[Music] [Bruno Soeiro] Good afternoon, everybody. Welcome to today's session. Just before I start, I just wanted to let you know we've had a bit of a content velocity issue. Our presentations didn't make it into the Adobe system for today. So colleagues just gone to fetch laptop just to show some visualizations, but I want to be respectful of your time that you've given us. So we'll be making a quick start. So in context to today's session, I just wanted to cover off a quick prelude on how we all know and listening in today's sessions and before arriving to the Summit, how AI is essentially reshaping marketing and the Customer Experience Management process. And it's essentially offering us new opportunities to scale our marketing initiatives and drive significant innovation. Where we're seeing the transformative effects is lying between the integration of creativity, workflow processes for content, tech, and data, and which is driving brands into a new area in enhancing the customer centricity, but predominantly on the operating model. And how the operating model is starting to integrate ethical AI practices across to enhance productivity, drive deeper engagements, but more importantly is how do we unlock efficiencies within the ecosystem to drive investment, to drive growth. And it's in this context that I'd like to drive the conversation today. So introducing to me on my left on your right, I'd like to introduce Claudia. Claudia is Director of the Innovation Studio at AT&T. Tom Portman, joins us as a Group Vice President and Head of Digital Channels for ABB. And we've got Bindya Raj, who is Global Head of Delivery for Digital Experience, everything from creativity right through to marketing technology stacks with Adobe, to marketing operations for Infosys. So just to start in this context, just to start with Tom, I wanted to ask you in your area, how are you prioritizing and tailoring your products, your services, and the journeys to your different audiences and personas? [Tom Portman] So before-- Wow, that's loud.
Before we even do that, the three of us are going to move over there because I'm conscious that the people on this side of the room can't see us. So a sound guy, if we getting feedback, it's all your fault. But, yes, good afternoon, everybody.
Remind me of the question, Bruno, and then we're going to keep on looking over to you. It's going to be like talking to the head teacher. Yeah. Yeah. Absolutely. So it's essentially when you're thinking of AI, GenAI scalability, innovation, how are you-- Are you okay? How are you prioritizing the contents and the journeys for your different personas and segments? So let's just put this into context for ABB. We're made up of 19 different divisions. We've got presence in 98 different countries around the world, and that gives us a huge, 19 different go-to markets with that across ABB. So already, we're starting to look at how we differentiate for each of our divisions because we've got lead times for sales that happen in a matter of seconds, matter of minutes for some of our fast moving products through to sales lead times that take years for some of the things that we sell. So we need to be looking at that differently for each of our divisions. We need to be looking at that very specifically for our customers. And the way that we're working through that within ABB is looking at that division by division. What is their specific go-to market and how do you deal with that? Now from a personalization side of things, we're very much at the beginning of that journey. And looking at, how do we deal with those different go-to markets? How can we inject AI into that to speed that up? And we're very much going through personalization proof of concept at the moment. I think the big thing around that for us is we need to learn quickly. I don't believe in the term fail faster, would rather learn quickly with that, what's working, what isn't working, and then move on from that perspective. And with 19 divisions, with that global spread, we need to be thinking about, what is it from our customers specifically in those territories and getting close to them. But then also recognizing that one size doesn't fit all. And I think, what was it, James Quincey's conversation this morning in the Keynote I thought was very out that one person's experience of Coca-Cola in one country is very different to in another country, and that's the same in ABB. Maybe not as much with having it over dinner, but what they need from us as a business will be very different in China compared to Europe compared to in the States. So very much on the beginning of that journey digitally, but looking to learn quickly in that space. Fantastic. Thanks, Tom. And Claudia, from your perspective, if you could just describe to the audience your role at AT&T, and how are you addressing and prioritizing your products, your services, your accessories in the sea of sameness that you're competing against? [Claudia Gill] Yeah. So for AT&T, when we talk about content, it's just so important to understand what are the needs of the consumer and to also understand what is their intent, so we can match that intent. We have millions of visitors that come to the website on a daily basis, and we know that 85% of those visitors end up in a transaction, not necessarily in digital, but 85% touches digital. So it is extremely important to understand what is that content that we're putting in front of our customers on every single channel, and how that content needs to be cohesive from all of the different touchpoints, and understanding if I start in retail, how can I finish in digital or vice versa. So we start seeing that omnichannel need for a seamless customer journey. Fantastic. And, Bindya, from your perspective, how are you finding-- Or what strategies are you finding most useful? You know, we've heard from Tom, very different industry to AT&T, but what are the common strategies that you would like to share with the audience that's delivering content velocity outcomes? [Bindya Raj] Yeah. So we try to focus on three things primarily, on how we can bring value to the customer, how we can engage, and how we can really understand the need of the customer. When I say value, it's about finding the differentiation that we can bring into the customer and we do that by multiple things. One is, of course, we do it by bringing in the focus that we need in and then as second, it's all about bringing the transformation that they need. So for that, we take a consultative approach about being a consultant, being a personal adviser to the customer, bringing the right stories that we can show on, how we have effectively done it as well as the transformation, extra examples that we can bring in with the latest technologies. And it's all definitely about bringing in creativity and technology together which drives the whole differentiation. Fantastic. So I think just in summary, we