How Mars is Using AI to Supercharge Market Insights 

Welcome to our webinar, how MARS is using AI to supercharge market insights on behalf of both Market Logic and our partners at MARS. My name is Carolyn Woods, and I will be your emcee for the next forty five minutes or so, as well as the person to answer any questions, housekeeping, etcetera. So we are currently in the introduction part. I will certainly be handing things over to my colleague, Jennifer, who will go a little bit into the history of Mars and Market Logic. Obviously, we have a lot to discuss about where they are today with their insights and knowledge management journey. But, of course, this didn’t happen overnight. So she will be providing a little bit of that story, some context to bring us up to the present day, at which point I will be bringing in Matthew to discuss where they are with their Market Insights journey, how the integration of new AI tools is transforming that, accelerating that to a certain extent. And then, as I said, we are moving into an audience q and a where you can get your own questions as asked and answered. So you see them to the side of your screen. Give a little wave, Matthew and Jennifer. From Mars, we have Matthew who is the global digital commerce customer in shopper insights. And he’s been at Mars for several years and has a wealth of knowledge that I’m excited to have him share with you today. And then on the Market Logic side, we have Jennifer McMahon, who’s the account director of customer sir success, and she has been working with Mars specifically for several years. So she will be able to offer our agency side, and view of things. So we’re really excited to be sharing the story with you guys today. Obviously, if you guys have joined our recent webinars, there’s a lot that we have been able to share just about our product, but it is always helpful to be to to see that in action and see from the customer side of things, how they’re deploying it, best practices, initial feedback of these new technologies, and of course, where we believe this industry will be going. So just one more slide from me, and then we’ll get the show on the road. If you if this is your first time, at a Market Logic event, I will just say a quick overview of who we are. We have been around for over a decade now. We’ve been in the insights management and knowledge management business, helping innovative companies run insights driven operations. So we work with a range of large businesses and enterprises who are global enterprises, as you can see here. We also have a whole host of integrated partners offering syndicated content, news, etcetera, to really build out this insights ecosystem. So we have, of course, been in the game for a long time. And then with the introduction of generative AI technology, we’re taking our suite of solutions to the next level with DeepSights. So I will now hand things over to Jennifer. She will tell you a bit about our product, the history of Mars and Market Logic, and then we will bring in Matthew. So without further ado, I will hand things over to Jennifer. Awesome. Thanks, Kelly, and nice to meet everyone. I’m Jennifer McMahon. I’ve been at Market Logic for about six years now and have had the honor of working with Mars the whole time I’ve been here. So I thought I’d kick it off with a brief introduction about DeepSights since it is a hot topic and will be discussed a lot in the interview portion. So I believe most people on the webinar know what DeepSights is, but just in case, DeepSights is an AI based solution by Market Logic that helps organizations unlock the value of their proprietary knowledge and insights. So users can ask natural language questions and receive answers drawn from their reports, studies, and other documents, making insights more accessible and actionable. So after asking DeepSights a question, the tool will provide one synthesized answer across all the selected sources that you deem relevant for that question. You can also create a report that summarizes key findings into one concise document, which is really handy to use, outside the platform for some of your presentations. Deepsights can also be directly integrated with other business apps like Microsoft Teams, Google Chat, or Slack. And we have an API available that feeds answers into business applications and workflows, so there is an opportunity for DeepSights to connect to your other, AI tools that you might be using internally. So outside of being an in platform tool, we have seen DeepSights be helpful in integration with other AI tools across our clients as well. Just a brief note on our partnership with Mars. We’ve been partners since, two thousand and fifteen with our first instance of our platform. And over time, we’ve added all of the major business units to become a truly cross segment enterprise insights management tool, which we call Synapse. Then in twenty twenty one, we successfully won a Mars leadership initiative called Zero Waste, which was a push to accelerate marketing and insights transformation across a lot of Mars key pillars and to build new capabilities and become more innovative. This was a really exciting campaign to be a part of. We actually have a webinar on this and some more information on our website if you want to get a deep dive into what the stairways initiative was about. But it was a