- il y a 1 jour
Human Touch, Smart Tech: How AI Elevates Customer Experience
Catégorie
🤖
TechnologieTranscription
00:00Thank you so much. Good morning, everybody, and thank you, Ariane, for such a lovely conversation.
00:05We thought before we kick off with our main panel, we'd do a bit of scene setting here.
00:11So, Rachel, thank you so much for joining me.
00:13Oh, I'm so excited to do it.
00:15Thank you. Well, you're fairly new at Adobe.
00:18Can you talk a little bit, just quickly, about your role and when you arrived and what you're responsible for?
00:24Yes. So, I started about six months ago, seven months ago,
00:28and I am part of the enterprise division, so thinking about marketing, technology, AI technology,
00:36even creative technology, but selling across businesses worldwide, enterprise businesses, so B2B, B2C.
00:44Fantastic. And, Rachel, as I say, we're talking all about how AI can enhance the customer experience.
00:52Obviously, you can't move for talking about AI at VivaTech this year.
00:57It was similar last year, but I feel like companies are really getting to grips with how to use the
01:02technology.
01:03How is Adobe differentiating itself?
01:05What kind of separates Adobe from the pack?
01:07I definitely think when you think about AI and you think about marketing, you think about creativity,
01:15all of these things coming together are very powerful, and that is what I think Adobe uniquely brings to customers.
01:24AI is, I think, an amazing technology.
01:27I'm very optimistic about how it can amplify creativity, amplify marketing internally,
01:33and then working with customers, we help them figure out, all right, how can you build the right creative experiences,
01:40how can you build the right marketing and customer experiences, and then using creativity, using marketing,
01:46and then using AI to scale it, bringing them to customers, engaging customers,
01:50and really making your brand resonate with customers.
01:54And the slide just there just shows how you run the kind of like the gamut of AI and marketing.
02:00It's kind of covering every boundary.
02:03Can we talk a little bit with this?
02:04This session is all going to be about human-centric AI.
02:08I kind of want to know what that means before we go a bit too deep.
02:12So, I definitely think that there's so much that AI can do, when I think about marketing, amplifying marketers,
02:21helping them move faster, be more agile, as they do everything from create messaging, content, campaigns,
02:28how do you experiment and how do you scale it?
02:31We internally, we're using all of our products, we obviously drink our own champagne, as the saying goes,
02:38but right now, like, I have marketers that are working on putting together sort of segmentation analysis, data analysis,
02:45and now with the agents embedded in Adobe's experience platform, they're able to do that very quickly.
02:50So, not that they wouldn't necessarily talk to data scientists or data engineers about putting together segmentations,
02:58but now they can do it themselves, right?
03:00And so that just means, like I said, they can be more agile and moving faster.
03:04They can do things like A-B testing, site experimentation, we're even working on figuring out how we put our
03:12brand concierge product
03:13that we announced at our summit back in March on our website, and really moving our website from maybe a
03:19traditional homepage experience
03:21to much more of an agentic experience.
03:23So, all of that, to me, starts with the individual, starts with the creativity and the imagination and the innovation
03:29that the individual brings to the, that the human brings to the conversation, and then AI amplifies it, helps you
03:35move faster,
03:37helps you scale, helps you be more experimental so that you can be more creative.
03:41Got it, got it.
03:42And then you did mention agentic AI there.
03:45Obviously, agentic AI has had a lot of momentum in the last 12 months.
03:49Again, how are you looking, you spoke a little bit, but how are you looking at integrating that within Adobe,
03:55but then also kind of helping your customers along that journey?
03:58Because there's a lot.
03:59Yes.
04:01So, I mentioned a minute ago our brand concierge product.
04:03So, one of the things that we've been working with customers, I think a really great example that'll resonate,
04:08is travel and hospitality.
04:10So, when you think about going into a hotel and talking with concierge, and you say, hey, I'm interested in
04:17these things,
04:18and then they can work with you on coming up with the perfect itinerary is.
