1
00:00:00,160 --> 00:00:02,960
Speaker 1: Hey everyone, it's Robert and Joe here. Today we've got

2
00:00:02,960 --> 00:00:05,240
Speaker 1: something a little bit different to share with you. It's

3
00:00:05,280 --> 00:00:09,400
Speaker 1: a new season of the Smart Talks with IBM podcast series.

4
00:00:09,720 --> 00:00:13,520
Speaker 2: This season on Smart Talks with IBM, Malcolm Gladwell is back,

5
00:00:13,520 --> 00:00:15,760
Speaker 2: and this time he's taking the show on the road.

6
00:00:15,880 --> 00:00:19,160
Speaker 2: Malcolm is stepping outside the studio to explore how IBM

7
00:00:19,239 --> 00:00:22,880
Speaker 2: clients are using artificial intelligence to solve real world challenges

8
00:00:23,079 --> 00:00:25,080
Speaker 2: and transform the way they do business.

9
00:00:25,480 --> 00:00:29,560
Speaker 1: From accelerating scientific breakthroughs to reimagining education. It's a fresh

10
00:00:29,600 --> 00:00:33,040
Speaker 1: look at innovation in action, where big ideas meet cutting

11
00:00:33,120 --> 00:00:33,960
Speaker 1: edge solutions.

12
00:00:34,320 --> 00:00:37,160
Speaker 2: You'll hear from industry leaders, creative thinkers, and of course

13
00:00:37,280 --> 00:00:40,920
Speaker 2: Malcolm Gladwell himself as he guides you through each story.

14
00:00:41,440 --> 00:00:44,680
Speaker 1: New episodes of Smart Talks with IBM drop every month

15
00:00:44,680 --> 00:00:48,279
Speaker 1: on the iHeartRadio app, Apple Podcasts, or wherever you get

16
00:00:48,320 --> 00:00:52,240
Speaker 1: your podcasts. Learn more at IBM dot com slash smart Talks.

17
00:00:52,880 --> 00:00:55,360
Speaker 1: This is a paid advertisement from IBM.

18
00:00:57,640 --> 00:01:01,000
Speaker 3: H'm Malcolm Gladwell. Welcome to Season seven of Smart Talks

19
00:01:01,000 --> 00:01:04,680
Speaker 3: with IBM. This year, we're exploring new stories about how

20
00:01:04,720 --> 00:01:08,480
Speaker 3: companies are using the latest advancements in AI and quantum

21
00:01:08,520 --> 00:01:12,520
Speaker 3: computing to create smarter business. For the first episode of

22
00:01:12,560 --> 00:01:16,160
Speaker 3: the season, I flew to Austin, Texas to join Sergei

23
00:01:16,280 --> 00:01:20,800
Speaker 3: Ghosh on stage at south By Southwest. Sergei is Chief

24
00:01:20,840 --> 00:01:25,600
Speaker 3: AI Officer at Heineken, the world's pioneering beer company. Founded

25
00:01:25,640 --> 00:01:28,880
Speaker 3: in eighteen sixty four. Heineken is deep roots, but it

26
00:01:28,920 --> 00:01:33,319
Speaker 3: continues to push the boundaries of innovation today. In twenty twenty,

27
00:01:33,319 --> 00:01:35,760
Speaker 3: the company came up with a goal to become the

28
00:01:35,800 --> 00:01:39,640
Speaker 3: world's best connected brewer. Surge plays a key role in

29
00:01:39,720 --> 00:01:42,720
Speaker 3: leading that transformation, and I sat down with him in

30
00:01:42,760 --> 00:01:45,640
Speaker 3: front of a live audience to understand what that journey

31
00:01:45,640 --> 00:01:48,760
Speaker 3: looks like and what it takes to reinvent a global

32
00:01:48,760 --> 00:01:55,160
Speaker 3: company from the inside out. And before we get to

33
00:01:55,280 --> 00:01:57,960
Speaker 3: the question of what you do in your job. So

34
00:01:58,120 --> 00:02:00,440
Speaker 3: I'm really interested in people who have jobs it didn't

35
00:02:00,440 --> 00:02:03,160
Speaker 3: exist for most of their life, and I'm curious how

36
00:02:03,200 --> 00:02:03,760
Speaker 3: you got there.