listening in on what are some of the foundational approaches and strategies to deliver content journeys to personas. But what I think would be a good topic to address, once you've got the foundations of your data, specifically your data in tech, what opportunities are you envisaging in your industry, Tom, for GenAI or the use of AI specifically? So, yeah. I think it's a good point that the way that we're rolling out from an online transformation perspective across ABB is very much getting those foundations in place, very much getting those more content. And that's made up of data foundations in place consistently across our business. And then our next step is then how can we accelerate that. So over the last, what is it, two or three years, we've been building the foundational capability so that our divisions can start migrating over to that. It's the year of migration for us this year. Whereas the next release that we're going to be bringing through is where can we inject AI into that experience so that we can be using GenAI for all content creation and speeding that up. Again, I thought the presentation, the keynote this morning was very interesting about how we can increase the velocity of that supply chain and bring that into it. Then from an analytics perspective we know already, and this has been around for quite some time, from analytics side of things, AI is really good at finding consistencies, finding trends that we as humans just wouldn't pick up on. And by putting that global ecosystem in place, we can find global trends that maybe somebody in China wouldn't necessarily notice for somebody in US or we wouldn't be looking at that globally. So it's all about putting those foundations in first and then leveraging AI to increase the velocity in across ABB. Fantastic. Thanks, Tom.
Bindya, from your perspective, as you have visibility across the market, across industries, what innovative use cases are you seeing with Generative AI that you could share with audience today? Yeah. So, I think, now as we all, it's more about not working in silos, and it's all about the connected experiences that we can provide to our customers. So we see a lot of GenAI use cases which are coming in. First, starting from the engineering side of it, where we are dramatically able to increase our productivity...
In the entire development cycle, we are able to bring in our accelerators and tools which helps us speed things, which will definitely speed up things and reduce the cost. That's one part of it. And, of course, in terms of, whether in terms of the chatbots which are doing or whether bringing an agentic AI in marketing. There are a lot of multiple use cases which we are plugged in and it's not just in one sector, whether it is in manufacturing, in healthcare, or in financial services. A wide range of use cases we have been able to see right now and some of this which are already being implemented today. Yeah. Fantastic. Thanks, Bindya. And, Claudia, we're having a discussion on this yesterday evening. Could you share some of the methods, tools, and some of the approaches that you're adopting at AT&T? Yeah. When we talk about GenAI and content creation, you can look at this at different ways. Right? It's not just about your Firefly and creating images, but it's also about what are those smaller or features that can help you with productivity. At AT&T, we started using a couple of features that are embedded within Photoshop that are actually Firefly tools to start working on the backgrounds or adding and removing elements from a photograph. And just by using those simple features, we have achieved 15% over year increase in output without extending our headcount, which is really impressive. And this is something that was already there. Right? So also when we talk about AI and content, it's not only about the creation, but also think about how it can enhance your workflow efficiencies and how you start becoming more productive with looking-- You saw today in the morning like Workfront. Right? How we have at AT&T been able to use Workfront, and what are all those capabilities that they keep coming with, and how you integrate it with your different other tools, and you start creating and unlocking opportunities of efficiency. It's not just about creating content, but it's how everything comes together with your workflow. Fantastic. And, Claudia, just to continue on that, how are you linking the creative process into the personalization realm of the content delivery and then aligning that or assigning that to your different segments and personas? Great question. Because as we start seeing and you heard it this morning, it's all about personalization. Right? So the demands of content are going to continue to increase and I don't know how is it in your industries, but I don't know if the headcount will also increase. Right? So how do we become more productive? And personalization, right? It's going to continue to grow. With AT&T, we started our personalization journey a few years ago, and so far we have seen 38% in increasing conversion by using personalized content. Now where the opportunity lies is that we keep thinking about personalization on the living and looking space, but we have not unlocked yet the opportunities with the personalization on the buying state. So also in the morning in the Keynote, the SVP of-- Lily was talking about the opportunities of unlocking that personalization and across the entire funnel. So that is a journey that we are on at AT&T and looking for as to how we're going to really unlock those possibilities. Fantastic. And I think the other thing that we talked about previously is the importance of experimenting with this, not just saying the days of how well we're going to invest millions and millions on something because we think that the business case is going to play out. That's not how we're doing it at ABB. And certainly from what you said at AT&T, that isn't the way that you're going either. It's try things, the term that I don't particularly like, which is fail fast, but learn fast, actually. What works well and then iterate from that. And, again, that was called out in the Keynote this morning, which really resonates with me. But you don't need to have a huge framework to be able to experiment with this stuff as long as you've got the guidelines across the business and at ABB when it comes to AI, I think we were very quick actually looking at how we can make sure we can leverage this, but leveraging it in a safe way, in a compliant way, in a legal way so that then our business feel comfortable with trying things out with AI, but making sure that we're keeping ourselves safe, and we can talk about that more further. But try these things out. If it works, great. You can do it for a relatively low cost. If it doesn't work, your loss is quite low as well. - Tell me-- Sorry. - Go on. Sorry. You bring an excellent point about A/B testing.