great thing for both of Market Logic and Mars to be a part of because it really helped us push our boundaries on what we can do together and what we can accomplish from an insights management perspective. So after the Zero Waste campaign, we continued to launch upgraded versions of the platform, added additional syndicated sources and connectors, and continued to see incredible asset and usage growth. And then finally, this year, we’ve launched our much anticipated AI assistant called DeepSights, which has been a great success so far. We’ve already released three new feature enhancements since we launched in February, and we have a plan to launch a handful more by the end of the year that we’re really excited about. And I think we have another webinar coming up in a few months with Olaf to talk about all those exciting new DeepSights enhancements that we’ll be launching through the end of the year and early next year as well. So a little bit more about the Synapse platform itself. There will never truly be one place that you go to for all your insights, but we like to say Synapse is pretty darn close to being a one stop shop. We have made big strides in connecting and capturing the information and data that really matters for Mars. Today, we could tout an incredible knowledge asset of seven thirty one million. That’s actually through Q3 of this year. And it includes primary research and insights from Mars plus content from our syndicated partners that we have listed here as well as some connectors that we have built into the platform as well so that all the valuable information and data flow into the platform easily. I work closely with a team of champion ambassadors like Matt to make sure that we continuously bring in all the information that is deemed valuable for decision makers so that the platform continues to be valuable and useful for not only insights curators, but also for the business folks who are making the decisions. I’m going to end my portion here with a little bit about the DeepSights approval process. So like with any new AI tool, getting DeepSights up and running was somewhat a lengthy process, very thorough process, which in this situation took us about six to eight months with Mars. First, we had to pass several security and architecture reviews. We needed to prove that our infrastructure was secure and show in detail how our large language models extract insights. We also needed to prove that our AI models are not trained to learn facts from Mars data specifically. So we have a great team that puts together quite detailed information. I don’t think there’s anything that anyone can throw at us in terms of a security review that we haven’t been able to successfully pass, but it does take some time to go through the full process. Then the Mars Champion team had to put together an official project charter for internal approval. So this contained information like the scope of the project, the business value, the estimation of resources, the timeline, and critical success factors. So what makes this project successful, and how do we measure success over time to continue having it in synapse. Once the project charter was approved, we moved on to IT, legal, and procurement alignment to make sure that all of our paperwork matched up appropriately with the security needs and legal needs and procurement needs at Mars. Finally, there was a final review of all materials by something Mars has called the responsible AI council. Needless to say, all the review and steps went well over time, and we officially launched DeepSights in February. So, again, happy to answer questions, or Matt can help with this a little bit more, but not unusual that we go through a process like this. This might have taken a little bit longer with some additional steps at Mars, but I think for the most part, this is quite standard. So that’s all I wanted to give at the kickoff of this today, so I’ll pass it over to Kelly and Matt for the more exciting interview part of the webinar. Alright. I will bring Matt on stage. Thank you, Jennifer, for getting us caught up to speed. As she shared, we have been with Mars for almost a decade now. So now the question is, you know, why am I sitting down with Matt? What is, this new development with their insights management and market insights process that that really, requires this this time to to tell the story. So obviously, lot has changed since February and the launch of DeepSights. But I think let’s just let you introduce yourself first, Matt. So thank you for joining me. And before we dive into the technical side of things, do you wanna just share a little bit about your role and position at Mars? Yeah, of course. Hi, Thanks very much for the invite. Great to be here, and, yeah, good morning, good afternoon to to everyone on the course. So so I think why everyone’s here probably most people know who Mars are, first of all, but, you know, just a quick bit of background. We’re obviously a large FMCG company. We’ve got tens of thousands of associates around the globe, so we operate in all the regions in in many, many markets. And also, I think what’s important as part of the story is that we also have different business segments. So we have a pet care, pet nutrition segment, we’ve got a food segment, and where I sit is within what’s Miles Wrigley, also a confectioner in the