37
00:02:03,960 --> 00:02:06,360
Speaker 4: Yeah, first of all, thanks for having me here. Yeah,

38
00:02:06,640 --> 00:02:09,959
Speaker 4: actually it did exist, and people sometimes don't realize AI

39
00:02:10,080 --> 00:02:13,640
Speaker 4: is not new. It's been there for seventy five years,

40
00:02:13,680 --> 00:02:14,680
Speaker 4: since nineteen fifty.

41
00:02:15,680 --> 00:02:16,919
Speaker 5: It just changed over time.

42
00:02:16,960 --> 00:02:19,720
Speaker 4: How the application is happening, right, So one thing to

43
00:02:19,800 --> 00:02:22,240
Speaker 4: keep up with is as AI became more popular and

44
00:02:22,280 --> 00:02:25,560
Speaker 4: more embedded in business, how do we upscal ourselves to

45
00:02:25,600 --> 00:02:28,960
Speaker 4: stay at par with the technology trends. So the preparation

46
00:02:29,080 --> 00:02:31,280
Speaker 4: for me personally started actually a long time ago, so

47
00:02:31,280 --> 00:02:33,359
Speaker 4: when I was in grad school in US, and I

48
00:02:33,400 --> 00:02:34,800
Speaker 4: also to live in US by the way, for a

49
00:02:34,800 --> 00:02:35,280
Speaker 4: long time.

50
00:02:35,560 --> 00:02:36,400
Speaker 3: You're Indian.

51
00:02:37,040 --> 00:02:39,200
Speaker 4: I'm an Indian originally, but it's news. I did my

52
00:02:39,200 --> 00:02:42,800
Speaker 4: grad school here and there. Actually I started taking courses

53
00:02:42,800 --> 00:02:46,160
Speaker 4: in newer networks and at fission diagents back in two

54
00:02:46,200 --> 00:02:48,880
Speaker 4: thousand and two, and it wasn't popular back then.

55
00:02:49,200 --> 00:02:51,040
Speaker 5: I was just curious, what is it? Maybe it's the

56
00:02:51,080 --> 00:02:51,720
Speaker 5: next big thing.

57
00:02:52,200 --> 00:02:54,560
Speaker 4: And I'm so glad I did that because that sort

58
00:02:54,600 --> 00:02:55,960
Speaker 4: of helped me build that foundation.

59
00:02:56,400 --> 00:02:59,519
Speaker 3: What was it you said you were? You were curious

60
00:02:59,520 --> 00:03:02,160
Speaker 3: about it? You're curious about it? Why what caught your

61
00:03:02,200 --> 00:03:03,200
Speaker 3: eye about it?

62
00:03:03,240 --> 00:03:06,760
Speaker 4: Was very different because the main difference was before that

63
00:03:06,840 --> 00:03:08,280
Speaker 4: I was an engineer by professions.

64
00:03:08,480 --> 00:03:09,760
Speaker 5: I went to engineering college.

65
00:03:10,200 --> 00:03:12,919
Speaker 4: Everything is rule based, Everything is based on a formula,

66
00:03:13,000 --> 00:03:16,880
Speaker 4: a physical equation. AI is something different because based on

67
00:03:17,000 --> 00:03:19,920
Speaker 4: data and statistics, it never gives you a clear answer.

68
00:03:20,000 --> 00:03:22,440
Speaker 4: It gives you a probability, and I just thought this

69
00:03:22,480 --> 00:03:25,639
Speaker 4: is very interesting because if you're trying to solve a problem,

70
00:03:25,919 --> 00:03:27,920
Speaker 4: you don't know exactly how to solve it. There is

71
00:03:27,960 --> 00:03:30,720
Speaker 4: no equation. How do you get around that? I think

72
00:03:30,720 --> 00:03:33,760
Speaker 4: that's where AI comes in. It finds those patterns within

73
00:03:33,880 --> 00:03:37,360
Speaker 4: data and comes up with some prediction that intrigued me.

74
00:03:37,800 --> 00:03:41,800
Speaker 4: So this is what year that you start I started dabbling,

75
00:03:42,040 --> 00:03:44,080
Speaker 4: let's call it dabbling in AI was two thousand and two.

76
00:03:44,400 --> 00:03:45,960
Speaker 5: It was almost twenty four years ago.