I have an example of how the things that we do digitally on att.com. Right? We let data drive our decisions. And the home page is a very impactful real estate, and there's a lot of priorities that we need to prioritize within that one page. And everybody in the business has their priorities and there's five different things that we need to position above the fold, which is really hard to do. So we turn into A/B and run tests and we merge with technology and AI, and we use machine learning. So it's like, does machine learning perform better than human dictated content? Me saying these audience should see creative A versus creative B, or should I let the algorithm dictate what the customer should see? And by doing an A/B test, we have data prove us like you should let the machine learning drive the content. And since we've done that, our performance for click-through rate and conversion from our homepage have gone up, and we don't have to have multiple discussions a week of what are the priorities. That's a fantastic example. And while we are in personalization, we discussed A/B testing. Tom, it'd be interesting for you to share how you're tackling personalization through search as a different means to deliver more predictability on outcomes? So talking about just wanting to go back to the A/B testing, we're again very much in our infancy with that at ABB. But the way that we're approaching it, and there's people in this room from ABB that have been involved in this, is basically marrying up two sides of the customer setup. And one is we actually go out and talk to our customers. We survey our customers. We do across our divisions and talk to them, find out what is it that you're liking about engaging with us, what is it that you're liking, whether it's from a digital side of things, what isn't going so well, and almost a bit of friendly fire with that, but then also matching that up with analytics. And when I said about this being in our infancy, we're only just starting to do that. The reason why we're doing that is because a customer may say, "I want it to work like this, and I want to work with you in this way," but the analytics may say something completely different. So we need to have that conversation, but to understand both sides and be able to start looking at next steps with that because they may be saying that, but their behavior is showing something different. So what do we need to be changing? Do we need to be having a different conversation with our customers? Do we need to be presenting it to them in a different way through our digital channels? Or is it actually we just need to go with the analytics? But you all should be marrying those two up. From a search side of things, again, this is something that we've taken quite big advances in over the last, what is it, six months since I brought in some dedicated search resource for ABB, 89% of our customers on our site search for products, so they're looking for product information. Now that's saying both positives and negatives that they're looking to get straight to the products, finding the product information that they need, find the support that they need. But also a large proportion of our business is through industries. Depending on who you ask, we've got 23 to 26 industries that we create content in and work through as a business. So why aren't we driving more content through our industry streams and driving that personalization through there? And that isn't reflected through search at the moment, but that's where we're going with it. Fantastic. And, Tom, as part of your search roadmap, we've had a discussion on the conversational AI element that we're looking to add on to your roadmap. Do you want to describe to the audience what advances you're looking to make there and the matching of search outcomes? It's quite cool, actually. We haven't launched it yet, so I don't really want to talk about it too much.