snacking segment as well. And that’s quite important as we go through as well. And in terms of the team that I’m in, as self, I’m in a global, we call it customer shopper insights team. So we have this global remit and then so we serve lots of teams around the business, but we sit within sales and focus on the shopper, but we have to work with global teams, local teams, different functions like marketing and innovation as well. So it’s sort of spreading it that way. And then suddenly my role is focused on the digital shopper, so really just working. So I suppose newer teams as well, but really important to upskill different teams with their insights and expertise as well. I’d say probably why I’m here, so thanks for the invite. It’s really I’ve been at Mars for a number of years, and actually my Mars career goes past even the Market Logic sort of partnership as well. But I’ve done a number of different insight roles, and it’s really been about leading expertise in certain areas. So I’ve led from an advertising development perspective, from a brand strategy and portfolio perspective, from a forecasting, and now I’ve been focused on sales on the shopper for about seven years. But for me, well, where I can add value but also where insights functions add value is really through that expertise and perspective. That’s where we can drive value to business, to drive that growth, to really sort of push it and add value versus what people know today. So, you know, given a different nuance, but with that consumer and shopper lens. And so I’m really energized by that. And that’s how I’ve signed up as, like, a knowledge ambassador, Jennifer Torso, knowledge champion. And so I’ve been looking to drive that. Yeah. So thank you for that. And and obviously it is a huge enterprise with a lot of different teams and and brands and data silos, I’m sure. So we are having your perspective as a knowledge champion, which I think is very relevant to today’s topic. But I know that that’s just part of the story. But from your perspective, can you give a little bit more of an intro into what your team’s specific approach to knowledge management and consumer insights is? Yeah, definitely. And I’ll talk about probably a couple of levels. So there’s the big mark or business level, so cross business level, but also we’re in the team. But say at across business, like the zero waste initiative that we highlighted a few years ago, we really focused on actually realizing that launch management has to operate what we call an enterprise level. So if just one team does it or a group of associates does it, then it’s not going to have any impact. It’s just going be a waste of time or that team might be slightly good, but it’s not really going to drive the value. What we really value as a business is understanding that the consumer. We want it to really be at the heart of our decision making. We want to make sure that we’re serving products and services that really serve the consumer. And so to do that, we really do need insights to reach as many associates as possible to really drive that impact as well. And that’s where knowledge management really comes. And so that’s why the partnership actually with Market Logic is a cross segment partnership. So it’s not just driven snacking, but it’s with pear and it’s with food as well. So it’s something that we look to drive at a business wide level. We do that by sort of identifying those ambassadors and champions that are really going to push and drive it. So we meet regularly on at least a monthly basis. We check-in with Jennifer, and then we’ve got very clear objectives and targets that we’re looking to do at an enterprise level. And so zero waste taps into that, but looking at how do we reach. And then within my team and smaller teams, actually, we look at how we drive performance of the insight teams as well. So not just on what we do at a day to day level, how do we do good research projects, but what we look at is how do we become as a high performing team. Knowledge management is one of those objectives that we work together on and then really drive. And so elements that we look to set up are the right infrastructures and environments, the right cadence in terms of our sharing within the team, but also how do we share with other communities and forums as well. That’s really important for us as well. Yeah. And as you just kind of alluded to, and Jennifer mentioned, this is not something that just happened overnight. I’m sure you’ve been improving these processes little by little since we began working together. But now let’s kind of pivot to the topic at hand, which is DeepSights. Again, this is an AI tool that sits on top of Synapse, your knowledge management platform. And you’ve had it deployed at Mars for several months now. So I’m wondering if you can offer any insight and initial feedback into how DeepSights specifically has changed your approach to knowledge management. Yeah, definitely. So I have to admit, the hype and just being able to say AI has certainly helped. So having something shiny and excitement, that’s probably why we’re all here on this call. But certainly that’s opened a few doors and it’s got a few more people attending those meetings about noise management. That certainly helped. I think in terms of the approach, it’s not yet changed the approach that much. But what it’s done