77
00:03:45,800 --> 00:03:49,120
Speaker 3: Four years so put what you were playing with in

78
00:03:49,160 --> 00:03:52,960
Speaker 3: two thousand and two was an extremely primitive version of what.

79
00:03:52,920 --> 00:03:53,560
Speaker 5: We have now.

80
00:03:53,920 --> 00:03:56,360
Speaker 4: I think it was very relevant because the way I

81
00:03:56,440 --> 00:04:00,080
Speaker 4: see it, should I have skipped all the foundations that

82
00:04:00,120 --> 00:04:02,040
Speaker 4: I learned over the years and just gone to the

83
00:04:02,040 --> 00:04:05,880
Speaker 4: current state. Maybe, But when I look back, I think

84
00:04:05,880 --> 00:04:09,440
Speaker 4: that foundation really helped me because back then, and surprisingly,

85
00:04:09,480 --> 00:04:11,560
Speaker 4: by the way, new on networks. When I talk about that,

86
00:04:12,040 --> 00:04:15,360
Speaker 4: it's still very ralliant and relevant within AI. The entire

87
00:04:15,400 --> 00:04:18,719
Speaker 4: foundation is new on networks. So I think that foundation

88
00:04:18,839 --> 00:04:21,080
Speaker 4: really helped. Yeah, and I still find it very relevant

89
00:04:21,120 --> 00:04:22,120
Speaker 4: and I apply it day to day.

90
00:04:22,400 --> 00:04:26,120
Speaker 3: Yeah, imagine in having a conversation with you twenty years

91
00:04:26,120 --> 00:04:29,280
Speaker 3: ago and I say what are you up to? And

92
00:04:29,320 --> 00:04:33,599
Speaker 3: you say, I'm playing with this thing neural networks. Early version,

93
00:04:33,640 --> 00:04:35,680
Speaker 3: would you have used the term artificial intelligence?

94
00:04:37,080 --> 00:04:37,599
Speaker 5: Probably not.

95
00:04:37,800 --> 00:04:40,520
Speaker 4: I probably would have used something is called statistics, which

96
00:04:40,560 --> 00:04:43,600
Speaker 4: is everyone is aware of. Back then, it was more statistical,

97
00:04:43,680 --> 00:04:46,520
Speaker 4: so you don't have these big algorithms at that point.

98
00:04:47,400 --> 00:04:49,200
Speaker 4: But then something happened. I don't know if you heard

99
00:04:49,200 --> 00:04:52,000
Speaker 4: of this company called Cago. They used to host these

100
00:04:52,000 --> 00:04:55,320
Speaker 4: sort of data science competitions and anyone can participate and

101
00:04:55,360 --> 00:04:58,039
Speaker 4: if you do really well, you get a price. That

102
00:04:58,200 --> 00:05:00,599
Speaker 4: was a good motivation, just to see, okay, learned something,

103
00:05:01,080 --> 00:05:02,680
Speaker 4: let me apply it and see how good I am

104
00:05:02,880 --> 00:05:04,880
Speaker 4: I'm getting at it. So I think that was my

105
00:05:04,960 --> 00:05:07,440
Speaker 4: first entry point where I really got hands on into AI,

106
00:05:08,120 --> 00:05:10,240
Speaker 4: and that probably stayed with me for a while. I

107
00:05:10,240 --> 00:05:12,800
Speaker 4: think that was back in two thousand and six. That's

108
00:05:12,800 --> 00:05:15,120
Speaker 4: what I started getting hands on. And the funny thing

109
00:05:15,200 --> 00:05:17,600
Speaker 4: is when you look at these Caggle competitions, the use

110
00:05:17,640 --> 00:05:21,240
Speaker 4: cases they used to give actual industry applications, so you

111
00:05:21,240 --> 00:05:24,400
Speaker 4: were really dealing with business problems applying AI to solve it,

112
00:05:25,000 --> 00:05:27,280
Speaker 4: and then you know, wait a minute, a medical company

113
00:05:27,320 --> 00:05:29,719
Speaker 4: is using it. A manufacturing company is using it, a

114
00:05:29,800 --> 00:05:32,480
Speaker 4: banking is using it. And this is two thousand and six,

115
00:05:33,040 --> 00:05:36,080
Speaker 4: so it already started and then it just yeah today,

116
00:05:36,120 --> 00:05:38,040
Speaker 4: it's a different ballgame now.