We're using AI to drive a more conversational search with our customers. But the things that we've been experimenting with this is, as we know, GenAI certainly in the search space can hallucinate. It can give duff results to customers. So it's how do we balance that up with traditional search results, making sure that they're prioritizing the right way for our customers, but also through a conversational side of things as well. The other benefit that we're seeing from that already is the language barrier. Again, that you saw some of that this morning, but not necessarily in search, is being able to provide search results in the customer's native language, which may have come from a different language to what they're searching in. So search in Italian, and the results are actually in English, but they get the results in their native language, which is pretty cool. - Yeah. It's very-- - More to come. More to come. More to come on that. And, Bindya, from your lens or aperture at the market, we've discussed A/B testing for personalization. We've discussed content, aligning it to different journeys, industries, and then we've extended on search. What other areas in the realms of personalization, can you share with the audience that are picking up a lot of traction but also driving meaningful outcomes? Yeah. I think, definitely with the Generative AI coming in, retrieval augmented search is something which we are seeing a lot of traction. And as Tom mentioned earlier, there is also the issue of hallucination and ensuring that the data is intact and the right search output is provided. But there is a lot of scope there, and it is definitely moving to the next level, with the features that RAG is bringing in. So we are seeing a host of opportunities where it's all about bringing, as I mentioned earlier, hyper-personalization at scale, and the customer demands it, wants it, and definitely asks for all the data which is available to be in a 360 degree format for them. And that's exactly what the new search techniques and RAG is providing today. So that's some of the-- That's one main thing which we're definitely seeing in the search area. Fantastic. Keeping in with the theme of personalization, I promise I'll go beyond this in a moment. So, Claudia, one aspect, we were discussing, it's we start testing, doing the A/B testing, we start learning. But where does that start for you? Does it start at when somebody does a search in Google, they land on your page, where do you actually start tracking and identifying your leads to start driving the personalization? So it should be in first customer touchpoint. Right? It's, again, going back to that seamless omnichannel customer journey is understanding what cohort is interacting with your content and then driving that same cohort across the different channels. So we're looking at AT&T, the usage of CDP to help us with connecting all of those different channels.
Again, it's a journey for us. There's so much more for us to uncover and perfect. I feel like even though we've been in the personalization journey for a few years now, it still feels relatively young. And all of the different tools and technology that keep coming into the market are really going to help and support that. Right? And also, the customers keep evolving as to where do they want to shop, what is the kind of information that they want to receive. Not every single customer wants the same amount of information. So how can you personalize that? It's not just about the messaging, but it's about the amount of messaging, also the journeys. Does this customer want to go through a flow that has 20 steps or this customer is savvy enough that they just want a bundle customized already for them? So in that personalization aspect, there's different, different areas that can be uncovered. So as you're describing some of the challenges while you're seeing personalization at an early stage, could you describe some of the challenges that you've picked up along your journey and how you tackled them and turned them into opportunities for yourself? I think the challenges become with being such a big company with so many different products, and the achieving omnichannel can be a hard part, right? Big company. But how do we uncover these opportunities and how you become that person that is able to champion change and be able to speak up and say this is what we usually should be doing for the customer. So it may take time, but sometimes repetition is what will get you to success.
So for us, it's understanding what the problem is, but also at the same time trying to start with a small use case. Right? So if I can show you in the domain that I control, which is att.com. If I'm able to come with a use case that shows you that by personalizing the content on the home page and carrying that message all the way through the buy flow, if I can increase click-through rate or progression, that is a win. And now I can take that to my peers on other channels and say, "Hey, I did this. What about if you help me and let's do this cohesive message from your paid campaign into the website, and let's see how that goes?" Right? So we pick a use case. We see that that conversion, the click-through rate and the conversion are lifted. Now I can start evangelizing this with data that will support that this is what we should be doing. Yeah. And that really resonates with me and my experience at ABB is that, again, starting small, showing an example, and something that one of my colleagues at ABB says so many times, seeing is believing with this stuff. So, yeah, we can go into board meetings. We can go into exec committees. We can go into steering committees and so on and talk and evangelize about this stuff. But until they start actually seeing the output, seeing how this would work, then they don't start, then they can start buying into this. And I think, again, going with divisions, ABB, very different divisions, very different go-to markets as I mentioned earlier, trying out with one of those, seeing what works, and then that could potentially work for other divisions. The other thing that, I've learned in my time, not just at ABB, but in general, one size doesn't fit all with this stuff. So just trying to go blanket while we're all going to do this approach to personalization, it doesn't necessarily work and it's trying to find the commonalities with divisions, and then they can distinctify themselves in those areas if that's even a word. Thank you, Tom. Thanks, Claudia. And, Bindya, if you could share