# Brewing Smarter: How HEINEKEN Is Using AI To Revolutionize Its Global Operations

An episode of [Smart Talks with IBM](https://invisiblenetwork.ai/podcast/1558889546), 21 Apr 2026, 43 min.

Audio: https://pdrl.fm/c50a0c/podtrac.com/pts/redirect.mp3/tracking.swap.fm/track/SxlTEPDY7xDg35RXkASs/mgln.ai/e/p457950/traffic.omny.fm/d/clips/e73c998e-6e60-432f-8610-ae210140c5b1/85547dfc-49bd-4e63-a6cc-ae32006cd832/aa79ddad-1874-4af6-b1c3-b432014f3ce1/audio.mp3?utm_source=Podcast&in_playlist=6d9e2ef8-e4e9-45e9-9e13-ae32006cd844

The HEINEKEN Company is one of the world’s pioneering global brewers. Founded in 1864, it continues to innovate as it pursues its ambition to become the world’s best‑connected brewer. For the Season 7 premiere of Smart Talks with IBM, Malcolm Gladwell sits down with Surajeet Ghosh, HEINEKEN’s Chief AI Officer, in front of a live audience at SXSW to explore how the company is using data and AI to transform its operations. This is a paid advertisement from IBM. The conversations on this podcast don't necessarily represent IBM's positions, strategies or opinions. Visit us at https://www.ibm.com/think/podcasts/smart-talks See omnystudio.com/listener for privacy information.

## Transcript

[0:02] Speaker 1: I'm Malcolm Gladwell. Welcome to Season seven of Smart Talks
[0:05] Speaker 1: with IBM. This year, we're exploring new stories about how
[0:09] Speaker 1: companies are using the latest advancements in AI and quantum
[0:13] Speaker 1: computing to create smarter business. For the first episode of
[0:17] Speaker 1: the season, I flew to Austin, Texas to join Sergei
[0:21] Speaker 1: Ghosh on stage at south By Southwest. Sergei is Chief
[0:25] Speaker 1: AI Officer at Heineken, the world's pioneering beer company. Founded
[0:30] Speaker 1: in eighteen sixty four, Heineken is deep roots, but it
[0:33] Speaker 1: continues to push the boundaries of innovation today. In twenty twenty,
[0:38] Speaker 1: the company came up with a goal to become the
[0:40] Speaker 1: world's best connected brewer. Surge plays a key role in
[0:44] Speaker 1: leading that transformation, and I sat down with him in
[0:47] Speaker 1: front of a live audience to understand what that journey
[0:50] Speaker 1: looks like and what it takes to reinvent a global
[0:53] Speaker 1: company from the inside out. And before we get to
[1:00] Speaker 1: the question of what you do in your job, So
[1:02] Speaker 1: I'm really interested in people who have jobs it didn't
[1:05] Speaker 1: exist for most of their life, and I'm curious how
[1:07] Speaker 1: you got there.
[1:08] Speaker 2: Yeah, first of all, thanks for having me here. Yeah,
[1:11] Speaker 2: actually it did exist and people sometimes don't realize AI
[1:14] Speaker 2: is not new. It's been there for seventy five years,
[1:18] Speaker 2: since nineteen fifty. It just changed over time. How the
[1:21] Speaker 2: application is happening, right, So one thing to keep up
[1:24] Speaker 2: with is as AI became more popular and more embedded
[1:27] Speaker 2: in business, how do we upskill ourselves to stay at
[1:30] Speaker 2: par with the technology trends. So the preparation for me
[1:34] Speaker 2: personally started actually a long time ago. So when I
[1:36] Speaker 2: was in grad school in US, and I also live
[1:38] Speaker 2: in US, by the way, for a long time.
[1:40] Speaker 3: You're Indian.
[1:41] Speaker 2: I'm an Indian originally, but it's news. I did my
[1:44] Speaker 2: grad school here and there. Actually I started taking courses
[1:47] Speaker 2: in newer networks and at fission diagents back in two
[1:50] Speaker 2: thousand and two, and it wasn't popular back then, but
[1:53] Speaker 2: I was just curious, what is it? Maybe it's the
[1:55] Speaker 2: next big thing. And I'm so glad I did that.
[1:58] Speaker 2: Is that sort of helped me? That foundation?
[2:01] Speaker 1: What was it? You said you were you were curious
[2:04] Speaker 1: about it? You were curious about it? Why what caught
[2:06] Speaker 1: your eye about it?
[2:08] Speaker 2: Was very different because the main difference was before that
[2:11] Speaker 2: I was an engineer by profession. I went to engineering
[2:14] Speaker 2: college everything is rule based, Everything is based on a formula,
[2:17] Speaker 2: a physical equation. AI is something different because based on
[2:21] Speaker 2: data and statistics, it never gives you a clear answer.
[2:24] Speaker 3: It gives you a probability.
[2:26] Speaker 2: And I just thought this is very interesting because if
[2:28] Speaker 2: you're trying to solve a problem, you don't know exactly
[2:31] Speaker 2: how to solve it. There is no equation. How do
[2:33] Speaker 2: you get around that? I think that's where AI comes in.
[2:36] Speaker 2: It finds those patterns within data and comes up with
[2:39] Speaker 2: some prediction that intrigued me.
[2:42] Speaker 1: So this is what year that you start.
[2:45] Speaker 2: I started dabbling, let's call it dabbling in AI was
[2:48] Speaker 2: two thousand and two.
[2:49] Speaker 3: It was almost twenty four years ago, four years so.
[2:51] Speaker 1: But what you were playing with in two thousand and
[2:54] Speaker 1: two was an extremely primitive version of what we have now.
[2:58] Speaker 2: I think it was very relevant because the way I
[3:01] Speaker 2: see it, should I have skipped all the foundations that
[3:04] Speaker 2: I learned over the years and just gone to the
[3:06] Speaker 2: current state. Maybe, But when I look back, I think
[3:10] Speaker 2: that foundation really helped me because back then, and surprisingly,
[3:14] Speaker 2: by the way, new on networks. When I talk about that,
[3:16] Speaker 2: it's still very valiant and relevant within AI. The entire
[3:20] Speaker 2: foundation is new on networks. So I think that foundation
[3:23] Speaker 2: really helped and I still find it very relevant and
[3:25] Speaker 2: I apply it day to day.
[3:27] Speaker 1: Yeah, imagine and having a conversation with you twenty years
[3:30] Speaker 1: ago and I say what are you up to and
[3:34] Speaker 1: you say, I'm playing with this thing neural networks early version.
[3:38] Speaker 1: Would you have used the term artificial intelligence?
[3:41] Speaker 3: Probably not.
[3:42] Speaker 2: I probably would have used something is called statistics, which
[3:45] Speaker 2: is everyone is aware of. Back then, it was more statistical,
[3:48] Speaker 2: so you don't have these big algorithms at that point.
[3:52] Speaker 2: But then something happened. I don't know if you heard
[3:53] Speaker 2: of this company called Cago. They used to host these
[3:56] Speaker 2: sort of data science competitions and anyone can participate and
[4:00] Speaker 2: if you do really well, you get a price. That
[4:02] Speaker 2: was a good motivation, just to see, Okay, I learned something,
[4:05] Speaker 2: let me apply it and see how good I am
[4:07] Speaker 2: I'm getting at it. So I think that was my
[4:09] Speaker 2: first entry point where I really got hands on into AI,
[4:12] Speaker 2: and that probably stayed with me for a while. I
[4:15] Speaker 2: think that was back in two thousand and six. That's
[4:17] Speaker 2: what I started getting hands on. And the funny thing
[4:19] Speaker 2: is when you look at these Caggle competitions, the use
[4:22] Speaker 2: cases they used to give actual industry applications.
[4:25] Speaker 3: So you were really dealing with business.