04:21Now, imagine that as a brand, you have that concierge experience for all of your customers,
04:26almost irrespective of your industry, but a way for customers to come to you and have a conversation
04:32about what they're looking for, what they're hoping to accomplish,
04:34and then your agent, your concierge will, is helping that customer do it.
04:40So, I think that's a great example of, you know, how you're using agents to reimagine that conversation
04:48and that experience customers are having with your brand and with your website.
04:52And that's been something we've been, you know, that we've been actively working with customers on
04:55over the last several months.
04:57And now, I see AI as being a kind of like a great leveler,
05:01and some of the stuff that you've been talking about there,
05:03where you can do away with a lot of rote tasks,
05:06but also you can really level up very quickly, especially in areas like marketing.
05:11Yes.
05:12Like you were just talking there about the A-B testing,
05:15and all the things you can do with synthetic data that you may have had to wait weeks
05:19for your research company to come back with.
05:21But do you worry that AI is kind of coming for the CMO job at the same time?
05:27I think they're really interesting.
05:31At this point, the way I think about this is, it's less about AI coming to your job.
05:36It's more about, if you use AI, if you don't use AI, I would say, then that's the problem.
05:42I think the more comfortable we get with the technology, the more interested, the more use it.
05:47I encourage my team, like, let's be using it.
05:50Let's be experimenting.
05:53Because I think knowing that and doing that, one, gets people more comfortable,
05:56but two, opens up so many new opportunities.
05:59So I'm less afraid of replacement and more afraid of people just aren't using it.
06:04Right, right.
06:05So kind of get on board, get with the party.
06:08Speaking of which, we have some more party guests about to join us on stage.
06:13If you could give a big, warm welcome to Laura Crittan, VP of Retail Excellence at LVMH,
06:19and Chris Narayanan, Chief Digital Officer at Verizon Business.
06:23Please come on free.
06:24They're going to join us.
06:25Move this slightly.
06:32Thank you.
06:32Welcome.
06:33Welcome.
06:34Sorry.
06:34I hope this doesn't screw up any camera angles, but I just need to see everybody if I can.
06:39So how about, you know, we've just got kind of warmed up in just getting the scene set
06:45around how AI is kind of affecting the customer journey.
06:48But I wanted to ask kind of all of you individually how that's happening at each of your businesses.
06:53Because obviously we have some very different businesses serving very different customer needs here.
06:59I mean, Laura, how about we start with you at LVMH?
07:02Sure.
07:02How are you thinking about the customer journey and how that gets changed with this AI technology?
07:07I mean, at LVMH in retail, it's really all about the emotions.
07:13So we're very proud of our retail excellence.
07:16And it's true that for my teams and for me, the main objective is really to increase client experience.
07:25And this for us goes through our client advisors, one of the greatest assets we have.
07:33So it's not really not about tech for tech.
07:36For us, it's really about the quiet tech.
07:39And how can we use AI to really improve the client experience and this through our client advisors?
07:45Can you give some examples of, like, how that works in practice?
07:49Yeah, sure.
07:49One very concrete example is on client telling in retail.
07:56So this is where we mainly use AI and Gen AI, actually.
07:59So we have our existing tools and apps that we provide to all of our client advisors.
08:05And we're, at the moment, have launched a few initiatives with it.
08:10So it's really about simplifying the life of our client advisors, helping them identify faster, like, what clients to contact,
08:20which client to contact, when to contact them.
08:24And what kind of product to contact them with.
08:28So really helping them identify faster, contacting, saving time, but also obviously being more efficient and then having the right
08:39tools to have to use it.
08:42And for the Gen AI part, we're actually just testing something quite new.
08:48Very recently, it's a thank you message template.
08:52Okay.
08:53Yeah, so it's after purchase for our client advisors to support them with different tone of voices, but also personalized
09:01contact based on the client profile.
09:03But, again, it's really a support tool.
09:07It's for them to use it and adapt it.
09:10And what's interesting as well, and I think it's important to know, is that we've been doing this in the
09:15past already.
09:16Like, we've shared already, like, client lists.