with the audience today, when everyone's faced challenge-- I think, it's fair to say, since beginning of humanity, we've been transforming, and we're just at a different pace of it today. Where do you see the biggest challenges that you're converting to opportunities? Yeah. So, I think whenever we convert from a legacy system to the new edge MarTech system, sometimes there is the issue of having data going into silos, and then we convert them into an opportunity by giving the customer a 360 degree view of the data and then ensuring that this is utilized to get a hyper-personalization, which is really making advantage of the challenge that we have in converting into an opportunity, which we definitely see coming in many circumstances. And it's also about, for any problem that we see now, it's all about going with the mindset that you have to go with the compostable architecture and have an AI for strategy, which definitely helps us, transform any challenge into opportunity. Fantastic. And what are the efficiency plays, Bindya, that you seeing when you're trying to convert the topics you've discussed into challenges to opportunities, what are the efficiency levers that you see, that customers are adopting, and how they're unlocking the value through that efficiency, and what are the areas you see the investments heading towards? Yeah. So, some of the-- Yeah, definitely, there's a lot of analysis which goes into it in terms of when we try to do this with the customer, we do analysis around the same. We do a lot of analytics around it and also around sentiment analysis, and a lot of that we do so that we give the right amount of data. And also the customers definitely now even the click-through rates or even the time on the page, and a lot of other factors are there, which are being tracked at the minutest level now so that there is no loss in the data. As you said, it's all about now the journey, and it's about orchestrating the journey for the customer end to end. And lot of these metrics are definitely tracked and followed through so that there is no loss of data while waiting things. Fantastic. So I just wanted to lead on to another question. So we've talked about addressing strategy, addressing foundational elements. What about adoption? Claudia, from your perspective at AT&T.com is a massive estate. You've got a lot of different product owners. You got a lot of different entities. How do you help adapt to the tech and data advancements you're making in communication, content, and personalization? It's not easy task. As I've said before, you need to really become an agent of change and be willing to leap that for the rest of the organization, especially in times like now that we keep seeing the advancements of AI. How do you turn a company that is set in certain ways and the processes do not exist to bring that new technology? So if you're passionate about it, which I hope you are because this is the future, is how you build that path and find a way to convince your manager, then your manager to convince their manager and start building these relationships and start creating a process to bring everybody along into this new journey, where it really in an era of revolution and it's not going to stop. Absolutely. And also another adoption tactic is internal talent. How are people geared to adopt this technology and take it forward and essentially reassign tasks of their daily jobs to the new tech that you bring in? Tom, from your perspective? In a decent-- Voice went. In a decentralized business like ABB...
In my view and the way that I'm moving forward with is all about the pull from the business. So what I talk about a lot is if you're trying to force something onto individuals or force something onto a department, then you're automatically going to get pushback for that. Anybody that's got kids will know about if you start trying to force them to eat the greens at dinner, it's going to end up on the wall at some point. At least it did for me. But it's about how do you get them enthusiastic in it. And the way that we're approaching our transformation is very, very much about what are the challenges, what are the opportunities that you're wanting to overcome or achieve across your division, and then how do we build that capability to enable you to do that. Now, of course, we do guide them with some of that to make sure that the releases that we're delivering the capability that we're delivering is going to enable them to optimize. But fundamentally, it's about the pull from the business. You get that in place, then you're pushing against an open door. They're wanting to take this capability from you. So always keep that in mind. And, Bindya, I'd love to hear your perspectives on when you we're delivering more advancements in technology, which is driving a lot of efficiency to unlock value. What are you seeing the opportunities for organizations in talent and re-skilling to really benefit from the promise of AI or GenAI into their organizations? Yeah. So, as an organization, we are really focused on bringing our AI first culture in our workplace and workforce, which means that we use AI to transform the way we do our day to day activities in work, whether it's on the engineering side or on the day to day applications that the team uses, as well as how you deliver to the clients. So it's really, it's not whether you want to adopt AI today in our organization, Infosys. So it's more about how well you can do it. So it's all about changing the culture of the organization to adapt to it in the complete way that we are doing it, and we have done that in a significant way. Second is about getting the team trained. So we have ensured that, we have completely trained and made our folks AI aware and then moving people to different levels of certification, so that technology, it is ingrained into them and built different accelerators, which can enhance their productivity and the realization for the people that it is not to replace them, but to actually bring them to a level which will make them in power with the person with a better experience and simplify their lives. So that has gone into the people, and I think it's all about the culture that you bring in. So that is the most important. And third is, of course, the-- I believe we need to have a responsible AI guideline and framework, and that should be in place to ensure that everything is within the checks and balances that needs to be there. So, Tom, I wanted to refocus on, you were speaking about pull from the