[4:27] Speaker 2: Problems applying AI to solve it. And then you know,
[4:30] Speaker 2: wait a minute, a medical company is using it, a
[4:32] Speaker 2: manufacturing company is using it, a banking is using it,
[4:36] Speaker 2: and this is two thousand and six, so it already started,
[4:39] Speaker 2: and then it just yeah, today, it's a different ballgame now.
[4:42] Speaker 1: So you you came to Heineken.
[4:45] Speaker 2: When twenty twenty twenty twenty right middle of COVID, right, and.
[4:51] Speaker 1: Where you brought in to be the chief AI officer?
[4:53] Speaker 2: Was you excorre it and the title changed, but yes,
[4:56] Speaker 2: I was the global leader.
[4:57] Speaker 1: Of Yeah, yeah, And what made you want to take
[4:59] Speaker 1: the job?
[5:01] Speaker 2: I was actually working for Amazon at that point, but
[5:04] Speaker 2: when I looked at Heineken and I thought, okay, this
[5:06] Speaker 2: is a legacy, traditional company, right, and AI was not
[5:10] Speaker 2: a capability embedded at that point of time. So it's
[5:13] Speaker 2: a great opportunity if I can start something from scratch,
[5:16] Speaker 2: really build it across the entire valuation of Heineken. I mean,
[5:20] Speaker 2: that's probably the best job anyone can even ask for. Yes,
[5:23] Speaker 2: it's of course a lot of responsibility that you have
[5:26] Speaker 2: to make sure that you really build the right products
[5:28] Speaker 2: and right capability. But that also happened, so I look back,
[5:32] Speaker 2: it's quite fulfilling.
[5:34] Speaker 1: But it's also if I might play that was advocate
[5:36] Speaker 1: for a moment, you're also taking a risk going into
[5:39] Speaker 1: an established how long has Heineken been around one hundred
[5:43] Speaker 1: and sixty two years to be specific, eighteen sixty four.
[5:46] Speaker 1: It's a very different proposition walking into one hundred and
[5:48] Speaker 1: sixty year old company and saying I want to bring
[5:50] Speaker 1: the future to the way you operate. Then it is
[5:53] Speaker 1: with the startup.
[5:54] Speaker 3: That is true, but it's also a challenge.
[5:56] Speaker 2: It's a good challenge and also that Heineken is also
[5:59] Speaker 2: looking external. There are companies that are picking up speed
[6:01] Speaker 2: and embedding and adopting AI. So should we fall behind,
[6:05] Speaker 2: not really, so we also need to pick it up.
[6:07] Speaker 2: So I thought it was a good challenge because the
[6:10] Speaker 2: use cases were there. The opportunity I put sense the
[6:13] Speaker 2: business really was having the appetite let's do something different.
[6:17] Speaker 2: When we apply AI in a corporate setting like this,
[6:20] Speaker 2: it's super important to understand how the business actually works.
[6:23] Speaker 2: What's the value Chaine looking like, what are the nuances?
[6:25] Speaker 2: Where can I And once you get an understanding, it
[6:27] Speaker 2: took me some time, by the way, to understand the
[6:29] Speaker 2: full business and the complexities. But once you cross that threshold,
[6:33] Speaker 2: you figure out what's feasible what's not.
[6:36] Speaker 3: Then it opens up.
[6:37] Speaker 2: Wait a minute, within the valuation and I see ten
[6:39] Speaker 2: areas I can optimize.
[6:41] Speaker 1: You say, once you want to see in the business,
[6:43] Speaker 1: describe the business the Heineken puzzle.
[6:45] Speaker 3: So well, puzzle. Let's see if it's puzzle.
[6:48] Speaker 2: After I explained it's actually we start with the procurement
[6:51] Speaker 2: where you get the glasses, cans and all the raw materials.
[6:54] Speaker 2: Then it comes to the brewery where the magic happens.
[6:56] Speaker 2: That's what behineqn beer is produced. Then it goes to
[6:59] Speaker 2: the tributors. Basically supply chain takes over. Then it goes
[7:03] Speaker 2: to the customers. And what we refer to as customers
[7:06] Speaker 2: are the bars, restaurants, retail stores, moment pop stores, convenience stores,
[7:11] Speaker 2: and that's where actually consumers then come and actually consume
[7:14] Speaker 2: the product. So that's actually the value chain. It's actually
[7:17] Speaker 2: pretty linear when you think of it, but there are
[7:19] Speaker 2: nuances depending on the country and the market. There are
[7:23] Speaker 2: some specific rules and guard rails that you have to
[7:25] Speaker 2: be aware.
[7:25] Speaker 1: Of, so you have that process going on all around
[7:29] Speaker 1: the world and across multiple brands.
[7:32] Speaker 2: Multiple brands, multiple countries, multiple operating companies.
[7:35] Speaker 1: Yeah, from your perspective as someone who is the chief
[7:39] Speaker 1: AI officer, what are the tasks in front of you?
[7:42] Speaker 3: What's your opportunity?
[7:43] Speaker 2: There any process that you think that is maybe not
[7:47] Speaker 2: digitized or maybe not data driven, you can optimize. I
[7:51] Speaker 2: look at it like a pendulum. So one side of
[7:54] Speaker 2: the pendulum you have complete gut based decision making. The
[7:57] Speaker 2: other extreme is completely data driven. So the idea is,
[8:00] Speaker 2: can we swing this spendil amtile bit towards data driven
[8:03] Speaker 2: From where we are?
[8:04] Speaker 1: Give me a specific example of a problem you set
[8:08] Speaker 1: out solve or address.
[8:11] Speaker 3: There are quite a few.
[8:12] Speaker 2: But if I want to pick one of give the
[8:14] Speaker 2: most fun one. Fun one may be the most most
[8:16] Speaker 2: complex one. Let's bring that one.
[8:18] Speaker 3: I think.
[8:19] Speaker 2: So we spend quite a bit on advertising, and Heineken
[8:22] Speaker 2: is a largely a lot of it marketing company, and
[8:25] Speaker 2: we really care about our brands.
[8:27] Speaker 3: And products, you know, we are almost obsessed with it.
[8:30] Speaker 3: Let's take an example.
[8:30] Speaker 2: Let's say you have X million dollars as your budget,
[8:34] Speaker 2: and you have two brands. Let's say Heineken and tossa case,
[8:37] Speaker 2: I think the audienceale will be familiar with that.
[8:40] Speaker 3: And then you have three.
[8:41] Speaker 2: Touch points TV, YouTube, Instagram, and I want to optimize
[8:45] Speaker 2: my advertising budget between different brand and touch points, so
[8:50] Speaker 2: Heineken on Instagram?
[8:51] Speaker 3: How much should I spend?
[8:53] Speaker 2: It's a very easy question to ask, but to actually
[8:55] Speaker 2: solve this we have to study historically how these performed
[8:59] Speaker 2: and then create a more rule and then predict if
[9:01] Speaker 2: I allocate my budget in this way that's probably more optimal.
[9:05] Speaker 2: Before it was more like somebody took a gun based
[9:07] Speaker 2: decisions saying okay, here goes xpillion, here goes five million,
[9:11] Speaker 2: and we said no, no, no, that's not the right proportion.
[9:13] Speaker 1: What did the AI tell you about the accuracy of
[9:16] Speaker 1: those spending decisions?
[9:19] Speaker 2: In the past, we looked at the return on investment
[9:22] Speaker 2: from this advertising, and so how much incremental volume or
[9:24] Speaker 2: volume of br reselling or revenue are recreating and we
[9:28] Speaker 2: find out can we improve that. It's a moment you
[9:30] Speaker 2: apply AI, and when we look at that this significant improvement.
[9:34] Speaker 2: In some cases we have thirty percent uplift, thirty percent,
[9:37] Speaker 2: thirty percent three zero not everywhere. In some places we
[9:41] Speaker 2: got but it ranges between ten to thirty percent uplift
[9:43] Speaker 2: depending on the type of AI product you're building, and
[9:46] Speaker 2: that impacts the top line. So it's very easy to
[9:49] Speaker 2: also realize that value. People get to see it, so.