09:20We shared templates with them as a support tool.
09:23So, this is more like, how can we use AI to make it, like, exponentially more powerful and really scale
09:30up in terms of client telling.
09:32So, they don't necessarily only contact the same clients over and over again.
09:36Got it.
09:37And then, Chris, is it fair to say that your application of AI is maybe kind of more under the
09:42hood than that?
09:43More, sorry?
09:44Kind of under the hood, behind the scenes.
09:46Yeah, I would say so.
09:47So, we've made a very deliberate effort to empower reps, both sales reps as well as service reps.
09:54So, we've taken a very conscious stance that Generative, in particular, isn't quite ready to go customer-facing, except in
10:02search.
10:02So, we've built a fair amount of capability to tool both sales and service.
10:09Service has had a far longer run of experience with this.
10:12But our work on AI began maybe 10 or 15 years ago in building predictive models like most other companies.
10:19And we've matured these to the point where predictive algorithms can actually recommend next best actions.
10:26Did you call it AI back then, or was it kind of called something else?
10:30I'm not able to hear you.
10:32Sorry, were you calling it AI back then, 10, 15 years ago?
10:34Oh, no, no, we just called it, you know, quant models and predictive models.
10:39We've realized now that as those models have matured and we're doing more with deep learning rather than just quant
10:46models,
10:47that these are genuinely what empowers AI.
10:50So, we've actually, the readiness to build AI and relevant experiences is founded on good data.
10:57And if the data is more real-time and your models are more comprehensive,
11:01it allows you to produce more relevant experiences.
11:05Generative creates an additional layer of relevance and context awareness that makes it more genuine to customers.
11:12So, when you take the magic of awesome customer understanding, which all lies in good data, as Rachel was saying,
11:19and you combine that with the ability to curate the best experience for customers, as Laura was saying,
11:25you end up with something that feels magical to customers, but feels easy for reps.
11:31And that's what we're actually working on.
11:33So, for example, if a customer calls with a frustration, your ability to anticipate that and to provide treatment strategies
11:41so that the problem or the resolution to that problem is already being prepared.
11:47The rep is able to say, we've taken care of the problem for you.
11:50AI is in the background working sort of magically to actually fix additional layers of the problem so that both
11:59the rep and the customer feel it's easier for them to do business.
12:03It works both for sales and service.
12:05So, our efforts have indeed been below the surface, but very thoughtfully so.
12:09Yeah, and you don't get that frustration of, I'll go away and talk to my boss and phone you back
12:14and all that kind of stuff.
12:16No, that's still good, right?
12:17I'll just be on hold for 10 hours.
12:18We run complex businesses.
12:19You're never going to get away from having, I've got to go away and solve the problem.
12:23But if you think about 85, 90% of the problems are relatively transactional.
12:29Yeah.
12:30And it should take no more than a click or two or a word or two if you're using semantic
12:34input to actually solve some of these things.
12:37And sometimes no words at all if you're able to do them predictively.
12:41And that's the work we're doing.
12:43Fantastic.
12:44And Chris was just talking about, like, some of the magic that AI can unlock in these situations.
12:51Rachel, I wonder if you could talk about, like, a time when, like, AI has kind of unlocked a magical
12:56moment in, kind of how you experience, how customers are experiencing Adobe products.
13:02One of the things that was very interesting that we found out, we sort of knew it intuitively on the
13:09marketing team.
13:10But AI helped us not only see it, but see the data and then be able to share the data
13:15with the sales team.
13:16So when we put together customer journeys, you know, we always think of them first, sort of digital first.
13:22Like, what are the ads?
13:25What are the emails?
13:26What are the social posts, social media?
13:29What are the digital things the customer is engaging with in the early stage of that journey?
13:34And so, you know, we had a hypothesis that, all right, this is, the digital footprint was good in the
13:41upfront, but at some point you have to insert either an engagement with another customer to have that customer share
13:46how they're using Adobe,
13:49or have the customer slash prospect engage in person or virtually in an event.