is is really supercharged the effectiveness that we have with our insights and the knowledge management as well. And the way that’s shown internally is I don’t know if we’re allowed to say this, but I describe DeepSights and our platform as being better than Google. So I know we go externally. There’s a lot of, you know, there’s always a lot of nonsense. There’s a lot of noise. You know, there’s a lot of stuff that’s not directly relevant. What’s really exciting is that, you know, we know that quality in equals quality out. DeepSights gets to really maximize what we’ve already created and the knowledge that we’ve really invested, and so it’s really helping us build that intellectual capital and really maximizing it as well. I think the other term, one of the terms that we’re really drawing is democratizing insights as well. So it’s made knowledge management, it’s taken us beyond just being an insights team role or a toolkit that they use. So in the past perhaps platforms have been researchers looking for documents using keywords to search. Actually the natural language element opens it up to lots of different teams. So it opens it up to marketers, to sales teams, people within finance, people within R and D as well. So you can search for what are my opportunities, what are the gaps, what are the trends that I need to look out for. And that moves the role of insights and knowledge and makes it much more exciting and much more engaging for lots of people as well. I think also it’s really just accelerated that zero waste sort of platform. So when we invest in insights and research, we don’t want it just to be for this year and then it moves and it gets lost in somebody’s hard drive or just disappears into a folder somewhere. We need that cumulative expertise. We want to build that knowledge asset base and make sure that we keep on building to really expand our knowledge as a business and expand our knowledge as a basis as well. And so that’s where it’s really sort of changed it and meant that it’s not just a file holding system. It’s somewhere that is actually a tool that’s really exciting for everyone to use as well. Yeah. I think what you said about the hype of AI, I think it’s exciting also for us to to be able to speak to people who are might not be that engaged with insights management, but suddenly see the ability to, for instance, surface data points from research reports that had been, you know, gathering dust on a shelf somewhere. Suddenly those are useful. You know, these are really conversations that are happening a lot more that we love to be a part of. But then, of course, with the the hype around AI, there’s been a lot of investment and a lot of large enterprises have, you know, decided to try to build tools internally. So I’m wondering, from your perspective, now that DeepSights has been deployed at Mars, what were the considerations that made you decide to go with a specialized tool like the one we at Market Logic had rather than build your own as we’re seeing at maybe some other enterprises? Yeah, definitely. So I think there’s a couple of things, but really, first of all, it’s the pace of change and getting that expertise to make sure that we get to the right solution. And that’s something that we’ve really looked to develop and drive. Perhaps, you know, a long time ago, we would have looked at trying to get everybody a solution that’s suitable for everyone and probably the cheapest possible option. But what we’ve really understood is there’s a lot of great technology and a great and we can use technology and external expertise such as Market Logic was a great enabler as well. We realized during our initiatives that we’re sitting on essentially a gold mine of insights and a gold mine of knowledge. And so I’d say it’s relatively off the shelf. I appreciate there’s a lot of work behind the back seat, back scenes, but essentially taking this platform and directly embedding it has really allowed us to deliver value straight away and really drive a lot of solutions already. We’ve got, you know, we’ve really already increased the number of users, we’ve increased the number of uploads, increased the number of use cases where it’s been taken to customers that will be embedded in strategies as well. That’s really unlocked sort of a lot of immediate value for us as well. But also in terms of we’re quite a principles led business as well and it really taps into a partnership, into a couple of principles. So one, we’ve got a principle of mutuality. So when we do well as a business, we want to do well with our consumers and our shoppers. We want it to work with them. We also want it to work with our partners such as Market Logic as well. So that’s one of the reasons and great reasons of this partnership is that we know that we’ve been very proactive in terms of driving innovation. You listen to our ideas, our feedback, and you’re proactively working behind the scenes. And with DeepSights, we’re still getting new innovations, so that’s amazing. And so that’s the thing we want to grow with our partners as much as possible. And you know obviously we’re challenging. Think some of our human process is quite challenging, but we also want to make sure that we’re working together to get to great solutions. Yeah. The other part of that I’ll just touch on is responsibility. So that’s something that we’re