- 6 hours ago
For years, digital advantage was built around access: access to technology, content, audiences, and data. AI is rapidly changing that equation. Companies are being forced to rethink how they stay relevant, distinctive, and desirable in a market where AI increasingly powers everything behind the scenes. In a market where intelligent tools increasingly power everything behind the scenes, what still drives relevance and competitive advantage? Bringing together perspectives from technology, consulting, and consumer innovation, this conversation explores how major organizations are redefining growth and differentiation as AI transforms not just how companies operate, but how consumers experience products, brands, and value itself.
Category
🤖
TechTranscript
00:00So, please welcome Luc Daman, SVP and President of EMEA at Adobe, and Prouffo, CEO of SNCF Connect & Tech.
00:12And guiding this conversation is Olivier Girard, Director of EMEA and President of France at Accenture.
00:20Please welcome them all on stage.
00:43Good afternoon, good afternoon to all of you.
00:46I'm actually extremely pleased to be with all of you today.
00:50Stay with us.
00:52This is the best panel of the day because of the topic, because of my guests.
00:55I am Olivier Girard, the Chairman of Accenture in France.
00:59And the topic of the day is the only topic that matters in the end.
01:02It's value creation in the age of AI with a specific angle here with my two guests,
01:09which is because it's the customer angle, the end customer angle.
01:13Because we know, we all, you all know that AI is a revolution, new paradigm of productivity, R&D is
01:20very different.
01:21But what's new on the customer angle, on the customer experience size?
01:25How do we use AI in that specific domain?
01:29To explore this topic, I have next to me Luc.
01:33Luc Darmann is the President of Adobe EMEA.
01:36You all know Adobe, but it's the world's leading creating software company, empowering millions of professionals around the globe.
01:45And Luc leads strategy and performance across Europe.
01:47And I'm also extremely delighted to have with us Anne.
01:51Anne Prevot is the CEO of SNCF Connect and Tech.
01:54You all know this application very well.
01:57SNCF Voyager, digital entity, the one that designs and operates the digital platforms that you are probably using every day,
02:05like millions of French people every day.
02:08My first question is for you, Luc.
02:11Luc, so every board, every CEO is investing in AI currently.
02:18From your perspective, what are the organizations that are creating the most value out of it?
02:26Hi, everyone.
02:27Great to be here.
02:28I think the honest answer is that they are not necessarily doing more with AI,
02:34or they are not necessarily going faster for the sake of going fast.
02:37I think they are doing it more coherently.
02:40And many organizations are doing AI in pockets, proof of concept here, use case there.
02:47I think the leaders have realized that the compounded value of AI comes when it's connected.
02:54And at Adobe, we have the privilege to work with global brands across the world,
02:59and that pattern is consistent.
03:01If you take customer experience and you mention the topic, that's our field.
03:04We see that when, for instance, you connect an AI-powered content supply chain with the customer data
03:12and the engagement journeys, and when that moves together, that is when the magic happens.
03:18That is when you start to reach the promise of personalization at scale, for instance.
03:23So that's, I think, one factor.
03:25But I want to mention the two shifts that are important,
03:28which I think are an extra opportunity to create value,
03:32but will also accelerate the gap between leaders and laggards.
03:36And obviously, this is top of mind for everyone.
03:39It's agentic and human.
03:40The first one is outside the enterprise.
03:42Our consumers are now interacting with the LLMs.
03:46I'm sure when you shop now, there is a good chance that you will go to an LLM.
03:50So there is a real re-intermediation of the customer journey or the relationship between the consumer and the brand
03:56through LLM and agents.
03:59And when you think of customer experience, you have also to adjust to do that for both human and agents
04:06as well.
04:06So that is a big shift.
04:09That is a big challenge for a lot of organizations.
04:12When you talk about brand visibility, how do you stay relevant in that context?
04:17I think right now, the ones who are tackling this right away are going to have an opportunity to accelerate.
04:22And the second shift that I want to mention is within the enterprise,
04:26where we have entered this era where the enterprise is now both human and agentic, right?
04:31And agentic is bringing a move, which is AI moving from being an assistant to really a co-worker,
04:40which is going to work along the marketing team, the operations teams.
04:43So that triggers a whole new reality in terms of how do you shift your operating model.