13:56But when we started to do experimentation, and we were creating customer journeys, and because of, this is where, you
14:03know, I think this, what you get with AI in terms of scale is very powerful.
14:07Because we were able to create assets, and we were able to customize journey almost down to the individual, we
14:13could quickly see how they were responding to various marketing assets we were putting out, campaign assets.
14:19And we really did see the spike, like it was a clear signal when you inserted an event, and it
14:26could have been virtual or in person, at a certain point in that journey, that prospect slash customer's desire to
14:32engage, I mean, it basically 3X'd our ability to take that, you know, take that deal or that engagement to
14:40conclusion and to positive conclusion.
14:41But that was the thing that, yes, we, you know, we may have thought of it intuitively, but being able
14:47to do the experimentation, being able to create the assets at scale, and then personalize them, really absolutely showed us
14:54that, yes, that thing matters.
14:57And it matters so much that it's worth, like, us almost rejiggering then the different ways we were doing the
15:02marketing engagement and the marketing journeys, customer journeys, to make sure we were embedding this insight.
15:07And how was that different from the way that you were perhaps optimizing marketing, you know, five years ago?
15:14Was it like the real-time element?
15:16Yes.
15:16Was it like, what's the difference?
15:18So, you know, in a lot of years that I've been doing this, I've been doing marketing, when you think
15:23about campaigns, everyone talks about it, it's put a marketing campaign out there.
15:27But, you know, if you even go back a couple of years ago, to create the assets, to have the
15:33assets in multiple languages, to have the assets span a variety of segments, it took weeks slash months to get
15:40all of those things together.
15:42And now it takes days.
15:44Like, I would have been able to do what we're doing now, it would have been taking me seven, eight
15:48weeks.
15:49Now it literally takes me about six days to create these assets and then scale them across the channels I'm
15:55going to put them in, get the data back, and then say, okay, this is working, this isn't.
16:00Let's swap these assets out.
16:02Or, to my previous point, hey, it turns out that when we do engage with customers and event experience, it
16:09helps us with closing, so let's make sure we're doing more of those.
16:12But just that stuff, it would have taken months and months and months to learn.
16:15And I guess also the stuff that, like, wasn't working, but you had that kind of sunk cost syndrome that
16:20we've kind of got to go ahead with it anyway.
16:22Exactly.
16:22So you're like, ah.
16:22Whereas if it only took six days, there's less of it.
16:24We don't love it, they don't love it, but let's keep going because we spend money.
16:27Right.
16:27Now we can just say, it's not working, let's pull it out.
16:30That makes sense.
16:31Yeah.
16:31And then, Laura, I wanted to talk more about the kind of retail experience.
16:35I mean, obviously, you know, in the luxury world, things like heritage and trust are just so paramount to everything
16:42you do.
16:43And authenticity.
16:45I mean, that's the heart of everything you do.
16:47And obviously, there is, you know, we're living in a more kind of synthetic world where things can be copied.
16:54But the thing that you can't take away is that moment when you walk into a store and you're greeted
16:59by, you know, a customer service agent.
17:02I shouldn't call them an agent.
17:03A representative, not an agent.
17:05They're not AI just yet.
17:07Can you talk about, like, any, like, aha moments you've had in that kind of blending, literal, real life, human
17:15interaction, but with the AI technology, but how that's actually really helped enhance it, not kind of take away.
17:22And also, like, you know, I'm guessing these people aren't just, like, looking at a script kind of constantly just,
17:28like, being fed all of these different attributes.
17:30So, yeah, can you just, like, walk me through how that's changed?
17:33Yeah, I think what's very interesting about your question is, I mean, for us, it's really our client advisors, as
17:39I said, are our biggest asset in stores.
17:42And so, what for me was kind of the aha moment was really the adoption.
17:48I was very surprised when we tested and we started with pilots.
17:52And as, again, I mean, for us, it's not tech by tech.
17:54It needs to make sense.
17:55So, we always go step by step, piloting, testing in a few stores.
18:00And I was really surprised by the adoption rate and the pace that the client advisors were starting to using
18:07all these great support tools that we were giving them.