117
00:05:38,080 --> 00:05:40,840
Speaker 3: So you you came to Heineken.

118
00:05:40,920 --> 00:05:45,560
Speaker 4: When twenty twenty, twenty twenty right middle of COVID, right.

119
00:05:46,120 --> 00:05:48,880
Speaker 3: And where you brought in to be the chief AI officer?

120
00:05:49,040 --> 00:05:51,600
Speaker 4: Was you explore it and the title changed? But yes

121
00:05:51,640 --> 00:05:52,520
Speaker 4: I was the global leader?

122
00:05:52,600 --> 00:05:55,520
Speaker 3: Yeah yeah, And what made you want to take the job?

123
00:05:56,360 --> 00:05:59,159
Speaker 4: I was actually working for Amazon at that point, but

124
00:05:59,200 --> 00:06:01,200
Speaker 4: when I looked at High and I thought, okay, this

125
00:06:01,360 --> 00:06:05,479
Speaker 4: is a legacy, traditional company, right, and AI was not

126
00:06:05,720 --> 00:06:07,520
Speaker 4: a capability embedded.

127
00:06:07,040 --> 00:06:07,880
Speaker 5: At that point of time.

128
00:06:08,360 --> 00:06:10,719
Speaker 4: So it's a great opportunity if I can start something

129
00:06:10,720 --> 00:06:15,240
Speaker 4: from scratch, really build it across the entire valuation of Heineken.

130
00:06:15,760 --> 00:06:17,960
Speaker 4: I mean, that's probably the best job anyone can even

131
00:06:18,000 --> 00:06:20,760
Speaker 4: ask for. Yes, it's of course a lot of responsibility

132
00:06:21,320 --> 00:06:23,040
Speaker 4: that you have to make sure that you really build

133
00:06:23,080 --> 00:06:26,000
Speaker 4: the right products and right capability, But that also happened,

134
00:06:26,480 --> 00:06:28,599
Speaker 4: so I look back, it's quite fulfilling.

135
00:06:29,200 --> 00:06:32,040
Speaker 3: But it's also if I might playable was advocate for

136
00:06:32,120 --> 00:06:34,960
Speaker 3: a moment, you're also taking a risk going into an

137
00:06:35,080 --> 00:06:38,800
Speaker 3: established how long has Heineken been around one hundred and

138
00:06:38,800 --> 00:06:41,480
Speaker 3: sixty two years to be specific, eighteen sixty four. It's

139
00:06:41,520 --> 00:06:43,960
Speaker 3: a very different proposition walking into one hundred and sixty

140
00:06:44,040 --> 00:06:46,240
Speaker 3: year old company and saying I want to bring the

141
00:06:46,279 --> 00:06:48,400
Speaker 3: future to the way you operate. Then it is with

142
00:06:48,480 --> 00:06:49,440
Speaker 3: the startup.

143
00:06:49,960 --> 00:06:51,800
Speaker 5: That is true, but it's also a challenge.

144
00:06:51,839 --> 00:06:54,599
Speaker 4: It's a good challenge, and also that Heineken is also

145
00:06:54,680 --> 00:06:57,080
Speaker 4: looking externally. There are companies that are picking up speed

146
00:06:57,120 --> 00:07:00,600
Speaker 4: and embedding and adopting AI, so should be all behind

147
00:07:00,760 --> 00:07:02,800
Speaker 4: not really, so we also need to pick it up.

148
00:07:03,120 --> 00:07:05,680
Speaker 4: So I thought it was a good challenge because the

149
00:07:05,800 --> 00:07:08,560
Speaker 4: use cases were there, the opportunity I put sense the

150
00:07:08,600 --> 00:07:11,040
Speaker 4: business really was having the appetite.

151
00:07:11,120 --> 00:07:12,160
Speaker 5: Let's do something different.

152
00:07:12,760 --> 00:07:15,040
Speaker 4: When we apply AI in a corporate setting like this,

153
00:07:15,480 --> 00:07:18,080
Speaker 4: it's super important to understand how the business actually works.

154
00:07:18,400 --> 00:07:20,680
Speaker 4: What's the value chain looking like, what are the nuances?