04:49But companies today who are getting ready for this, rethinking how they can approach work,
04:56leveraging those technologies, but rethinking how they operate are definitely going to have an edge.
05:01So I would say connecting brand visibility, customer content creation, as well as brand visibility
05:10is really an accelerant to drive value from AI in a customer experience world.
05:16Thank you very much.
05:17Luke, Anne, next question is for you.
05:20How AI is changing the way you do your job?
05:23I'm not talking about your everyday job, but more in terms of your strategy and the customer experience.
05:31I would be a bad customer on your panel if I said nothing, but still.
05:38Let's start with what SNCF Connect and tech is.
05:42SNCF Connect, you mentioned it, it's an app, it's a website to buy train tickets,
05:48but we also have two strong arms, I would say, in tech providing solutions and also providing services
05:58around mobility and tech in the mobility sector.
06:04AI is important for us, but we were born 26 years ago as a digital company, a digital native company.
06:12So we've seen a lot of new technology, new habits, new usages from our customers that we need to adapt
06:21to.
06:22So we used our habits of adapting, adjusting to what was coming.
06:31And we were also using artificial intelligence even before we had this hype, because this hype is just focused on
06:42generative AI.
06:43And so if we come to generative AI, which I think is the core of your question,
06:53it's transforming the way we interact with our customers, but AI is not the purpose.
07:03The customer interaction is the purpose.
07:06And so the AI should be serving this interaction, serving this purpose.
07:12So what it creates for us, it's mainly opportunities for all of our activities, the train distribution and mobility distribution
07:22application and website,
07:24but also all of our tech activities, because it's transforming also the way we are serving our customers.
07:32Thank you, and I'm staying with you for my next question.
07:35You serve millions of passengers every year.
07:39What challenges were you trying to solve when you began to modernize your digital customer experience 26 years ago?
07:47I understand.
07:48So let's talk about some figures, some numbers.
07:52We served in 2025 approximately 25 million users, so people that used our app or our website.
08:0517, more than 17 million customers, what we call a customer is basically an email address and a credit card
08:15number,
08:16altogether purchasing something on our website or our app.
08:21Sold more than 233 million tickets.
08:25We had more than 1 billion and 6 visits last year.
08:32So we have a huge number.
08:34We have a lot of interaction.
08:37But still, we want to manage this interaction with all those customers like they were the only one we care
08:45about.
08:46We have a program around customer first, and this is exactly what we want to achieve with all the systems,
08:54all the tools that we are building for our teams, but also for our customers.
08:59So we tried to solve that, how to improve the interaction we have with our customers.
09:06I will give you an example.
09:09As many websites, as many e-commerce websites or e-commerce apps, we have a chatbot.
09:17As many other businesses, we have implemented a new chatbot that is powered by AI.
09:27We did that with only the customer interest in mind, not willing to just add some AI in our features.
09:38And so we implemented it very carefully.
09:43Domain by domain, we measured the impact that we made on the customer.
09:48How do we do that with our large numbers?
09:51We have a lot of people asking questions through our chatbots.
09:54If you have used it, you know that we are always asking the question, was the answer useful?
10:01And so we have a usefulness index that we are able to measure.
10:06We did basically an A-B testing, implementing the new AI-powered chatbot, leaving for the other half the old
10:17chatbot,
10:18and we doubled the usefulness index between the two solutions.
10:24Since then, we have continuously improved that chatbot, and we are continuously gaining some new points in this usefulness index.
10:36But what we implemented for our customers, we started it with our agents in our customer relationship centers,
10:46because we needed to challenge whether it would bring any value, and we needed to start with our teams as
10:54well.
10:55The answer was very useful.
10:58Thank you very much, Anne.
11:00My next question is for Luc, but Anne, feel free to add something.
11:05It's a question about the ecosystem and about the architecture, because we see in this domain every month, every week,
11:11there's something new popping up, new models, new agents, new platform are emerging constantly.
11:16And clients, companies, they don't want to be locked into one technology.
11:22They need the right model for the right use case, right level of performance, sovereignty control at some point in
11:28time as well.
11:29So, from your view, Luc, how should business leaders think about building an AI strategy that remains open and flexible,
11:38but that does not become fragmented at the same time?
11:41So, how do you deal with that?