18:10And I think what's really important there is that, for us, the point is to take away from, like, operational
18:16tasks and fluidify these and give them more time to really, on high added value interactions, to create, like, this
18:25special bond with their clients.
18:27And is that similar at Verizon, too?
18:30Like, how have you been a cheerleader within the organization to encourage people like your customer service agents and that
18:37whole organization to adopt this tech?
18:39Because lots of people just don't like change, right?
18:42They've been doing this for a while.
18:43They're good at their jobs.
18:44I don't want this new thing coming in.
18:47So, we focused a lot on simplifying the user experience to the point where the cognitive load is very low.
18:55There's been a fair amount of testing on this over the last five, seven years.
18:59But with AI coming into the fore, we've got it simplified even further.
19:05I keep telling my teams that if you've got to train people on how to use simple technologies, it's not
19:12designed well enough.
19:14Any technology that requires an expert team or a center of excellence or, you know, in theory, I would set
19:22a very high bar for design in such a way that it's adopted fairly easily.
19:27So, I'll give you an example.
19:29With our customer service reps, we've now created a fairly simple, this is not new in some ways for customer
19:36service, to give them red, yellow, green for customer state of mind.
19:41But our current customer emotional state is fed by real-time data.
19:46It's fed by things they might have said one hour ago or 15 minutes ago, as well as things they
19:52might have done online.
19:53So, transcribing the audio in real-time?
19:56Yes, we process audio in every channel where we're legally allowed to capture and process it.
20:01We transcribe, summarize, derive insights, merge it into the quant models so that it feeds back into the red, yellow,
20:10green.
20:10So that the representative facing the customer knows if the customer is safe to sell to or is going to
20:17be less receptive because they've had some issues.
20:19So, that's an example of taking a fairly old sort of practice of saying, is the customers, is they going
20:25to be, like, are they going to be happy or not, to something that's much more real-time without changing,
20:31perhaps, how the rep should think about it.
20:33So, we've made a lot of headway in simplifying experiences.
20:35On the sales side, we've had some similar headway in sort of taking very, like, massive amounts of data and
20:42making it super simple for reps and giving them one page on their daily login screens where they know exactly
20:48who they're going to reach out to and why.
20:50And we've created emails and scripts for them to follow that are AI-driven.
20:54So, AI is sort of folded into their daily experience to make it easier rather than something else they've got
21:01to go learn how to do.
21:02So, it reduces cognitive load and it increases adoption the way, yeah.
21:07And what does the tech stack look like there?
21:09Like, what do you build versus rent versus buy versus, you know, like, how does that work out?
21:17So, this is so fast evolving that you almost cannot buy your way into this.
21:25Crucially, I think data is the underpinning.
21:27And I think the focus the company has put over the last dozen, 20 years is on data quality and
21:35bringing key pieces of data together.
21:38Transactional data, web data, customer account level data, interactions data, and sentiment data, which is sort of, and now we're
21:47getting better at predicting sentiment based on all of the previous things.
21:50So, all of this being not just in one place, but in a way that it actually connects with each
21:56other, makes the development of AI algorithms that much easier.
22:00So, yes, that's one underpinning.
22:02The second is an AI stack that's fast developing, that's partially rented and partially built.
22:08No, we're not building LLMs at Verizon, but with the massive data we've got, we partner closely with Google and
22:15others to empower the creation of large-scale models
22:19and to bring LLMs into the daily lives of our reps and our customers.
22:25And then there's a whole set of operational movements that need to happen, to your point about adoption.
22:30Those are not taken lightly.
22:32So, we do take training very seriously, but we try to design in such a way that training isn't the
22:38primary focus.
22:39It's design that's the primary focus.
22:41Right.
22:41Training becomes an important thing that happens, but easier if you've designed it right.
22:46For sure.
22:46So, that's the way we approached it.
22:48And how are you approaching it at LVMH?
22:51Like, where does AI kind of sit within the org?
22:54I thought it was a very interesting, actually, what you said, Chris, about you shouldn't have a user guide on
23:00how to use it.