[9:52] Speaker 1: You say, oh, we we can do a way better
[9:54] Speaker 1: job if we spend x more or x less in
[9:56] Speaker 1: this particular area. Give me another example of another.
[10:00] Speaker 2: No one would be we have a very big large
[10:02] Speaker 2: salesforce within Ainiken. So these sales reps what they do
[10:06] Speaker 2: They go to the outlets, the bars and restaurants, and
[10:09] Speaker 2: they maintain that that human to human relationships with these
[10:12] Speaker 2: our customers is super important. To maintain that, and they
[10:15] Speaker 2: go solve the customer problems. Let's say someone is out
[10:18] Speaker 2: of stock, someone is about to churn, or the surprise mismatch,
[10:21] Speaker 2: something like this. And before there is to go like this,
[10:24] Speaker 2: let's say a sales rep on a day to day
[10:27] Speaker 2: job has to visit five places A, B, C, D E, five.
[10:30] Speaker 3: Different outlets and he used to go ABCDE.
[10:33] Speaker 2: Turns out, the model tells you on any given day,
[10:37] Speaker 2: if you optimize taking into account the traffic conditions, instead
[10:41] Speaker 2: of going from that linear route, you go.
[10:43] Speaker 3: To D first and then to B, then to C,
[10:46] Speaker 3: then to E and then to A.
[10:47] Speaker 2: And the reason for doing that is the model tells
[10:50] Speaker 2: you if you visit customer D first, he has the
[10:52] Speaker 2: biggest problem that needs the most amount of time to
[10:54] Speaker 2: be solved, and that's how it optimize.
[10:57] Speaker 3: And also the sales reps now they are big.
[11:00] Speaker 2: I'ming so educated with some of these AI models. They
[11:02] Speaker 2: are now becoming business advisors, so they are no longer
[11:06] Speaker 2: just solving little problems. They are having the time to say,
[11:10] Speaker 2: what else can I do for you as the customer?
[11:12] Speaker 2: So that I think it was a big change within
[11:15] Speaker 2: Heineken because it impacted a lot of people that were
[11:17] Speaker 2: using that.
[11:18] Speaker 1: In an instance, it requires not just building a model
[11:21] Speaker 1: that can be smarter about how people should spend their
[11:24] Speaker 1: time and what they should say, but you have to
[11:26] Speaker 1: obviously educate your salesforce to believe in what I selling.
[11:32] Speaker 1: Tell me about that piece? Is that a piece that
[11:34] Speaker 1: you that you're a part of or is someone else?
[11:36] Speaker 2: Yeah, that's also part of because that is super important.
[11:38] Speaker 2: I think we can build the best models. Best algorithm's
[11:41] Speaker 2: highest accuracy doesn't mean anything if it's not used the
[11:44] Speaker 2: right way. So what we do we have within our
[11:47] Speaker 2: company a pretty big upskilling program. So bring everyone along
[11:51] Speaker 2: in common understanding, basic understanding over AI does not everyone
[11:55] Speaker 2: needs to understand newer networks or algorithms, right, But what
[11:59] Speaker 2: we do is give them a handheld device and an
[12:03] Speaker 2: app which is driven by AI. Play with it, have fun,
[12:06] Speaker 2: see how it changes your life. And once you start
[12:09] Speaker 2: liking the product, liking the UIUX, then you start getting
[12:12] Speaker 2: more and the insights also telling the story because once
[12:15] Speaker 2: you start getting the value, I am not having to
[12:17] Speaker 2: pitch my models anymore. The sales reps and the markets
[12:22] Speaker 2: they are pitching on behalf of us. And that's such
[12:25] Speaker 2: a good place to.
[12:26] Speaker 1: Be, is it. It's interesting. So in that instance where
[12:30] Speaker 1: you're designing a more efficient form of interaction and fruitful
[12:34] Speaker 1: for of interaction between sales reps and customers, I could
[12:37] Speaker 1: see a version of that where it is really clear
[12:41] Speaker 1: looking up from a high level that things are working better,
[12:45] Speaker 1: but it might not be clear to the salesperson. Is
[12:48] Speaker 1: the salesperson who's now following the direction of the AI
[12:51] Speaker 1: aware that they are more efficient? They are? How are
[12:55] Speaker 1: they aware them?
[12:56] Speaker 2: Worke they realize few things that they were visiting customers
[12:59] Speaker 2: just because they had to because it was in the schedule.
[13:02] Speaker 2: Now they go there and they find out, wait a minute,
[13:04] Speaker 2: I never tackled this big problem that was not being addressed,
[13:08] Speaker 2: and they solved it. And the customer feedback also comes
[13:11] Speaker 2: back saying we are really happy. So for all these products,
[13:14] Speaker 2: we give the feedback not just from the sales reps,
[13:16] Speaker 2: but also for the customers. Do you really like the
[13:19] Speaker 2: recommendations we're giving you? And that's the best validation you
[13:22] Speaker 2: can think of, because that's for stand feedback.
[13:25] Speaker 1: When the AI is doing this ranking, it wants you
[13:28] Speaker 1: to focus on the customer with the biggest problem first.
[13:32] Speaker 1: Or is it much more complex than that.
[13:34] Speaker 3: It's a little bit more complex than that.
[13:35] Speaker 2: Yeah, but usually it's a rank porderut in terms of
[13:38] Speaker 2: which one is the biggest problem that needs the most
[13:40] Speaker 2: amount of time.
[13:41] Speaker 3: That's how it's ranked order.
[13:43] Speaker 2: But sometimes you can also override the model, right, We
[13:45] Speaker 2: also give options to people. You don't have to all
[13:48] Speaker 2: the time one hundred percent follow recommendation. If you have
[13:51] Speaker 2: some urgent priority, you can overwrite that. That's also possible.
[13:54] Speaker 1: With something like that. Is there a next level you
[13:57] Speaker 1: can go to? So you design the system and you say, oh,
[14:00] Speaker 1: I can make ourselves staff a lot more effective in
[14:03] Speaker 1: the way they operate with their customers. And then you
[14:07] Speaker 1: see that it works, and then it comes back and
[14:10] Speaker 1: then you say, okay, what's two point oh? Is there
[14:13] Speaker 1: a two point oh?
[14:13] Speaker 2: It could be so it's always about innovation. Then you think, okay,
[14:17] Speaker 2: today we go and solve the problems that have already happened.
[14:21] Speaker 2: What if we solve the problems that are likely to happen,
[14:24] Speaker 2: that will be the next step. So this customer hasn't
[14:27] Speaker 2: been very active for a while, there's a high chance
[14:29] Speaker 2: that that customer might churn out of Heineken. So what
[14:33] Speaker 2: actions can I actually recommend to make sure And we
[14:36] Speaker 2: do this, by the way, we also gather a lot
[14:38] Speaker 2: of customer feedback and complaints and feedback, and we use
[14:41] Speaker 2: lllms to extract and glean information. Okay, what are the
[14:44] Speaker 2: real pain points? What's the theme and the topic that
[14:46] Speaker 2: needs to be addressed. And once you do that, then
[14:49] Speaker 2: also you can prepare ahead of time. We're already there,
[14:52] Speaker 2: by the way, when I say two point zero, we're
[14:53] Speaker 2: already testing it. You solve problems or you'll try to
[14:57] Speaker 2: solve problems before da even occur. So I think that's
[15:00] Speaker 2: a little bit of a two point zero. Then we
[15:01] Speaker 2: have to see what else we can do with it.
[15:04] Speaker 1: Tell me, you've had this partnership at Heineken with IBM
[15:09] Speaker 1: for since twenty thirteen, twenty thirteen, so you came in
[15:13] Speaker 1: and there was already a strong working relationship. Tell me
[15:17] Speaker 1: about how that relationship started and what does it mean
[15:19] Speaker 1: on a practical basis. You're building all these tools, how
[15:23] Speaker 1: does the interaction with IBM work?