really important to us in terms of doing things in the right way and perhaps we’ve been a bit risk averse in the past and we’re probably still quite risk averse as a business. But we have set up a couple of years ago a responsible AI council that Jennifer touched upon And this is a team who are responsible just making sure that we conduct AI in the right way. So of course from a corporate reputation, we obviously don’t want our secrets or an intellectual capital to seep out, go out to the market either. But actually, the key way is also knowing that we can work with you to deliver great results. That for me is the key part of responsible AI is that we’re using it for positive reasons. So we want to get good products and services that really understand the consumer and shop as well. And that’s what we’re looking to try and deliver. And that’s what AI is really great. It’s not just about doing things quickly and cheaply. It’s getting better answers, really bringing our knowledge together and really driving it as well. Yeah, yeah, I think that’s definitely what we’ve seen as well. I think the first selling point is always, you know, efficiency and time to insights. But I think as the AI develops, we’re also seeing just massive improvements on the quality of the answers. As you said, this is not a static product that we sold you, but something that will continue to develop as as the technology does. Moving on to drill down a little bit into, you know, how DeepSights, works on a day to day level. Can you shed some light on how DeepSights addresses your specific stakeholder needs, whether that’s, you know, experts within the insights and, knowledge management teams or end users within, you know, business roles, marketing, etcetera. Yeah. Definitely. And and before I do a proper answer, I’ll give a little bit of a surprise and, like, a special like a nice surprise that we have. So I think many on the call probably work within insights teams and we probably get lots of emails, lots of requests, you know, via email, via Teams, or via the office as well. And often it’s just like looking for the magic number, like can you provide this data point? And that’s what I love about DeepSights is that you can just type it in there and it will find that number as well. Love that as a way it’s got lots of great uses as well as these great engagement reports. But that really taps into how we address our stakeholders. So for me, you know, a blocker that we have as an insights team is that we’re limited by our time, so how much time that we have. We all feel sort of rushed and are limited in time. But also in terms of what we actually know and what we’re aware of in terms of what projects have taken place and what to search for. So if a sort of relevant deck in Petco has happened in another market beforehand, you know, I probably wasn’t aware of it and I’m probably not going to end up finding it during my search. But DeepSights is really amazing at connecting those, we call them insight silos. It closes down those silos, first of all, that really connects those insights and that really helps to build that intellectual capital, getting to really much more powerful, much more informed answers and really great expertise as well. I think the other area is really around accessibility as well. And what I think is great for stakeholders is actually they now get to reach really experienced, clever insight directors twenty fourseven. They don’t have to wait for me to give an answer, which can take a couple of days or might even ignore it. They can email DeepSights, either on the platform, within Teams, whenever, wherever, any part of the day as well. And then so it’s really easy to get those instant results. But also the other benefit of accessibility is that it’s the natural language really moves it away from keywords into actually what you actually need. So you’re already shaping the answers in terms of what you actually need, so be it the opportunities, the gaps, your particular question. And that then enables other teams and stakeholders to really sort of use it as well. So again, you could be a marketer, you could be someone in research and development, or you could be someone in sales, you could be a junior person, you could be a top level manager or director within the business. The natural language really improves the search and really improves that viability and so that really broadens the stakeholder base as well, so reach many, many more stakeholders with this rather than just being a tool that the insight teams use as well. And so that’s example of how we drive value You know, to be more strategic, that’s like the vision of most insight teams. You have to move beyond your database stakeholder group. You have to drive the enterprise. And DeepSights really enables us by reaching more people, driving accessibility, but really getting, you know, essentially smarter, better answers by closing down those silos as well. Yeah, and I think that segues very nicely into my next question because obviously, we have this very accessible tool that you can you can access, whether it’s on Teams or email or any other the API functionalities. But saying it’s accessible is one thing, and then actually driving your colleagues and then people from across teams to access it is a whole another challenge. So I think especially, on the enterprise