23:01And it's exactly this, I think.
23:03I mean, we have two key apps in stores, and we really try to test everything in advance with key
23:09users to really make sure that when we launch it,
23:12everyone knows exactly where to click, how it works.
23:14And, I mean, it's a bit weird to say, but for us, it's really we want our client advisors to
23:20spend as much time as possible on our applications.
23:23So, they need to be super seamless, so they spend more time really with the clients.
23:27And training is a very important part.
23:30I mean, obviously, for our client advisors, but also in the head office, I mean, we have a really great
23:36tech team that are working on trainings that are very concrete and personalized on all the different teams.
23:42because, at the end, AI is a very abstract concept.
23:46So, you really need to make it user-friendly when you explain it.
23:50And you need to understand how you can use it in your daily lives.
23:53What are the benefits?
23:54How can you use it?
23:55What are also risks?
23:56And really to kind of have a concrete example of what are really the benefits.
24:02So, I think there we had great support from the tech teams that really did, came into all the different
24:07departments and spend really a lot of time with everyone explaining it.
24:12And then, obviously, we also have from LVMH group side a lot of training modules and, I mean, our own
24:18AI assistant, Maya.
24:20Nice.
24:21And how do you keep up personally?
24:23Like, you can't go a day without a new model, a new breakthrough, a new consumer app, you know, OpenAI
24:31has gone down, Google's going up.
24:33Like, how do you kind of stay abreast of everything?
24:36I mean, viva tech, no?
24:38Ah, very good part.
24:41Good answer.
24:41No, indeed.
24:42I mean, obviously, I think it, actually, it's a fun story.
24:45I'm going to tell something, a personal story, but I was very surprised the other day.
24:49My parents that are over 75, they went on a trip to Japan and they told me, like, yes, I
24:55use ChatGPT to understand what to do.
24:57I was like, oh, wow, who told them?
24:59They have a good time, though.
25:02It sent them to all the right places.
25:04It didn't hallucinate.
25:05Definitely.
25:05Oh, good.
25:07And, Rachel, I mean, you have a lot of interaction with CMOs, obviously, at Adobe, one of your biggest customer
25:12bases.
25:13Where are CMOs at on this journey at the moment?
25:17And also, where are they kind of taking ownership of AI in their organizations?
25:23Technology is often the kind of the domain of the CIO, for example.
25:28So, yeah, what are you kind of seeing?
25:30And obviously, I imagine it's a spectrum and some are further forward than others.
25:34Yes.
25:34And we were talking about this a little bit backstage.
25:37So, I definitely think that, you know, with AI, it's CMOs are asking constantly, like, how can they use it?
25:44They're experimenting with it.
25:45They're actually moving, you know, in certain cases to production on certain things.
25:48But I definitely think the conversation between CMOs and CIOs around AI impacted work streams, where should AI be implemented,
26:00like, who owns what particular implementations, I think those are really important conversations to have.
26:06Two weeks ago, I was in New York, and I was doing a dinner with CMOs and CIOs to talk
26:10about just these things.
26:11So, what are the CMOs helping to do?
26:15What are they hoping to accomplish?
26:16Kind of, what are the partners they want to work with?
26:18And then, where do CIOs have questions?
26:21And almost vice versa.
26:23So, like, when you think about companies, tech companies like OpenAI, you're like, okay, how, when they talk to CIOs,
26:30they're going to have one sort of, you know, engagement or talk track, if you will.
26:33But how do we actually make sure that we're talking to both audiences at the same time and bringing them
26:39along on where AI shows up in, like I said, various work streams, and just making that partnership between the
26:45two groups that much stronger?
26:47I think so they can move that much faster when they are not only experimenting with, but then deploying to
26:53production AI workloads.
26:55And is there a common thread in terms of the questions that CMOs are asking you about at the moment?
27:01Like, is there a pain point that seems to go across industries or, you know, excitement that they want to
27:08be, like, moving towards and they can't move quickly enough?
27:11I think there's a lot of excitement.