[15:25] Speaker 2: Yeah, so I think good to give a little bit
[15:27] Speaker 2: of context. Heineken started this digital transformation journey in twenty
[15:31] Speaker 2: twenty formally, but the tech was already there. We had
[15:34] Speaker 2: our systems, platforms, data, everything was there. So all the
[15:38] Speaker 2: IT four ID systems is where IBM was partnering with
[15:42] Speaker 2: us from the get go, from twenty thirteen, and it's
[15:45] Speaker 2: a very long standing partnership because as we found the
[15:47] Speaker 2: tech is evolving, our partnership also kept evolving because we
[15:51] Speaker 2: need to keep up up to the speed. So it
[15:54] Speaker 2: was more about it for ID systems, cybersecurity platform, data,
[15:59] Speaker 2: incident management, service level, you name it.
[16:01] Speaker 3: All of that.
[16:02] Speaker 2: IBM was supporting us both in terms of hands on
[16:05] Speaker 2: and also in.
[16:05] Speaker 3: Terms of strategy to create together.
[16:08] Speaker 2: But that also evolved, like I said, when we went
[16:10] Speaker 2: into this digital transformation journey in twenty twenty. Then we
[16:14] Speaker 2: started building this digital core, which is the central nervous
[16:18] Speaker 2: system software system of Finikin. That's where IBM is really
[16:22] Speaker 2: partnering with us and helping us not just shape the
[16:25] Speaker 2: whole thing in terms of building it hands on, but
[16:27] Speaker 2: how do we strategize that so that it lands well.
[16:29] Speaker 2: So that's yeah, it's a long trusted partnership. I think
[16:33] Speaker 2: we are going to go a long way together.
[16:35] Speaker 3: Yeah.
[16:36] Speaker 1: Heineken Space just sells out of Amsterdam.
[16:38] Speaker 3: The head office is absolutely so.
[16:40] Speaker 1: The IBM people who work with you are they on site?
[16:43] Speaker 2: There are some on site and there are some teams
[16:45] Speaker 2: in India some times spread across the globe. But for
[16:48] Speaker 2: for tech, I think the location doesn't matter. But you
[16:50] Speaker 2: still need people on site to actually talk with the
[16:52] Speaker 2: business and really understand what the problem for.
[16:55] Speaker 3: Those interactions are also very important.
[16:57] Speaker 1: When you said you wanted to when you got there,
[16:59] Speaker 1: you're wanted to build the digital nervous system, what does
[17:02] Speaker 1: that mean?
[17:03] Speaker 2: Maybe good to give you an example, Let's say iPhone, Right,
[17:06] Speaker 2: it's a central platform, but you can download thousands of
[17:11] Speaker 2: apps there and all of them once you download, seamlessly
[17:15] Speaker 2: integrates with the system and you don't see any difference.
[17:17] Speaker 2: This is the same thing. So what we want to
[17:19] Speaker 2: build within Heineken is a central software system, which the
[17:22] Speaker 2: old school way of saying it is the ERP Enterprise
[17:25] Speaker 2: resource planning. It removes the fragmentation of different platforms, it
[17:28] Speaker 2: brings it all together. It makes sure all the business
[17:31] Speaker 2: applications within supply chain, commerce, finance HR all in one
[17:35] Speaker 2: place and coordinates them everything orchestrates them. The benefit of
[17:39] Speaker 2: doing that is two one across the value chaneing of
[17:42] Speaker 2: one way of doing business because everything is standardized, but
[17:46] Speaker 2: we also have multiple markets globally. Across the multiple markets
[17:50] Speaker 2: also it becomes one way of doing business, so it's
[17:53] Speaker 2: both ways. And once you standardize it, we can embed
[17:56] Speaker 2: new apps which will seamlessly integrate. Then it just keep
[18:00] Speaker 2: scaling further and we scale very quick and having that
[18:03] Speaker 2: digital core will really help us scale because the value
[18:06] Speaker 2: from AI and insights is not just building one product
[18:10] Speaker 2: in one place, it's how quickly can you scale it?
[18:13] Speaker 1: And are you still building it or is it an
[18:14] Speaker 1: ongoing thing or it's.
[18:15] Speaker 2: Ongoing thing Because there are nuances in markets. There are
[18:18] Speaker 2: nuances in tax systems and currency systems, so it takes
[18:22] Speaker 2: a little bit of As much as we want to standardize,
[18:25] Speaker 2: you also have to bake in some of the nuances
[18:26] Speaker 2: otherwise people cannot use it. So those sort of outliers
[18:30] Speaker 2: we have to also bake in.
[18:33] Speaker 1: You must learn something when you suddenly, suddenly, when you
[18:36] Speaker 1: standardize a bunch of things that have not been standardized before,
[18:40] Speaker 1: presumably you have a basis for comparisons you couldn't make before.
[18:44] Speaker 2: That's correct. So we also get a lot of external inspiration.
[18:48] Speaker 2: So sometimes these large projects, we don't start by Overholman,
[18:51] Speaker 2: so we get inspiration from partners like IBM or someone else.
[18:55] Speaker 2: How have they done it in somewhere else and where
[18:57] Speaker 2: it's really working. So then you get those ideas, the learnings,
[19:01] Speaker 2: and you start.
[19:02] Speaker 3: Building that way.
[19:03] Speaker 2: And while doing that you figure out that, wait a minute,
[19:05] Speaker 2: we might have don something different and maybe it's even
[19:08] Speaker 2: better than what others have done. Yeah, so it also
[19:11] Speaker 2: creates creativity.
[19:12] Speaker 1: Yeah, I'm just curious whether there was an insight that
[19:16] Speaker 1: you learned from that process That comes to mind A
[19:20] Speaker 1: big one.
[19:21] Speaker 2: I think we don't look at tech for the sake
[19:24] Speaker 2: of tech and embedding it, you know, just just as
[19:27] Speaker 2: one would say it's one single core, one single platform,
[19:30] Speaker 2: everything coordinated. What's the big deal. The big deal is
[19:33] Speaker 2: bringing people along to actually believe that there is a
[19:36] Speaker 2: benefit of doing one way of business, and that actually
[19:39] Speaker 2: means the entire company, not.
[19:40] Speaker 3: Just the leadership team.
[19:42] Speaker 2: So to bring everyone on board and say, tell us
[19:45] Speaker 2: how this platform should look like, what are the components
[19:47] Speaker 2: we should build. It's a pretty big task. Yeah, that's
[19:50] Speaker 2: where the chain management comes in. Yeah, what was hard
[19:53] Speaker 2: about that? Did you have bumbs or we did? I
[19:56] Speaker 2: think it's about convincing people the benefit of doing this.
[20:00] Speaker 2: Why do we say, if you standardize something, we can
[20:02] Speaker 2: go at high speed in scaling. It's not very easy
[20:05] Speaker 2: to visualize it at first. But what you do is
[20:08] Speaker 2: you show some proof of concepts. And that's I won't
[20:10] Speaker 2: call it a trick. This almost bread and butter of
[20:13] Speaker 2: what we do. Show a small proof of concept, show
[20:16] Speaker 2: that it works, show that we can scale, and then
[20:18] Speaker 2: automatically if people start having the faith, and then we'll say, okay,
[20:21] Speaker 2: I see it. Yeah it makes sense.
[20:23] Speaker 1: Sergerty, at least half of what you've talked about is
[20:26] Speaker 1: not about the check itself, but about being a kind
[20:28] Speaker 1: of evangelist for the check. It is half the right percentage.
[20:32] Speaker 1: How much of your time is spent convincing an organization
[20:36] Speaker 1: and people in the organization to see the value in
[20:38] Speaker 1: what you're doing, as opposed to building the thing that
[20:40] Speaker 1: has value.
[20:41] Speaker 2: Yeah, I think that proportion changed over time. When I
[20:43] Speaker 2: first joined, I was very much into the products itself.