11:42That's a good question for you as well, Anne.
11:44I'd love to hear your perspective as well, Anne, but I think it's a real struggle that we hear from
11:48all our customers today.
11:50I mean, the fragmentation is real, and it's getting worse, right?
11:52You mentioned it, new models, new agents everywhere, new platforms, in many cases, disconnected customer data with content,
12:02and teams may be working in silos.
12:04So, it's a very complex landscape and patchwork for our customers.
12:09But at the same time, I think no one wants to really standardize or just bet on one model, right?
12:15The tech evolution is so fast that it's not necessarily the right decision, and there is probably a fair vendor
12:22lock-in fear in this.
12:23So, I think clearly organizations today are calling for flexibility on interoperability.
12:31I think that's clearly the message that we hear, and that's the approach that we have developed at Adobe.
12:36We want to meet our customers wherever they are, and to that extent, we have signed and developed extensive partnership
12:44with all of the leaders, whether it is Anthropic, OpenAI, Gemini, Microsoft, NVIDIA, and all the leaders,
12:54so that they can access, our customers can access our workflows through any surface, and vice versa.
13:00So, I think this is the future right now, making sure that you simplify and you have an ecosystem that's
13:07going to be open.
13:08And it's the same with agents as well.
13:11Right now, the ability and the necessity to orchestrate multi-agent collaboration is going to be key.
13:18We strongly believe into this as well.
13:20We want to, you know, we are very proud of all the purpose-built agents that we have for customer
13:27experience, workflows, marketing.
13:29But in the end, we are very conscious that there is a broad ecosystem.
13:32So, when we approach this, we are developing solutions like Coworker is a good example of that,
13:38where you have the ability to leverage the Adobe agents as well as all the agents that will be required
13:44to perform a task with a business outcome that is clear.
13:48So, I think that's the future.
13:50It's really flexibility at the model layer, but consistency at the experience layer that's going to make a difference.
13:57Thank you for that.
13:58And do you want to complete that?
14:00Just to create a parallel with the story I was telling about the chatbot, we are still in the infancy
14:09of all those technologies.
14:11So, we need to experiment.
14:14So, what we are doing a lot, and we have a kind of a motto for that, which is dare
14:20to experiment, deploy responsibly.
14:24So, we experiment a lot.
14:27We share a lot of feedback on what could be useful, what is not, et cetera, et cetera.
14:33And we deploy when we gain some certainty.
14:37And we are also able to move backwards and go back to another model.
14:46But you're right.
14:47There are a lot of models available on the market.
14:52They are evolving quite fast.
14:55And depending, and I don't believe, this is a personal statement, I don't believe in a one model solving all.
15:03So, we need to keep our options open.
15:09And this is something very strong in the way we need to consider everything that is related to AI.
15:15So, we all agree.
15:17At Accenture, we said that 50% of the agents of the future will come through the software world.
15:2150% will come from the client world that you will need to build on your own because you have
15:26specificity, industry specificity.
15:28So, thank you for that.
15:30My next question, Anne, is for you.
15:32Technology is one part of the equation.
15:35You even say AI is not the purpose.
15:37Can you share examples of how SNCF Connect is using AI and data to create real value for end consumers?
15:47So, I already gave the example of the chatbot, but we have many other experiments and implementations.
15:56Some with Adobe, by the way, and we are using now AI in a lot of our jobs and activities.
16:12Of course, customer interaction is one.
16:15We have all that is related to marketing.
16:18Why is that important in marketing?
16:20And this is something that we are using Adobe for, especially.
16:26It's because you are targeting now almost a segment of one.
16:31This was a very old topic, but now it's a reality.
16:35And now we have the tools to create content that is really adapted to those small segments.
16:42And those small segments, you will touch them with various touch points that you need to adapt your content for.
16:52Once again, AI is a really valuable tool to do that.
16:57Sometimes it's just adapting to your own standards.
17:01I will give you a very concrete example.
17:04Very recently, we had a French city that gave us some marketing content about bike rental in the city.
17:16The content they gave us was a young woman, very happy to be on her bike, but she had her
17:25feet not on the pedals.
17:27And we have a value in SNCF and in SNCF Connect and Tech, which is security.
17:34It's not very secure to not have your feet on the pedals when you are riding a bike.