27:13I do think that one of the things we talk a lot about and one of the things that Adobe
27:17Price is self-honest, you know, making sure that its guardrails are adhered to.
27:22So, if you're a CMO, you want to make sure, hey, I want people to be able to use technology.
27:26I want people to be able to build campaigns, build creative, build content, but obviously being within brand guidelines, so
27:31that's important.
27:32But I actually think that there's just so many questions still right now because it's such an area of, I
27:38mean, it's growing.
27:39There's a lot of experimentation.
27:41So, when I talk to CMOs, lots and lots of questions come up, not just about marketing, but then also,
27:46again, working with their tech counterparts.
27:48What can they learn?
27:49And what can they be doing better?
27:50And then I get a lot of questions on how, when we sort of internally, you know, sort of use
27:55Adobe on Adobe, how do we encourage people to experiment?
28:00And one of the things that we started doing on my team is I encourage people to fail.
28:05I say, look, if we're not experimenting and we're not, you know, failing, it means we're not learning something.
28:12So, don't be afraid to try things.
28:14Don't be afraid to recommend things or come up with an idea that we can try.
28:17Because I would rather try it and learn from it, even if it doesn't go the way we thought, but
28:22at least we've learned something and then we can take it into the next experiment.
28:25So, I think that's really important, especially when you're talking about rolling it out or training folks, get them comfortable
28:30with, it's all right if it doesn't go right the first time.
28:33We're going to learn and we're going to move fast and learn, and that's goodness.
28:36Which is a new mindset for a lot of companies, right?
28:39Like, move fast and break things is not, you know, it doesn't work in every industry quite as simply as
28:46it sometimes does in tech.
28:47And often it doesn't work in tech either, as we well know.
28:50We've only got a couple of minutes left and I wanted to end, it's been really great that we've been
28:55talking about a lot of kind of concrete examples about what your companies have done so far and have been
29:01working on.
29:02I wanted you to kind of look in your crystal balls a little bit and give me some examples of
29:08kind of what you're hoping AI can unlock in terms of, you know, improving the customer experience in, say, the
29:14next three to five years.
29:15Like, what's like your grand vision, your grand dreams for what this technology can do?
29:21And perhaps, Chris, I can start with you.
29:23Start with me?
29:23Yes.
29:24Oh, sure, why not?
29:28We're working very hard to ensure that AI feels very seamless to our customers so that within 18 months to
29:37two years, a lot of the things that customers today expend effort on,
29:42calling to understand the status of a bill, of a payment, or, you know, adding people to their plan, what
29:49have you, should become as easy as giving simple commands on search.
29:55So we, you know, we see that happening now.
29:57We're actually testing some of these things now.
29:59So I think it's very easy to foresee that within the next two to three years, you will see the
30:06level of automation go up.
30:07Automation is one where you give a command and the system is able to follow through.
30:12But autonomy is a little bit further away.
30:14We want to move towards autonomy more consciously, thoughtfully, so that a system doing things on your behalf without necessarily
30:23you commanding it is something that we're working on as well.
30:27But that will be coming, that's what we call agentic, right?
30:30When you have multiple agents working either in sequence or in parallel to take actions on behalf of humans, that's
30:37something we're working on both for customers.
30:39So to build delight into experiences is the key for us and to figure out how we can do it
30:45proactively.
30:45If you can anticipate customer needs, we've done our jobs.
30:49Very quickly, Laura.
30:51Yes, very quickly.
30:52I think it's really, and I'm going to repeat myself, about really supporting our client advisors, liberating as much time
30:59with low added value tasks and giving them more time to spend really with the client, engaging and creating these
31:05emotional bonds.
31:07Fantastic.
31:08I'm afraid we're all out of time, but I think that was a really inspiring conversation.
31:12Hopefully, it's given you lots of food for thought about where AI is going in the customer experience.
31:17Thank you so much, Rachel, Chris, Laura.
31:20Thank you, everyone.
31:21Thank you.
31:21Good morning.
31:23Thank you.
Commentaires