[20:46] Speaker 2: I was to review codes myself, let me check what's
[20:48] Speaker 2: going on. And over time, of course, then you focus
[20:51] Speaker 2: on somewhere else. You realize, like I said, best codes,
[20:54] Speaker 2: best models, I use this, if not use the right way,
[20:57] Speaker 2: then I said, okay, now my time is to actually inspire,
[21:00] Speaker 2: show people the value of it. And what I realized
[21:03] Speaker 2: is explaining AI in very simple language really goes a
[21:07] Speaker 2: long way because you take away that that anxiety that
[21:11] Speaker 2: a new product is coming in and we humans.
[21:14] Speaker 3: A little bit have this.
[21:16] Speaker 2: I don't know if it's the right thing to say,
[21:17] Speaker 2: but it's inertia of rest. We like status, poke, we
[21:20] Speaker 2: don't sometimes like change that strupts our. So every time
[21:23] Speaker 2: you build a new product that will change our way
[21:25] Speaker 2: of working. Yeah, there's inherent little bit of anxiety. Yeah,
[21:28] Speaker 2: take that away. Yeah, a job becomes a lot easier.
[21:31] Speaker 1: Are you a good evangelist?
[21:34] Speaker 3: So far it's working?
[21:35] Speaker 2: I think I can do better for sure, because it's
[21:38] Speaker 2: about understanding what's the reason people sometimes might be reluctant
[21:42] Speaker 2: to actually onboard or adopt a new technology. And once
[21:46] Speaker 2: you sort of understand that, then it that anxiety goes away,
[21:50] Speaker 2: it becomes easier. How many people work for httikine about
[21:53] Speaker 2: eighty five thousand, nine thousand globally.
[21:56] Speaker 1: So you have essentially a city pretty much. If you
[22:00] Speaker 1: look at that universe of every five thousand, is there
[22:03] Speaker 1: anyone in that universe who is not touched by what
[22:05] Speaker 1: you're doing?
[22:06] Speaker 2: I think the way we do it is we prioritize
[22:09] Speaker 2: based on the size of the market and the potential opportunity. Yes,
[22:13] Speaker 2: if I had infinite resources, I would go everywhere within
[22:17] Speaker 2: the Heineken company and do everything, But we cannot. We
[22:20] Speaker 2: don't have infinite resources, so we say, let's be a
[22:23] Speaker 2: little bit picky and choosy where the biggest opportunities are.
[22:26] Speaker 3: But it's a matter of time.
[22:27] Speaker 2: Right today we touch upon the big market's biggest scope.
[22:30] Speaker 2: Over time, it's going to be pervasive through the company.
[22:33] Speaker 2: But the appetite is already there. So people are really
[22:36] Speaker 2: even if they have not really embedded some product, they're
[22:39] Speaker 2: asking for it, which is a fantastic place to be.
[22:43] Speaker 3: Yeah.
[22:43] Speaker 1: Yeah, what's been your biggest disappointment so far?
[22:47] Speaker 2: So far, it's been very fulfilling, I must say, but
[22:49] Speaker 2: I think what I would look back. Can we do
[22:52] Speaker 2: things a little bit quicker? Can we go a little bit.
[22:54] Speaker 3: At high speed?
[22:55] Speaker 2: And that's why this whole concept of digital backbone. Can
[22:58] Speaker 2: we standardized everything, We can speed that up. If you
[23:00] Speaker 2: can really scale quickly, I think that'll be the best
[23:04] Speaker 2: because today I have a very good problem.
[23:05] Speaker 3: People are asking for products.
[23:06] Speaker 2: Sometimes I say, yeah, I need to put it on
[23:10] Speaker 2: a timeline and a roadmap because I cannot just cater
[23:12] Speaker 2: to it immediately.
[23:14] Speaker 1: Presumably that's one of the things that the people you're
[23:17] Speaker 1: working with at IBM can tell you. They can give
[23:19] Speaker 1: you a sense of how quickly others have adopted some
[23:23] Speaker 1: of these technology.
[23:24] Speaker 2: That's correct, and that's actually one of the bench funds
[23:26] Speaker 2: that you're referring to. Yeah, we see are we losing
[23:28] Speaker 2: pace and which other things can we go forward? And
[23:31] Speaker 2: in some case, when you look at the digital core
[23:33] Speaker 2: and the backbone, maybe specific areas we can speed up
[23:36] Speaker 2: because those are the areas that are maximum potential value
[23:39] Speaker 2: and some of.
[23:39] Speaker 3: Them we can deprioritize a little bit. Yeah, that we
[23:41] Speaker 3: do all the time. Yeah, just a pragmatic approach to.
[23:45] Speaker 1: I'm curious about a high can specific question, which is,
[23:49] Speaker 1: so here you have a legacy brewer pace in the
[23:54] Speaker 1: Netherlands hundred and sixty years old. If I were to say,
[23:59] Speaker 1: I want you to take an entire new job, I
[24:01] Speaker 1: want you to do what you're doing, but I want
[24:03] Speaker 1: you to do it for an American company in a
[24:07] Speaker 1: completely different industry that's thirty years old, maybe a company
[24:12] Speaker 1: that makes the vacuum cleaners thirty years old in America.
[24:16] Speaker 1: How much of what you're doing. I guess what I'm
[24:18] Speaker 1: trying to say is, are there's things that are particular
[24:20] Speaker 1: to Heineken that have made your job sort of challenging
[24:25] Speaker 1: or interesting or that just wouldn't be an issue in
[24:28] Speaker 1: another environment.
[24:30] Speaker 3: So it's a good question, and thanks for the enticing offer,
[24:33] Speaker 3: but I will polite.
[24:35] Speaker 2: I tried to make it as un average exactly, but
[24:38] Speaker 2: I won't polite to reject offer.
[24:40] Speaker 3: But I'll tell you why I'm reject.
[24:41] Speaker 1: You're going to be in Nebraska, they're making just one
[24:44] Speaker 1: kind of vacuum cleaner.
[24:45] Speaker 2: What I went to school in Iowas, Yes, exactly, so
[24:48] Speaker 2: I'm quite a bit. I think there's there's a cultural difference.
[24:51] Speaker 2: We're all very passionate about our brands and products, and
[24:55] Speaker 2: there's a lot of it is connection based in the
[24:56] Speaker 2: sense we create these connections with our customers sometimes consumers,
[25:00] Speaker 2: and it's all about maintaining that and once you get
[25:03] Speaker 2: a feel of it, you feel part of the family.
[25:05] Speaker 3: That's a very good feeling to have.
[25:07] Speaker 2: And the fact that today where I am, if I
[25:08] Speaker 2: look back, I probably will happy to say very fortunate
[25:12] Speaker 2: to have probably one of the best jobs in the
[25:14] Speaker 2: world in the current times. And there is no end
[25:16] Speaker 2: to innovation, by the way, and even within Heineken, yes
[25:18] Speaker 2: it's a traditional company who is stopping innovation. There's a
[25:21] Speaker 2: lot more do so I'll be very busy for the
[25:24] Speaker 2: next few years.
[25:26] Speaker 1: What are the Dutch like? This is one of the
[25:28] Speaker 1: oldest and most successful commercial cultures in the world, A
[25:32] Speaker 1: tiny country that's been solidly successful us for it. I'm
[25:35] Speaker 1: curious about innovating in that kind of environment. How is
[25:39] Speaker 1: that different from innovating in a huge country like the
[25:42] Speaker 1: United States, or in a different kind of national culture.
[25:45] Speaker 2: I think it's a question of opportunity because within Netherlands.
[25:48] Speaker 2: By the way, Netherlands has one of the most highest
[25:52] Speaker 2: number of startups within Europe, if not the highest, So
[25:56] Speaker 2: there is this culture of innovation that's already embedded in there.
[25:59] Speaker 3: It happened all the time.
[26:00] Speaker 2: Companies like Phillips ASML, some of the very big big
[26:03] Speaker 2: players already there. So it could be a little bit different.