business. Now we see that when we're driving long-term transformations that are being delivered over multiple releases. As the business starts to learn, adopt to tech advancements, there's something called AI out there. How do you balance releasing the baseline capability while adopting the demand from the business? And I remember you discussing this as it's about distinct and about being core. Yeah. Could you get into that a little bit for us? I think so. So there's two in the transformation that we're going through, there's fundamentally two streams. There's the build of the capability, which is all driven by the business, prioritized by the business so that then they're getting regular drops of capability on a monthly basis so that there's new stuff coming through, and the priorities of those, as I say, are very much set by the business now. Then the second part is around the adoption, making sure that they've got the right skill sets, got the right resources to be able to adopt that capability. And there's reason why I call it adoption, not migration, because migration is just moving from one place to another. Adoption is about changing the way that they're actually going to be working, changing the tools that they're going to be using, and changing the ways that they're going to be interacting with their customers and really putting the customer at the forefront of everything that they're doing. Now what we're finding and what we predicted as a result of the adoption pieces, they're starting to learn, "There's this new stuff that I can use. There's this new capability, but now I want more." Remember what I said about the pull. I'm wanting more of this. These are the types of tools that we're wanting that are distinct to our business, and what we're spinning up as part of the transformation is now what I foresee going into more business as usual sort of a factory model where we're the third stream of iteratively continuously improving. And the capacity of that, again, because they're the ones that are paying the bill, is down to the divisions. So if they want this capability to be rolled out more rapidly, then we increase the capacity, increase the cost that goes with that, unfortunately. Well, maybe you could help me with that. Yeah. Increase the cost of that so then we can roll it out quicker. If we're having tougher years or if they're wanting to slow down, they're not able to take that change on, we can ramp that capability down. So program through releases, division adoption within the divisions, and then moving into the third stream, which is continuous improvement. Thanks, Tom. And, Bindya, be good if you could share with the audience some insights on some of our large customers who are on a multiyear transformation. They see opportunity to drive disruption to their own transformations, through the use of AI and GenAI.
How would you safely integrate these capabilities mid transformation? How do you-- Sorry. Repeat the last part. So how do you balance integrating new capabilities halfway through transformation? Yeah. So I think there are a couple of customers. Of course, we have long running transformation journeys and, of course, ABB is one and definitely we break it into parts and we ensure that we have a continuous-- We do something known as a real model brainstorming with them, where we ensure that a whole set of ideas are completely worked through. And we also ensure that the existing data, the existing functions which are there are totally utilized and that kind's a baseline for us to integrate into the newer capabilities and the new offerings. So it's a continuously evolving process, and whenever new offerings are defined, we keep in mind the existing opportunities which are there and bind it and tie it together into it so that we can take it forward. So that's the process. And, of course, studio model is another big thing that we are working on, and that'll definitely ensure that the whole process moves into the future augmented way. - Yeah. - Fantastic. Just doing a quick time check. We've got 20 minutes, and we all plan for roughly 15 minutes of Q&A, with the audience just to make it more interactive, address some questions you may have, we may not have addressed, or would like to address. So just like to-- We got the microphone in the middle. Just wanted to see if there's any hands up. Any questions to our panelists today? Go ahead. [Man] What do you think is the main difference between B2B and B2C, better from each one regarding how they are perceived. Right. Stop. Yeah. Okay. So in my opinion, B2C and B2B...
There's not that much difference. There's just more complexity in the B2B side of things. Let's be honest. I think over the last few years, few decades...
The B2B side of things has become more focused on customer, and I think that's absolutely the right way to go. But what we're learning from that is, yes, we're more focused on customer, but there's more barriers that we need to be getting over as a business in the B2B space because the credit checks, there's more than one person from that business that are buying from us. So how do we deal with that? There's the whole documentation side of things, more complex products, and so on. So I think it's the level of complexity, but still fundamentally, we've got business buyers that are comparing their experience to us as ABB to-- Well, it's so easy for me to do this through Amazon. Why is this so complicated to do this through you as a company? Not ABB because we're awesome. But there is that comparison. If I can do this easily in my personal life, then I expect the same thing from B2B side of things. So I think that the fundamentally, in a word, is the complexity, Maurice. That's what it's down to. For AT&T, I think is the...
Between consumer and business is the amount of people that you need to serve. Right? So the product and the quality should be the same, but because you're not serving a family of four, you're serving a business of 20, 50 plus, what is that offering that is still competitive? From a experience perspective, there should not be difference as an individual, how are you purchasing this product digitally or in a retail space and business. You should have the seamless experience as well. But from a product offering, it's very different. Yeah. And another thing just that I've thought about from a B2B side of things, the trends in B2B buying are interestingly enough from a procurement perspective, the procurement group, not the procurement team, the procurement group is growing.