level, this is a really, really interesting topic. And I’m wondering with you as a knowledge champion, how are you, individually or within your team, how are you driving the adoption of these new tools? Yeah, definitely. So we talk a lot. A phrase we use is in terms of innovating with an eye of scale. So I’ve said in the past, perhaps we always try to find the right solution before that pleased everyone, but we’ll end up with the cheapest and not necessarily the best. What we try to do now is be much more focused on use cases, but at the same time make sure we’re aware of how can we scale it up. So teams will take leadership and sort of practice, you know, think about a specific use case to drive that value, prove that value, but always think about can it be scaled into other markets, regions, or other use cases. And that’s really an approach that we’ve been using with DeepSights as well of really honing in on different stakeholder groups, tuning our communications, just saying this is how you’d use it, these are example questions, this is your best way of reaching it as well. So we’re doing it sort of group by group so it does take time but what we find is that when people, when you get those curious people, those enthusiastic people starting to use it, they become the ambassadors as well or unofficial ambassadors. And so we’re getting them nice and we’ve got very positive trends in terms of usage as well. And so not quite a snowball but hopefully we’re getting there and the knowledge management has been embedded across the business and we’re getting a lot much more energy as we go on. Yeah. And of course, as I said at the beginning, this AI sits on top of the knowledge management, so it wouldn’t be as impactful or as usable without the work that you guys are doing to make sure that knowledge management foundation is really robust and relevant. So in that the AI is kind of an agent and a helper to that, an accelerator to that. Obviously, we’re still seeing the adoption statistics, like, pick up and people get more comfortable with using AI in their daily workflows. But, just to wrap up this portion of the event, before we move into audience q and a, I was wondering if we can just look a little bit into the future and just from your perspective. I know you don’t have a crystal ball, but, how are you seeing the future of this technology impacting the strategic role that insights teams play within an organization, whether it is just, you know, elevating knowledge management or change management or so on? What would you what would you hypothesize at this at this point in time? Yeah, so I’m obviously very excited about it. I I enjoy using it, and it’s been a great tool, and I think it’s gonna make insight teams, you know, fast moving, quicker moving, much more engaging and and better informed. So I think that’s got to be a good thing. I think the key way it should enable us to enhance our value to the business is through curation and storytelling. So it manages those day to day questions and they can get managed within the business, but where we can really play a role is in terms of presenting it back to the business, using the knowledge assets that we’ve built, but then really communicating this is what it means for us as a business. You know, here’s all the great knowledge in terms of what we need to do now, and that’s where we can really add value, as well as quickly understanding what enables us to really understand what do we have at this moment in time, and again, mapping that path for the future. We can say where those gaps are and where do we go next, but we’ve avoided all that duplication and tidying as much. So I think the speed, the pace, the different perspective that we can get out of it rather than just being delivered files it is really gonna improve our storytelling ability as well. So I’m looking forward to it. Alright. Well, with that, I know that we have a lot of questions now, so, you can stay on the screen. But I think that was a great snapshot into, you know, how Mars is using these AI tools. Again, this is a conversation that we’re all sort of having in real time together. So I’m sure if we reconvened in just a few months, you’d have even more to share or more insights. But it is it is a very exciting topic. So with that, I think I’ll bring Jennifer back on, and we will move to the audience questions. We have about ten minutes. So if you have a question and you haven’t asked it already, by all means, just use the tab in the bottom right hand corner. So I’ll try to go in order, but let me start with his question at the beginning. And I think maybe, Jennifer, if you’d like to take this one. It’s about product differentiation. So obviously, Gen AI, that’s a word we’re using a lot. And the national language sort of prompt and answer model is something that people have seen before, whether that’s with Copilot or ChatGPT. So can you just explain what you think differentiates DeepSights from its competitors? Yeah. Not a problem. I think the the biggest thing, another a lot of little things, is that DeepSights is grounded in customers’ trusted and proprietary content. Right? So our models are specifically built and trained on market research and insights specifically. So it’s within the realm of our insights management platform. As a general rule of thumb, most of the content that’s available