[26:07] Speaker 2: I think in Netherlands we want to make sure what
[26:10] Speaker 2: we are doing really is going to work, so there's
[26:11] Speaker 2: a little bit of discussion alignment. It's more structured, but
[26:15] Speaker 2: also agile in a way we do things, and us
[26:18] Speaker 2: was more like let's do let's go quick, experiment learn fail.
[26:22] Speaker 2: So I think there's pros and cons on both sides,
[26:24] Speaker 2: but so far it's quite good.
[26:27] Speaker 1: Give me a sense of what you're with a day
[26:30] Speaker 1: in a life like for you, what does it look
[26:33] Speaker 1: like to have the job that you have in a
[26:35] Speaker 1: place a Canney.
[26:36] Speaker 2: First of all, it starts with the calendar and the
[26:39] Speaker 2: number of meetings I have, which is usually filled for
[26:42] Speaker 2: forty hours so longer in the week. So that's the
[26:45] Speaker 2: starting point, and then I have to pick and choose
[26:47] Speaker 2: which meetings I need to prepare for what. And usually
[26:49] Speaker 2: these meetings are mostly about where are we with the product,
[26:52] Speaker 2: what are the challenges? How can I help and solve it?
[26:55] Speaker 2: And then sometimes also pitching new products or convincing something,
[26:59] Speaker 2: and also sometimes chain management. I also do sessions where
[27:02] Speaker 2: I present internally, quite often go to different places, because
[27:06] Speaker 2: it always helps to be in front of the audience
[27:08] Speaker 2: when you're presenting something. We also started something recently which
[27:12] Speaker 2: we call AI boot Camp, which is you use Jane
[27:15] Speaker 2: AI as an interface for all these big AI models
[27:18] Speaker 2: and people can interact in a very fun way. That's
[27:22] Speaker 2: our new way of really convincing the rest of the
[27:24] Speaker 2: company that hey.
[27:26] Speaker 3: This is fun to play with and let's go.
[27:29] Speaker 1: So yeah, how many people would you cycle through that
[27:33] Speaker 1: kind of book camp?
[27:34] Speaker 3: Anyone?
[27:34] Speaker 2: Usually we keep it a small group just to make
[27:36] Speaker 2: sure everyone is doing something hands on and nobody's just listening.
[27:41] Speaker 2: So usually a twenty to thirty people max, and then
[27:44] Speaker 2: go from one place to other and it's all hands on.
[27:46] Speaker 3: You cannot sit and watch. You have to participate.
[27:49] Speaker 1: Are you directly involved at all in the design or
[27:51] Speaker 1: creation of any of these tools, so.
[27:54] Speaker 2: I review it and then I used to review also
[27:56] Speaker 2: the codes before and now I'm mostly like trying to
[27:59] Speaker 2: get the feedback from the people that are using it,
[28:02] Speaker 2: because that's my best validation point. If I get high
[28:05] Speaker 2: net promoter score on these products, I know the job
[28:08] Speaker 2: is well done. But I do check accuracy model. Some
[28:10] Speaker 2: of the basic things you'll check in AI. Is the
[28:13] Speaker 2: model drifting over time? What's the accuracy you know? How
[28:15] Speaker 2: is it hosted on a platform? These things I check.
[28:18] Speaker 2: But we also have mechanisms on those, so it's not
[28:20] Speaker 2: like every time you have a dive deep and look
[28:22] Speaker 2: into everything.
[28:23] Speaker 1: Yeah.
[28:23] Speaker 2: So once you have these mechanisms in place, then these
[28:26] Speaker 2: sort of tasks become easier.
[28:28] Speaker 1: You've used the phrase that you want to make honek
[28:30] Speaker 1: in the Best Connected Brewer. What does that phrase mean?
[28:33] Speaker 3: Yeah?
[28:33] Speaker 2: So I think it started with the ambition in twenty
[28:35] Speaker 2: twenty when we said we are going to digital transform.
[28:38] Speaker 2: Remember the pendulum I was talking about from gut page
[28:41] Speaker 2: to all the way to data driven. And in today's world,
[28:44] Speaker 2: when you think of digital transmission, there are few components,
[28:46] Speaker 2: cybersecurity being one of them. The digital core, like I
[28:49] Speaker 2: was saying, is one of them. Simplification and automation of
[28:52] Speaker 2: systems is one of them. Our breweries, how can we simplify?
[28:55] Speaker 2: And then comes data and AI, which is the really
[28:57] Speaker 2: one of the biggest components.
[29:00] Speaker 3: Think or best connected brewer.
[29:01] Speaker 2: The idea is we have been serving our consumers and
[29:04] Speaker 2: customers for one hundred and sixty two years. What's different
[29:08] Speaker 2: if you leverage tech in today's world. I think you
[29:11] Speaker 2: can really enhance the experience the customers have. The example
[29:15] Speaker 2: I was giving you earlier about the salesforce going in
[29:17] Speaker 2: different places and optimizing their out that's a good example.
[29:21] Speaker 2: But the relation is maintained just simply by data driven insights.
[29:24] Speaker 2: So if you can connect all the different applications, all
[29:28] Speaker 2: the platforms, remove fragmentation, scale very quick, make sure your
[29:32] Speaker 2: company is CyberSecure, things are simple and automated. That's what
[29:36] Speaker 2: we call the best connected brewer. That's the ambition.
[29:39] Speaker 1: Actually, how do you measure this success of what you're doing?
[29:43] Speaker 1: Is do you expect that your efforts will have a
[29:49] Speaker 1: measurable and tangible effect on the bottom night of the
[29:51] Speaker 1: company and can you actually figure out what the impact
[29:56] Speaker 1: of your efforts is?
[29:58] Speaker 3: Yeah, we do.
[29:58] Speaker 2: I think it is super important to measure, because the
[30:01] Speaker 2: first when I was referring to proof of value that
[30:04] Speaker 2: when I'm embedding some model, does it really work? So
[30:07] Speaker 2: we do a B testing, which is basically you keep
[30:10] Speaker 2: aside some sample and you actually launch the product on
[30:13] Speaker 2: a different sample, and you see the difference between the two.
[30:15] Speaker 2: The assumption is those that had the product and those
[30:18] Speaker 2: that didn't have the product, both of them went through
[30:21] Speaker 2: the same experiences because of market, seasonal etc.
[30:24] Speaker 3: That's one good way of doing it.
[30:26] Speaker 2: And if you cannot have the luxury sometimes of doing
[30:28] Speaker 2: a B testing because everyone is having high appetite, give
[30:31] Speaker 2: me the product I don't want to sit aside. Then
[30:34] Speaker 2: you do some sort of causal models like we say,
[30:37] Speaker 2: so you kind of look at what would have happened
[30:39] Speaker 2: if the model was not there, and then you predict
[30:42] Speaker 2: that and since the model was there, something else happened.
[30:45] Speaker 2: The difference between the two is the incremental value the
[30:48] Speaker 2: model is creating. A B testing is more accurate the
[30:50] Speaker 2: causal models. The other one, like you said, which called
[30:53] Speaker 2: time series model, a little bit less accurate, but directionally
[30:56] Speaker 2: both give you the sense that yes, it's working.
[30:58] Speaker 1: What happens if you do ABT testing on a new
[31:00] Speaker 1: idea and you don't see a difference.
[31:04] Speaker 2: In that case, we will move on to something else
[31:06] Speaker 2: because it means it's already optimal. Then we said, good
[31:10] Speaker 2: check that. Now let's move on to something else. But
[31:12] Speaker 2: we need to just make sure that the process is
[31:14] Speaker 2: still running optimally. So time to time it. Keep doing
[31:17] Speaker 2: every testing anyway, every six months or whatever the timeframe is.
[31:21] Speaker 3: Yeah, just to make sure that it's still still relevant.
[31:24] Speaker 1: What if this is We're getting out on a little
[31:26] Speaker 1: bit of a digression here, but it's something I've often
[31:28] Speaker 1: thought about. What if the value that is being created
[31:32] Speaker 1: is not measurable? So I'll give you a dumb example.