I would have thought, until I saw these stats, that the procurement groups would be getting smaller because you'd be able to move quicker. Actually, they're getting bigger, and that's because of the generational changes where we've got buyers that are coming through that are not used to the traditional ways of sales and going to salespeople and somebody turning up at their factory door to talk about things and so on. No. They want to do the vast majority of their research online and only really get in touch with the company that's selling at the last few stages. And I think the stats were that the buying groups were around between four and six. They're now predicted to be more around 10 to 12. And that's because more people within the business that is buying want to have an input in, "I want it to be like this, and I need it like that, and it needs to be compliant. It needs to be in a sustainability considerations," and all the regulatory side of things, and they're wanting to give inputs. So it's a group thing and not necessarily working well as a team, and that's another challenge from B2B. And another point around this, and if you see how things come together, is you ask what's the difference between consumer and business. But in AT&T, there's lots of instances where a consumer owns a business. Yeah. And when we talk about personalization, when that person comes to att.com, what content should I be showing him? Should I be cross-selling him? Does he have two accounts, that he has his personal account, his business account? Does he go to business.att.com? Like, all of those nuances are so interesting and opens a lot of opportunity. Yep. Especially for this audience, they require very different value propositions, from an individual, from a B2C with quicker, faster buying decisions. No one has to look over their shoulder to ask or confer.
On the B2B side, to Tom's point, you've got 6 to 10 to 12 buyers who are looking for different things. So really, really interesting. Any other question from anybody? Hopefully, that that answer your question.
Any hands up? No? Anyone? So Sorry. Go ahead. [Man] Claudia, it's good results with Firefly and definitely Photoshop. Just curious what are the conversations you have with creators, creative teams, the executives who are using it? You sell them into it or they excited to do it? Is the message, "Hey, I have this thing that will let you do more work?" Is that always due to debate? So there is an excitement there because you have identified an opportunity in an area that is very tedious. Right? You spend hours removing a background when instead you can be creating beautiful art. Right? So now with a touch of a button within seconds, that background is gone. I'm not spending my time there. I'm spending my time doing actually creating stuff. So it is very exciting for them, and it's how you talk to them about things and how you let them be on the driving seat.
And I think that I've not met a creative team or marketing team that have been sat there waiting, going, "Well, I've got nothing to do. It's all right. I've got loads of time to be able to do this thing that you're asking me to do." It's always within a stack of things. So if we can free up time and make them availability for the more interesting stuff, the creative side of things, then more for it. I think in the experience that we've had at ABB, certainly just in the AI and GenAI, over the last year, it's been getting over the fear.
As we've all seen in the press, this AI is going to take out jobs. It's judgment day, all of that sort of thing, and moving on from, well, actually, this is the-- Seeing is believing. This can actually make our lives a lot better. It's more of a-- Stealing Microsoft's term, more of a copilot. Get rid of the tedious stuff that you said, Claudia, and allow us to do more of the creative things and move forward that way. One thing that I heard once and I found extremely...
Interesting is AI is not going to replace you. It's going to replace a person that doesn't know how to utilize AI. So you have to be on the driving seat.
Fantastic. Thank you. Go ahead. [Man] I'm curious to know how this might apply to smaller organizations, specifically nonprofits, local nonprofits and profits like your area's schools, usually it's that kind of thing. The role it says, I'm a digital marketer, I'm working at positions, that used to be position of public relations or PR specialists, upper category people, anyone. I'm starting to notice that all those jobs descriptions really are striving, digital marketers and social media narrators, that kind of thing. You said there's place for, is it affordable for those smaller agencies to be able to do this transformation? I'll have a go.
I think there's-- This might be a bit contentious for the Adobe colleagues in the room. I think there's a responsibility for the large software providers to help with nonprofit organizations and educational organizations. And you see that through Microsoft. I don't know what Adobe do in that space, but I think there's a bit of a responsibility there to able help them in that space because I think that...