17:40So we adapted, we changed the content, the marketing content they gave us to fix that.
17:48We would have been able to do that before, but it would have taken, I don't know, one day, two
17:54days with some content creativity tools.
17:59It was two seconds with AI.
18:03So this is the type of activities that we are using it for, and also everything that relates to development.
18:13All the development cycle is now powered in some way by AI tools.
18:18Very impressive answer with your example.
18:22Thank you, Anne.
18:23And so my next question is for both of you.
18:26Accenture's experience is that the hardest part, and you say that already, but Accenture's experience is that the hardest part
18:31of any AI transformation is not so much technology itself.
18:36It's the reinvention around that.
18:37It's leadership, alignment, it's operating model.
18:40It's the skills, it's the adoption, it's the governance, it's to do it in a responsible way.
18:45And it's all this.
18:46It's what I use the expression, what's above AI.
18:50So once the foundations are in place, you know, that's the famous scale question.
18:54How do you move from pilot, you mentioned dare to experiment, to scale, to all your customers, and all your
19:02customers as well?
19:02So maybe, Luke, you start on this one.
19:05I think you pointed it.
19:07One of the misconceptions of AI is that it's a take-all-out, while it should be a business transformation
19:14and a cultural transformation as well.
19:18It's, I think, right now, where we see the ability to scale, it's when first you need to choose a
19:24workflow that will drive tangible outcomes, and it's true that customer experience and digital marketing workflows are enabling that.
19:31But there is a real appetite to drive a business outcome that's going to be a game-changer, a completely
19:38advantage.
19:39You can use AI for efficiency, and sometimes the tip of that is bare, it's just efficiency, doing the same
19:44thing faster.
19:45Fine, that's great.
19:46But I think we see customers right now making a difference by rethinking, as we mentioned before, how they're going
19:52to operate in their own space.
19:55Tesco in the UK is a good example where we are co-innovating together to do something really disruptive.
20:01So I think there are three elements that stand out.
20:04That means, in that case, you need to have really a mandate from the top at CEO level and XCOM
20:08level that this is going to be a business transformation.
20:12We have clarity and a vision that's going to be driven by tangible business outcomes, and it's not just efficiency,
20:19it's gross, and it starts there.
20:22The second point, which is a bit less glamorous, is establishing the right foundation, right?
20:27Because AI, without the right foundation, it just doesn't mean anything.
20:31And that foundation is made of a trusted data layer that is 360 enables you to power meaningful journeys in
20:41the instant that are going to be relevant.
20:44It means also a strong governance that is in place.
20:47And for a lot of marketers in the room, I know it's top of mind, it's also brand control.
20:52When you start deploying AI at scale, you need to ensure you're going to be able to maintain the brand
20:57gallery, the brand consistency, and the authenticity of your brand.
21:01So if you add to that a genuine appetite to amplify your DNA as a brand in the first place
21:08and use AI to amplify that ingenuity, free up some time for people, whether they are marketers, creators, to think
21:16about their core business, we see that as a distinctive approach that pays dividends.
21:21And the third one, and I think Anne and SelfConnect are a great example of that, it's change management as
21:28well.
21:28It is not just a training program, it is a matter of sharing that vision, making sure all the layers
21:34in the organization are totally infused with what we are doing.
21:38And it's rethinking the operating model that will take you to those type of results.
21:43So that's something I think you have done exceptionally well.
21:49I will not paraphrase what you've just said.
21:54I will go on the side of what you said.
21:58Artificial intelligence amplifies.
22:01It amplifies your strengths, but it also amplifies your weaknesses.
22:07So you need to be very strong and to know what makes you so strong on the market, what makes
22:15you so valuable for your customers.
22:18And this is what you need to focus on, and this is where you need to put the effort in
22:25terms of artificial intelligence.
22:27And I will go back to my chatbot example to demonstrate that in some way, because it's a bit simple.
22:38When we started the first deployment of our new AI-powered chatbot, we discovered that some answers were not relevant.
22:49Why that?
22:51Because the chatbot was looking and the AI was looking in our FAQ, and it was looking at the details
23:01and the content.
23:02The AI was not able to understand what was written in those FAQ.
23:07Guess what?
23:08Human people in this room were not able to understand that as well.
23:11So we were not strong in our content, and so AI was amplifying our weaknesses.
