we’re working on with the personas agents. And there’s a couple different like larger topic areas that I have down here. Maybe if we skip to the last one, API support, depending on who you are, this will or will not be exciting for you, but we see a lot of the IT teams and so on in our customer base who are working on tools that can communicate with one another, a strong excitement and realize we have AI support. Long story short, that means that other different tools, for instance, working on innovation ideas or campaigns, etcetera, at organizations can be communicating behind the scenes with these personas, getting feedback and wrapping it into those applications. And that’s all facilitated with the API. So that’s a really cool offering that’s coming out, or actually it’s gone live this quarter, to speak, and one customer is really leveraging Back to the top, image generation. So I mentioned that earlier on. In coming weeks, we will be releasing the ability to generate a variety of different images with these best in class image generation models, which many of you have probably seen demos on and so on. So imagine from a persona chat, you could now generate a storyboard or an infographic or packaging and so on using these models. So that’s something that’s really exciting and we hope all of our existing customers will get a lot out of. A couple other areas. Yes. AI moderation is coming. So imagine rather than needing to facilitate the entire chat with one or more personas, you could just brief the system, select the persona, and have it actually carry out the conversation for you. And not only would this be convenient, we think it actually would bring a little bit of rigor into some of the conversations, investigations depending on the user who’s using the system if it can be configured to the way your organization works. So AI moderation, definitely exciting. The on the fly persona generation piece, spoke about a little bit earlier. The idea is that in addition to these persistent personas that we set up, like the SBB ones that Alexandra took us through at the beginning of her presentation on the personas, we would let users come to the system, prompt or filter, and then generate a persona from the knowledge base that’s really specific and niche to the particular task they wanna carry out at that moment. So that’s a very interesting area that we we think will take off, especially in coming quarters, and we have some pilots on there right now. And finally, integrations with third parties. I mentioned MindLine. Next Atlas is a great provider that we’re partnered with, and we’re able to go to their content now and bring it back into the persona chats soon to release a couple other such integrations. So I think the ability to go to other tools and so on where different types of data sits, whether it’s social listening, foresight data, etcetera, and have that be integrated in a seamless way for the user in the personas environment is something that will be key and that we’re really pushing on. And with that, I think we wanna Maybe back to you Maria, and we can kind of take this home. So great, thank you everybody. Thank you so much. Thank you so much for the presentations, Alexandra, Michael, and Joe. We’ve already started receiving some great questions for the audience and keep them coming. I’ll start with the one for Alexandra. Alexandra, if you were advising if you were advising another insights team starting with personas today, what would you emphasize? Good question. Well, I think I would start with defining the use case. So why do you want to use the personas? What is the goal you want to achieve? And based on that, you can proceed. Either you have to find insight based personas so you can do your search light case, or you say, well, we want to use it in a for improving customer centric mindset, then you can use ad hoc personas. You have to keep in mind, your output is only as good as the input is, as it is always with IT and artificial intelligence. And maybe you think about the communication as well. We’ve made the experience that personas are very well received, so they get a great attention. But there is also a danger as people not always understand what persona can and cannot do. So you need to think carefully who is the target group. What do I need to explain to them? How do I explain to them what personas are and how they can be used? Thank you so much, Alexandra. And the next question goes to Joe. Are the personas assessing the visual itself or the associated text summary of the concept? I think it goes to one of the slides that you were referring showing. Yeah. Right. So sure. I think the ability to take a look at these images and then provide feedback on them, be it concepts or ad campaigns and so on, is obviously something that a lot of customers really really focus on. And indeed, we use multi model multimodal, that’s the terminology in the AI space, models, meaning we’re not just text describing the image, so to speak. We’re really passing the image through to the large language model, and it’s actually assessing the image. So when it gives feedback on packaging or a campaign that it’s seeing, it’s really taking a look at both the text, but what’s happening in the image, the color scheme, etcetera. So I think there’s something that’s really strong in how we’re implementing those models. Thanks, Joe. And then there’s another one for you. When when you do this group chats on the do the participants interact with each other at all? Do they build on each other’s input, or do they just answer each one at a time as through it’s two or more in-depth interviews? So do the participants basically interact with each other and do their answers impact each other’s answers? Yeah. Sure. Good question. So indeed they they do. So, you know, as you saw there in in the demo, when the personas are speak are answering, they’re not just individually answering the person who’s asking the question, you as the user. They’re actually able to see one another’s responses, and they’ll build on one another if it makes sense. So that gives a nice dynamic that we we find is, you know, more realistic to how actual group group setting would work. You can also, of course, speak directly to just one of the personas if you wanna hone in on what they’re saying within a group chat. So yeah. That’s good to know. And we have a great one for Alexandra. Alexandra, you talked about ad hoc personas and research like use cases. In which in which situations are this especially helpful? And when would you still insist on deeper research? Well, there are a lot of use cases. Think it’s easier to say when I wouldn’t use the research light product cases when I have a million dollar decision. Then yes, I would do some additional quantitative research. I maybe would use the personas for getting first ideas what to ask, which topics to cover, but then I would do research. Whenever I have a first idea, like pretest ideas for a campaign or for a product before I actually start work, then yes, Research Light is a really great case. And I think it’s not so much about replacing research, it’s about doing more research. It’s doing research when I previously wouldn’t have done research because it’s too expensive, too time consuming. So I just can ask a persona, and it’s better than nothing because I’m getting into this customer centric mode of working instead of just thinking about the features. And as I said, you can also use it as a conversation starter, and the conversation is completely different like this than when you just have your features and your solution list. That’s a great perspective. Thank you so much for sharing. And we have one more before we have to wrap up. This one goes to Market Logic. Joe, could you talk a bit more about on the fly personas? What types of content is being used? Can existing personas also be updated with that technology? Yeah. Sure. So just to stress this, this is not what SBB is using and not what we largely spoke about today, but it’s something I mentioned as a roadmap item and we’re receiving a lot of excitement from some customers and prospects. And the idea here is, one, either looking at the DeepSights repositories, remember we’re holding potentially thousands of market research reports and other types of data on behalf of our customers in our environment already. Or additionally adding in survey data, so take your usage and attitude survey work that often underpins occasion or demand space based programs. And we can then allow users to type in a prompt, hey. I wanna speak to, like, this niche persona and go to those data sources, the structured survey data, the DeepSights repository of content, and comb them in order to to generate these personas. So that’s the two types, let’s say, data sources that we’re really seeing. This is, again, something that’s coming that we think will be a strong play for all of our Personas customers, which is the ability to take existing Personas. So now here I am referring to the ten SPV ones, for instance. And if you wanted to take them to another market or age cut or whatever be it that you want to have a different version of the Persona, you could sort of take that existing description and on the fly change it for a temporary persona that you would then speak to, but always keeping those core personas there to to underpin the the the real research and backing. Those are the directions we’re looking at here, and I think both are are pretty pushing the envelope, let’s say, in terms of what Jenny and can do. And, you know, we don’t see them on the market currently, but exciting for us. Cool. Thank you so much for sharing. Thank you so much all for your questions. If we didn’t have time to answer some of some of your questions, we will follow-up separately after the session. Also, feel free to reach out to us in case you have any further question to our guests. Thank you so much, Alexandra, Michael and Joe for joining us today and for sharing your experiences and perspective. I wish you all a great day. Thank you so much. Thanks, guys. Take care. Bye bye.