As with large organizations, there is a real opportunity for smaller organizations to get a bigger bang from their book without having to invest, quite simply, they don't have it, hundreds of thousands, millions of dollars in technology to enable them to do that, and they just don't have the money to do that. Whereas if there's the right leveling of that software, the leveling of that capability, which I'm sure there is, then they can get more out of it. So straight answer to your question is, yes, I think there is an opportunity within education, within local authorities, within nonprofit organizations to really get more out of this if the software companies come to the table with amicable or reasonable offering, which I'm sure they will. No pressure. Excellent. Hopefully, that's provided answer to you. Any further questions? Go ahead, please. [Woman] Claudia, I like what you were mentioning about the A/B testing and using machine learning for that. Kind of have a two part question. I'm curious, one, which software, Adobe software that you're using to leverage that? And then the second part is how to get buy in from those properties that you were talking about. Everyone wants piece, buy things on top, and hold a piece of the pie, how to get that buy in to test these things and kind of figure out they don't want to pay for it? Great question because our partners, we operated in that way of that manual segmentation and prioritization for many, many years. And then it's like, "Hey, there is this little thing called machine learning. What is that?" It's like, "It's heaven for us." So the way that we have a test and learn culture in our company, so that is a big, big, advantage. But even with that, you need to convince your partners. So one of the tactics that we've used and continue to use when something seems very risky is that we say, we're going to test it in 10% of the traffic, 90% is going to see the manual segmentation. It's 10%. Let me start with 10% of the traffic. I'll do the 50-50 split on the 10%. If it starts working, we will agree in increasing the size of the audience, and that's how we started that. And...
Then you have the data to support the decision. And it runs really, really well. We still have to turn it off sometimes when we have commitments with our partners. Let's say that we're launching, the new Samsung device, and there is a partnership and commitment with Samsung that says they need presence on the homepage a 100% of the time, then we turn off machine learning for that period of a time. But we know that it's not going to be the same performance, but it's a business use case that we need to support.
Fantastic. Anyone else? We've got roughly five minutes left.
Go ahead.
[Man] We talk a lot about this conversation but... My question is, what's next? What is coming next? Is the pattern recognition, with the AI image, etc. How do you see that? What comes up next? The crystal ball question.
Good question. Good question. So my personal view, not ABB, so disclaimer in there, is I think we've got what you described, there is going to be an error of hyper-personalization. I think that what Claudia answered in the previous question is within large organizations, everybody's vying for the same space, and we're certainly seeing that at ABB with some of the behaviors in the past and addressing that through personalization so that your experience is different to their experience is different. So you all get your own view, and that deals with almost cannibalizing AdWords and cannibalizing each other to get to the top of the list. So there's that piece, and that's we know that that's happening now. From a hyper-personalization side of things, I do think there's a risk, and I'm interested to-- It'll be interesting to see how this plays out where from an SEO side of things, you start injecting AI into that to make sure that your content is more optimized, the SEOs there, making sure it's got the right tagging and so on. You're not telling me that the Google's out there aren't going to start seeing that and going, "Hang on. You're using AI to do the SEO. Is that ranked to the right level? Should we start dismissing that?" And that's something that I foresee will be starting to play out over the next few years because there's a lot of really cool tools out there, some of them from Adobe, that can deal with the SEO side of things and put AI in that space. But what's the counter to that? Equals and opposites, and I think it's going to be interesting to see where the balance of, you still need to have the human input for, from an SEO side of things to make sure the personalization is relevant and actually it's coming from a human is not just putting rubbish in front of them because of an algorithm. Yeah. How do we balance that out? And I don't know. I think that's going to be really interesting. And I also-- Yeah. I also think it depends on your product and who your audiences are. Right? I think that you need to listen to what your customers need.
I'll say a year ago for one of our programs, we did some personalization from a messaging perspective, and we went very, very niche. And we crafted five different messages for five different segments, it didn't work. We didn't see increase in click-through rates or conversion. And it's as lovely of these messages are, for all of these five audiences, why am I spending my time on creating all of these variations if they're not moving the needle? So it goes back into testing and understanding what is your audience really looking forward. Yeah. And I think it really comes down to what we've all been saying about the using the AI as the copilot, but there still has to be that human element. It can deal with the monotonous work. But if you don't have the human element in there, then you're going to be missing the point and missing fundamentally the customer. It's also about agentic AI, playing a role in ensuring that, of course, using hyper-personalization to bring the right content to you and orchestrating your journeys. But as you mentioned, Tom, I think it's also about ensuring that you have that human touch so that there is the right data which comes in. I mean, you can have agentic AI doing it, but it's all about orchestrating it also and to ensure that the right message, the right focus goes in as well. So, yes, the future is definitely for agentic AI to replace a lot of work that we are doing with AI, but there's more to go in that. Yes. Definitely. Fantastic. Thanks, Bindya. Just for that, I just wanted to say thank you all so much for joining our session today. I know you had 10s and 20 other places to go, but I appreciate you giving us your time, and hope you enjoy the rest of your day. Thank you so much. - Thank you very much, Bruno. - Thank you. [Music]