23:20This is a very simple story to illustrate that.
23:23But when you are improving the foundation for AI, you are also improving the foundation for all your teams, for
23:32all your customers.
23:34That's a great story.
23:35Thank you for that.
23:36I will come back to that in my conclusions.
23:39We have five minutes left, so I have one last question for both of you.
23:43Take the time to think about what you want to share with the audience, and then I will have a
23:47few words to conclude.
23:49Can you share with the audience?
23:51There's probably a lot of CMOs in the room.
23:55So what are your two lessons learned on your journey, on what you see at clients and your personal journey
24:02at SNCF Connect and TED,
24:03on this whole journey from piloting, daring to experiment, to deliver tangible value at scale for your millions of customers
24:12and your hundreds of customers?
24:17I think the – I wouldn't think if you're a CMO, like, don't do an AI project for an AI
24:23capability.
24:24Start from the business outcome.
24:26Look at probably your customer journeys.
24:29What are the friction points that you want to solve?
24:31What customer experience do you want to drive?
24:34Start with that and go, you know, backwards from that to how you are going to get there.
24:41And so I think that's the first point.
24:43And then the second one is, you know, there is always a struggle between finding quick wins, tangible, fast outcomes,
24:51as well as being changing the game totally.
24:55I think when we look at CMO, like content production, content velocity is probably one of the low-hanging fruit.
25:01You can start there.
25:02But I think it's important to expand the scope and the vision for an end-to-end vision from a
25:08marketing standpoint.
25:09If you deploy a full content supply chain, for instance, and you connect it, as mentioned, to the customer journeys,
25:16then you are going to start to accelerate the results.
25:19And I think it's a huge opportunity for marketers and CMOs to move from being maybe a cost center to
25:26that to a profit center really driving growth and owning that customer experience really tight to the CEO.
25:35Thank you for that.
25:36Anne, two lessons learned.
25:38So once again, not paraphrasing what Luke just said.
25:44Do not confuse the tool for the purpose.
25:49Purpose is important.
25:50In our case, once again, it's customer, customer interaction, customer first.
25:55And AI is just a tool in our toolbox.
25:59And we have many tools.
26:04And second, measure, measure, measure.
26:09Because what you are not able to measure, you are not able to improve.
26:13And as AI is here to improve, is here to amplify what you are doing, you need to measure what
26:21you are doing.
26:22Thank you very much.
26:23So it's the time to conclude.
26:25I was able to keep a few things from you.
26:28Great insights, I have to say.
26:30Well, I could say I could put the Accenture signature below that.
26:35The first one is what you said, Luke, about there will be more gap between leaders and laggers with this
26:40revolution.
26:41Because if we are honest with ourselves, in the past 25 years, I have more than 35 years of services,
26:47you would not die for missing or for being late on the technology revolution.
26:51This is very different today.
26:53So you are absolutely right.
26:54There will be much more distance and gap between the leaders and the laggers.
26:58That's the first point.
26:59The second point, and you mentioned it many times, AI is not the purpose.
27:02And we tend to confuse that because there is such a hype on that.
27:07So let's not forget that we need a real purpose, a business purpose.
27:10That's the second point.
27:12The third point, you use the word, Luke, interoperability.
27:15In the world, that will be hybrid because many agents, many cloud, very different service level, very different sovereignty service
27:23level as well.
27:24To be sovereign, you need to be interoperability.
27:27So that's another question, another point, I think, very insightful.
27:32The next one is, Anne, dare to experiment.
27:37We can do POC.
27:38That's the beauty of this technology.
27:40We can do a lot of POCs and learn from that, from a cultural standpoint, but also from a business
27:44standpoint.
27:45So dare to experiment is a great one.
27:47I also love, Anne, what you said about this is amplifying who we are.
27:53This is amplifying strengths and weaknesses.
27:55And we tend to think about only strengths, not about weaknesses.
27:59So that's a very good insight from you.
28:02And I will finish with the last one.
28:03That is one from Accenture.
28:05Human needs to stay in the lead.
28:07So that's absolutely critical for us.
28:09Thank you very much for your presence here today.
28:13Thanks a lot.
28:13Thank you, Anne.
28:14Thank you, Luke.
28:14Thanks a lot.
Comments