in the platform is quality. It’s approved. You search and you find the information and you know it’s valid. So we don’t really bring in any evidence outside of that quality approved content. There’s no hallucinations. We’re not interpreting data that’s from the outside or that’s not approved by the insights organization. So it won’t hallucinate anything if we don’t have content to back up an answer. It simply will say, we’re sorry. There’s not enough content here to provide a strong answer. So I would say it’s building on insights and market research specifically and the walled garden of that quality approved data that really differentiates our AI tool compared to some other great AI tools that probably are just used for a bit of a different purpose. Yeah. Yeah. And this is not to say that they don’t have their own use cases, but, obviously, within the market insights industry, you know, people who work, within there see themselves as guardians of, you know, the truth. So if there is gonna be an AI tool that works with them, then it also has to, you know, be something you can trust. So, let’s jump up. Alright. We have one, Winnie from Dyson. So this is going back to, the question that we asked you, Matthew, about, scaling adoption. And you said that you’re working from sort of a use case model where rather than just trying to get as many people to click into the platform as possible, you’re trying to deliver specific use cases that have a value prop for different stakeholders. So can you offer any more or, like, an example of that or any more specifics about how you choose what use cases to to develop and drive? Yeah. Definitely. Yeah. Thanks, Rini, for the questions. So I I think really a top one is in terms of identifying top, like top trends or key trends that for your business area. So I obviously fix on the digital, let’s say what the top trends that we need to think about digital online shoppers. Where the value really comes onto us. Obviously, I think we’re all used to receiving lots of external debts from the big named agencies, but what DeepSights really does is overlay our Mars knowledge on top of that. So it really makes it relevant, really makes it useful and impactful as well. So it does that in a way obviously overlooked to that anomaly, but it does a great starting point in use case. We’ve used these outputs in terms of taking it to customers and to retailers directly. We’ve used it directly within workshops as stimulus and inputs as well. That’s a really practical way, and I’d say quite a valuable way for us as an insights to really take. I’d say a couple of my top three. The other would be those missing data points. If you’re trying to say what percentage of people like snack healthily, you’re going to find that answer pretty quickly rather than looking through all the different snacks. So different depths, perhaps those sort of quick data points actually saves a lot of time. It’s a quite scary time. I say it that way. And then the other is actually, so I find out a lot of insights of hidden away index. So say if I did a review in the UK market in different ways, there’ll be insights in different areas such as perhaps healthy snacking or treating yourselves in different categories. And beforehand, you’d really need to understand what you can find where. The DeepSights or the AI search really has helped unlock these or find these insights that are hidden away as well. So you know, you might be sure that you’ve done it before, but you might not be unaware, but actually just finding, getting that true value about everything you’ve done is really valuable as well. Yeah. Thanks for that. And again, I I’m sure in a few months time, you’d have even more use cases that you can share, but it’s a good snapshot. Let’s jump up to Michael’s question. So maybe, Jennifer, you can start with this, but then also, Matthew, if you have anything to add, you guys can maybe share this question. But, since deploying DeepSights, did you notice a change in behavior among the end users in terms of how they use your insights portal? Have you seen an uptick in, like, let’s say, marketers or other business users switching from regular searches to DeepSights searches? I don’t know if Jennifer you’d want to So I’ll kick it off since I I have all the reporting side of things. And then, Matt, you can have some real life experiences about maybe how some more marketers approach you differently now. But we have our core search, and then we have DeepSights. And, I think a hypothesis was that search would go down, DeepSights usage would go up. But actually, like, search has remained pretty consistent just because it’s now has a different use case. So I feel like people use search to explore if they don’t know what they’re looking for. They want to see what content’s in the platform. They use DeepSights when they have more of a focused question. And what’s great about DeepSights is that it provides source evidence. So we’ve seen some document interaction changes, right, where maybe there’s not as much interaction after they find things in search, but there’s more interaction when they find things in DeepSights. Right? Because, it opens up their mind to explore the topic further, then they go back to search and find other content. Or DeepSights might open up a few sources or projects that they’re like, wow, I I had no