[31:35] Speaker 1: When you were talking earlier about the salesman and giving
[31:39] Speaker 1: them a new, you know better instructions about how to
[31:43] Speaker 1: basically spend their day, what if you tested that, discovered
[31:47] Speaker 1: it it didn't have any effect on the bottom line.
[31:50] Speaker 1: But in fact, what was happening was that the salesmen
[31:53] Speaker 1: were a lot happier with their jobs. And we're satisfied
[31:57] Speaker 1: and we're excited to come to work. Do you measure
[32:00] Speaker 1: something like that? Something intolerable?
[32:01] Speaker 3: Measure?
[32:02] Speaker 2: One way is npscore, which I said, netromot score. Are
[32:05] Speaker 2: you really happy with the product. Has it changed your life?
[32:08] Speaker 2: That gives you a good indication. And by the way,
[32:11] Speaker 2: it's a numeric output, so it gives you a score
[32:13] Speaker 2: between minus hundred to plus hundred.
[32:17] Speaker 3: And sometimes it's not even tangible.
[32:19] Speaker 2: Let's say we do something for corporate affairs because they
[32:21] Speaker 2: want to get external signals of consumer insights and then
[32:24] Speaker 2: just lean some information. Maybe we act on it, maybe
[32:27] Speaker 2: we don't, but this is for a good cause. Sometimes
[32:30] Speaker 2: you just want to study the market. There's no immediate
[32:32] Speaker 2: value if you don't create a product out of it,
[32:35] Speaker 2: or something to do with legal If there's a reputational
[32:37] Speaker 2: risk for Heineken, can I extract some insight that will
[32:40] Speaker 2: prevent us or create the best briefing or summary or
[32:44] Speaker 2: external briefing that using AI that will help us protect ourselves.
[32:47] Speaker 2: That's also reputational damage.
[32:51] Speaker 1: Last question before we go to questions. I'm curious when
[32:54] Speaker 1: you look at the very beginning you talked about this
[32:58] Speaker 1: linear value chain, where along that chain are you having
[33:04] Speaker 1: the most impact and where are you having the least impact.
[33:06] Speaker 1: I'm more interested in the second half of that.
[33:08] Speaker 2: Yeah, I think we covered few things, but one area
[33:12] Speaker 2: I think we can do more is really understanding consumer sentiments.
[33:17] Speaker 3: And the reason for.
[33:17] Speaker 2: That is Heineken is people go to the bars and
[33:20] Speaker 2: outlets and you're not really leading your first hand data there. Right,
[33:24] Speaker 2: You're enjoying a beer, then you walk away. I don't
[33:27] Speaker 2: know exactly what you did. I can get some aggregated
[33:29] Speaker 2: data to make some sense.
[33:30] Speaker 3: Out of it.
[33:32] Speaker 2: But if we can really get consumer insights as to
[33:35] Speaker 2: what the consumers like and dislike, what sort of ad
[33:37] Speaker 2: you like? How should I design my Hanneken campaign so
[33:40] Speaker 2: it resonates with a cluster of individuals.
[33:44] Speaker 3: That would be a little bit of holy grail.
[33:46] Speaker 2: As the next step, like you were talking about two
[33:47] Speaker 2: point zero, and to get consumer insights first party data,
[33:51] Speaker 2: it's not super easy. So what we are trying to
[33:53] Speaker 2: do is create digital twins of consumer So at an
[33:58] Speaker 2: aggregate level, they give you a sense of Also with
[34:01] Speaker 2: agent KI, which is also you hear a lot about
[34:04] Speaker 2: to get a sense of how consumers might react to
[34:06] Speaker 2: certain campaign or certain product, and that should give us
[34:09] Speaker 2: quite a bit of insights that right now we don't
[34:11] Speaker 2: have access to. I think that's one of the areas
[34:13] Speaker 2: we could really do a lot more.
[34:14] Speaker 1: Yeah, yeah, So if I said lask question if I
[34:17] Speaker 1: sat down with you, it's twenty twenty six. Now we
[34:20] Speaker 1: did this over five years hero now twenty thirty one,
[34:23] Speaker 1: we're sitting in this chair. Tell me what the kind
[34:27] Speaker 1: of what's going to be the next big score.
[34:30] Speaker 2: I think one area will be how we make our
[34:33] Speaker 2: lives as employees of finding a lot easier. So the repetitive,
[34:37] Speaker 2: boring task, manual task. Can we automate those things and
[34:40] Speaker 2: just use the time to do something more creative and
[34:43] Speaker 2: think big about the business itself. That will be one area,
[34:46] Speaker 2: most on the productivity side. But the other area would
[34:48] Speaker 2: be Indeed, when we look at gen Z and this
[34:51] Speaker 2: is fact, I'm not saying something my own opinion, there's
[34:55] Speaker 2: a trend of distinct trend of alcohol as a beverage,
[34:59] Speaker 2: the consumption on a decline. So then what's the next
[35:02] Speaker 2: best thing for the new generation consumers?
[35:05] Speaker 3: What will resonate?
[35:07] Speaker 2: Those are the pockets we need to find and I
[35:09] Speaker 2: think that's where we will transition very quickly over the
[35:11] Speaker 2: next five years.
[35:12] Speaker 3: And if you get there, I think that will be
[35:13] Speaker 3: big success.
[35:14] Speaker 1: Do you think that your specific department responsibility can help
[35:21] Speaker 1: the company in discovering what the answer to that question.
[35:23] Speaker 2: Is about definitely. That's the ambition. That's what we're trying
[35:26] Speaker 2: to do. That's what we are really trying to get
[35:28] Speaker 2: this consumer insights. I think that's the last mile. That's
[35:30] Speaker 2: the one part that is left.
[35:33] Speaker 1: Sergery. This has been fascinating. Thank you so much. I
[35:35] Speaker 1: should say thank you for a question, sir. My uncle
[35:38] Speaker 1: was a Heineken salesman in Jamaica. He was the local
[35:41] Speaker 1: distributor and I have so many child memories of going
[35:45] Speaker 1: to Jamaica and he would show up in his Heineken truck.
[35:48] Speaker 1: So we're resonating deep in my inmorary circle with this conversation.
[35:55] Speaker 1: He would come and he would have a Heineken right
[35:57] Speaker 1: there on the table and would drink it at.
[35:59] Speaker 3: The end of the day.
[36:00] Speaker 1: But we have we have a few moments for questions.
[36:04] Speaker 1: They're all on the screen and I don't have my glasses.
[36:06] Speaker 1: Can you read them?
[36:07] Speaker 3: Yeah, I can read them. Then they should be going
[36:09] Speaker 3: order to the first one.
[36:10] Speaker 1: Yeah no no, no, no no no no, rookie air,
[36:13] Speaker 1: never do that. Okay, read the first four and pick
[36:17] Speaker 1: the one you want to answer.
[36:19] Speaker 3: Okay, go tip.
[36:21] Speaker 2: But I gave it away already, so I'm going to
[36:23] Speaker 2: now do what I said no, I think the first
[36:25] Speaker 2: one is quite relevant.
[36:27] Speaker 3: So it's a question for both of us.
[36:29] Speaker 2: If you were advising a twenty year old, what three
[36:32] Speaker 2: skills would you tell them to start developing right now
[36:35] Speaker 2: to stay relevant in an AI driven world?
[36:39] Speaker 1: Oh well, well you don't you have a twenty year
[36:42] Speaker 1: old or a near twenty year old. You have a
[36:44] Speaker 1: fifteen year old. You told me I have a fifteen
[36:45] Speaker 1: year old. All right, what do you tell your son?
[36:47] Speaker 2: I thought you were going to answer them this first, but.
[36:50] Speaker 1: My kids are two and four. I tell them to
[36:52] Speaker 1: put away their toys. You this is more you start.
[36:57] Speaker 1: You have a is your fifteen year old center.