idea this was related to this topic. So surprisingly, search usage was flat. I will say we’ve seen some uptick in in the business users and the marketing marketers usage, especially with our MS Teams connector that we have. We find that people will ask DeepSights questions in big meetings and get answers directly in the chat, which we’ve seen great usage too. So I think it’s the ways of working have changed a bit, and so we try to be sticky where those folks are that want those answers quickly. Yeah. I’ll just build on And then, Matt, I don’t know if it’s Yeah, just to build on The key thing is we’ve got more people using the platform. So we’ve got more energy, more excitement, and so that increases actually the people using the call search, the old keyword search, and of course the people using DeepSights. And we’ve got more people using it and we’ve got more users spending more time on the platform as well. So I think in the past, if you’d have somebody going on to Synapse our platform, they would spend like a minute to give up and then they’d come and email me. But actually now people are getting used to DeepSights, they better understand the platform and the uses and so they’re investing a little bit more time as well and so that’s protecting me a little bit, protecting the insight teams a little bit, that is really making sure their insights of being used by by more people as well. Great. Yeah. And I think we have just time for one more question. I don’t wanna keep us over our allotted time, but, I think this is also a good place to end it because it talks about just our responsibility towards AI. And I know that we have the or Mars has the responsible AI council. I won’t make you go into that too much, but, you mentioned that there is a there’s a connection between quality input and quality output when you come to these natural language processing machines. So I’m wondering how is Mars or your team approach this either the responsible use of AI or coaching people on how to prompt it effectively. Is that right now a part of your, you know, adoption drive with DeepSights or is that still being worked out in real time? So we’ve got we’ve obviously done a lot of checks to make sure that DeepSights isn’t leaking our knowledge elsewhere so that’s important. I think it’s a very real concern that I’ve heard from a number of my colleagues actually is that they’re actually scared of uploading their documents in case DeepSights gives the wrong answer so it doesn’t give the recommendation that they’re looking for. And so I see it much more as a change management piece in terms of yes, if you control your document and you only share it yourself then of course you can have a very clear recommendation. However, if no one sees that document, then literally no one is going be listening to that recommendation. So we’re doing a lot of trying to do reassurance. We’re trying to highlight those positive use cases and demonstrate where answers make sense and we’re trying to sort of identify where those concerns are, know, listen to them, understand them, but also try and encourage them to try it. We have platform the ability, if you don’t like the answers, you can indicate sort of highlight that, so I think that’s helping, yeah, it is an ongoing change management and people are concerned about the answer not being right, but I think the more knowledge we have, we’ve to trust it in some ways as well. Yeah. I think that’s a good place to end it. Obviously, we can have a whole another webinar about, you know, change management and obviously, this hesitation around AI, I don’t think that’s DeepSights specific. I think it’s just generally within the industry. So it’s good to have champions like you, Matt, who are, you know, nudging people in the right direction. And I think in general, this is this is a process that will make the whole enterprise a little bit more connected to insights, which is, in the end, what we want, whether that’s through AI or other technologies. Alright. So thank you everyone who participated in the q and a, who tuned in. I will just do a quick wrap up. But thank you again to Matt and Jennifer for sharing their expertise on this topic. This is an ongoing conversation, so I will, of course, just plug some in some future events, and then I swear I will let you guys go. So, obviously, we heard from the CPG side of things. If you are interested in hearing how this is being used on the retail side of things, in about two weeks, we have a conversation with Reva Group. And again, this is about how they’re taking their knowledge management foundation and then improving it through the incremental deployment of AI tools. And then if there’s anyone in the audience from the health care and pharma side of things, we have a brand new Novartis case study that again is looking at how they have transformed their data solution. This is relevant obviously to those of you in the industry, but anyone outside as well, especially as we talk about the privacy and security considerations that come with these, platforms. I think it’s always good to look at the pharma industry with who have very stringent regulations. It’s it’s it’s very useful. So with that, I I want to thank our speakers again. This was a very interesting conversation. And to everyone else, we hope to see you at a future event. So thank you for joining us.