[37:00] Speaker 2: His sun and he is already tinkering with AI, is
[37:03] Speaker 2: doing his own Python coding, etc. Which I couldn't imagine
[37:06] Speaker 2: when I was fifteen. So I think I'll give a
[37:09] Speaker 2: high level answer. To be actually successful. Depending on whether
[37:12] Speaker 2: your hands on within AI, building models yourself or not,
[37:16] Speaker 2: there are three things I think super important. One is
[37:20] Speaker 2: having that tech background, having a common understanding of.
[37:23] Speaker 3: What AI really is. It always helps.
[37:25] Speaker 2: Not everyone needs to have the details and algorithms and
[37:28] Speaker 2: how models work not needed, but having that basic understanding
[37:31] Speaker 2: always is good. Then you know exactly how to gauge
[37:34] Speaker 2: what AI is really doing. And I think the other
[37:36] Speaker 2: thing is if you are in a corporate setting and
[37:39] Speaker 2: you are doing something for the business, work backwards from
[37:42] Speaker 2: the business and understand whatever you're building should actually touch
[37:46] Speaker 2: the business and make it beneficial for them. It's not
[37:49] Speaker 2: AI and modeling for the sake of it. That's for
[37:51] Speaker 2: a separate research and development. If you're in a corporate world,
[37:54] Speaker 2: try to build something beneficial for business. And I third
[37:57] Speaker 2: one I think which myself I learned quite a bit
[37:59] Speaker 2: in my in last six years.
[38:01] Speaker 3: It's communication. Talking about AI.
[38:04] Speaker 2: If you used a lot of tech jargon and mathematics,
[38:08] Speaker 2: sometimes people lose you. It's about how you really narrate
[38:11] Speaker 2: the story in a very simple way so people can
[38:14] Speaker 2: relate to it. I think if a combination of these
[38:16] Speaker 2: three has worked very well for me, so I can
[38:19] Speaker 2: say that anything you want to add.
[38:22] Speaker 1: It's funny because I I met this guy who's the
[38:26] Speaker 1: headmaster at a Jesuit school in Manhattan. We've been chatting
[38:30] Speaker 1: and I want to do a little program at his school,
[38:33] Speaker 1: and it's all about asking questions because we're now into
[38:38] Speaker 1: the era of asking questions, right, That's correct. AI is
[38:41] Speaker 1: this incredibly good tool, but you have to answer the
[38:44] Speaker 1: right questions. But this is not just true of AI,
[38:47] Speaker 1: but it's also true of the world we're living in.
[38:50] Speaker 1: Is a world that's so interconnected and everything involves so
[38:54] Speaker 1: many different people that your distinguishing feature in many context
[38:59] Speaker 1: is not where they're not the answers you have, but
[39:01] Speaker 1: the quality of the questions that you ask.
[39:03] Speaker 3: That's fantastic what you said. I fully agree.
[39:05] Speaker 2: I think it's about asking the right questions that really
[39:08] Speaker 2: tells you, you know, looking for depth, that unique thing
[39:11] Speaker 2: that that you're missing.
[39:12] Speaker 3: Yeah, I full agree.
[39:14] Speaker 1: Maybe I'll advite you to this class and like you
[39:18] Speaker 1: have that kind of time on your hands if you
[39:22] Speaker 1: brought if you brought up you know, Heineken for all
[39:24] Speaker 1: the kids in school, that would that would really.
[39:27] Speaker 3: Yeah, surely we have to build a special product for that.
[39:31] Speaker 3: Well let's see, all.
[39:32] Speaker 2: Right, next question, let's go to this one or Malcolm,
[39:36] Speaker 2: is there a particular AI capability you are each excited
[39:40] Speaker 2: to explore?
[39:42] Speaker 3: That's for you, my friend.
[39:44] Speaker 2: I think in the in the short term, I'm really
[39:47] Speaker 2: looking forward to Agent KI is hearing a lot of
[39:49] Speaker 2: noise and the hype and there are a lot of
[39:53] Speaker 2: feedback that I'm getting from a lot of companies. Have
[39:55] Speaker 2: you really embedded agent ki within your systems? There is
[39:59] Speaker 2: a very mixed feed some say yes, some say no.
[40:01] Speaker 2: I think the potential of agent ki when we look
[40:03] Speaker 2: at this task we do day to day. Let me
[40:07] Speaker 2: gives some example, invoice management or transactional finance or very
[40:11] Speaker 2: repetitive task. If you can really automate augment those things
[40:13] Speaker 2: with agent KI, I think it's going to be a
[40:15] Speaker 2: game changer. If you free up thirty percent of our
[40:18] Speaker 2: time just by embedding these things, then I can really
[40:22] Speaker 2: think big. Everyone can think big. What's next? Then the
[40:25] Speaker 2: creativity comes in. Otherwise all day you are stuck with
[40:28] Speaker 2: the repetitive task. So I think that's what I'm really
[40:31] Speaker 2: looking forward to. And this is very short term. Within
[40:33] Speaker 2: the next few years.
[40:34] Speaker 1: Yeah, we have I think, what time for one more question,
[40:39] Speaker 1: Sergey go for Let's see this is it's going to
[40:43] Speaker 1: be the last one always has to be the best one.
[40:45] Speaker 2: Certain Let for people hearing the phrase for the first time,
[40:50] Speaker 2: what is the real example that shows Heineken being the
[40:53] Speaker 2: best connected broer. Basically you're asking for a proof point.
[40:57] Speaker 2: Are we really becoming the best connected brower. When we
[41:00] Speaker 2: look at our markets, Heineken Mexico is a very good
[41:03] Speaker 2: example across our value chain. If you work backwards from consumers,
[41:07] Speaker 2: customers and so on. We have advertising optimization for consumers.
[41:12] Speaker 2: For customers, we have next best action. Actually for customers,
[41:16] Speaker 2: we have pricing and promotion optimized. For the salesforce, we
[41:19] Speaker 2: have next best action. For the breweries, we have connected brewery.
[41:22] Speaker 2: We are getting signals from these machines and optimizing them.
[41:25] Speaker 2: I think it covers a significant portion of a value
[41:28] Speaker 2: chain that's fully automated end to end. So that would
[41:31] Speaker 2: be a good example where we really saw the benefit
[41:34] Speaker 2: of taking it to the next level when it comes
[41:36] Speaker 2: to automation. So Mexico, Heineken Mexico is a good example.
[41:40] Speaker 1: Thank you so much for joining us. Thank you to
[41:42] Speaker 1: all of you who came to listen.
[41:45] Speaker 3: Tank for It Expert, thank for very much. Yeah.
[41:50] Speaker 1: That's it for the first episode of season seven of
[41:53] Speaker 1: Smart Talks with IBM, But stay tuned. There's so much
[41:56] Speaker 1: more to come this season as we die further into
[41:59] Speaker 1: how AI and quantum Computing are creating smarter business. Smart
[42:05] Speaker 1: Talks with IBM is produced by Matt Ramano, Amy Gains
[42:08] Speaker 1: McQuaid and Jake Harper. Engineering by Ninabird Lawrence, Mastering by
[42:13] Speaker 1: Sarah Buger, music by Gramoscope, Strategy by Cassidy Meyer and
[42:18] Speaker 1: Sophia Derlin. Special thanks to Sergei Ghosh and Michelle Ganji.
[42:22] Speaker 1: Post from the Heineken Company. Smart Talks with IBM is
[42:26] Speaker 1: a production of Pushkin Industries and Ruby Studio at iHeartMedia.
[42:30] Speaker 1: To find more Pushkin podcasts, listen on the iHeartRadio app,
[42:35] Speaker 1: Apple Podcasts, or wherever you listen to podcasts. I'm Malcolm Gladwell.
[42:40] Speaker 1: This is a paid advertisement from IBM. The conversations on
[42:44] Speaker 1: this podcast don't necessarily represent IBM's positions, strategies, or opinions.

Shows are listed from Apple Podcasts' charts and directory, and Podcast Index where it is used. Every show and episode belongs to its publisher and plays from the publisher's own feed; this site hosts none of them.