155
00:07:20,680 --> 00:07:22,760
Speaker 4: Where can I And once you get an understanding. It

156
00:07:22,800 --> 00:07:24,720
Speaker 4: took me some time, by the way, to understand the

157
00:07:24,760 --> 00:07:27,760
Speaker 4: full business and the complexities, but once you cross that

158
00:07:27,800 --> 00:07:31,040
Speaker 4: threshold you figure out what's feasible what's.

159
00:07:30,840 --> 00:07:32,960
Speaker 5: Not, then it opens up.

160
00:07:33,080 --> 00:07:35,440
Speaker 4: Wait a minute, within the valuetion, I see ten areas

161
00:07:35,600 --> 00:07:36,520
Speaker 4: I can optimize.

162
00:07:36,600 --> 00:07:38,400
Speaker 3: You say, once you want to see in the business,

163
00:07:38,400 --> 00:07:40,920
Speaker 3: describe the business. What the Heineken puzzle?

164
00:07:41,160 --> 00:07:43,920
Speaker 5: So well, puzzle, Let's see if it's puzzle.

165
00:07:43,960 --> 00:07:46,720
Speaker 4: After I explained it's actually you start with the procurement

166
00:07:46,760 --> 00:07:49,520
Speaker 4: where you get the glasses, cans and all the raw materials.

167
00:07:49,880 --> 00:07:52,000
Speaker 4: Then it comes to the brewery where the magic happens.

168
00:07:52,000 --> 00:07:54,920
Speaker 4: That's what behinneqn beer is produced. Then it goes to

169
00:07:55,000 --> 00:07:58,560
Speaker 4: the distributors. Basically a supply chain takes over. Then it

170
00:07:58,600 --> 00:08:01,280
Speaker 4: goes to the customers. What we refer to as customers

171
00:08:01,320 --> 00:08:06,000
Speaker 4: are the bars, restaurants, retail stores, moment pop stores, convenience stores,

172
00:08:06,560 --> 00:08:10,080
Speaker 4: and that's where actually consumers then come and actually consume

173
00:08:10,120 --> 00:08:10,560
Speaker 4: the product.

174
00:08:11,000 --> 00:08:12,320
Speaker 5: So that's actually the value chain.

175
00:08:12,400 --> 00:08:14,720
Speaker 4: It's actually pretty linear when you think of it, but

176
00:08:14,800 --> 00:08:17,760
Speaker 4: there are nuances depending on the country and the market.

177
00:08:18,160 --> 00:08:20,120
Speaker 4: There are some specific rules and guard rails that you

178
00:08:20,160 --> 00:08:20,760
Speaker 4: have to be aware of.

179
00:08:21,080 --> 00:08:24,400
Speaker 3: So you have that process going on all around the

180
00:08:24,400 --> 00:08:27,600
Speaker 3: world and across multiple brands.

181
00:08:27,480 --> 00:08:30,560
Speaker 4: Multiple brands, multiple countries, multiple operating companies.

182
00:08:30,680 --> 00:08:34,440
Speaker 3: Yeah, from your perspective as someone who is the chief

183
00:08:34,480 --> 00:08:37,440
Speaker 3: AI officer, what are the tasks in front of you?

184
00:08:37,880 --> 00:08:38,840
Speaker 3: What's your opportunity?

185
00:08:38,880 --> 00:08:42,760
Speaker 4: There any process that you think that is maybe not

186
00:08:43,040 --> 00:08:46,400
Speaker 4: digitized or maybe not data driven, you can optimize. I

187
00:08:46,440 --> 00:08:49,480
Speaker 4: look at it like a pendulum. So one side of

188
00:08:49,520 --> 00:08:52,880
Speaker 4: the pendulum you have complete gut based decision making. The

189
00:08:52,920 --> 00:08:55,920
Speaker 4: other extreme is completely data driven. So the idea is,

190
00:08:56,040 --> 00:08:58,280
Speaker 4: can we swing this pendulum it will be towards data

191
00:08:58,360 --> 00:08:59,319
Speaker 4: driven from where we are.

192
00:09:00,200 --> 00:09:03,360
Speaker 3: Give me a specific example of a problem you set

193
00:09:03,400 --> 00:09:05,520
Speaker 3: out solve or address.

194
00:09:06,559 --> 00:09:08,080
Speaker 4: There are quite a few, but if I want to

195
00:09:08,559 --> 00:09:11,560
Speaker 4: pick one of the most fun one fun one may be.

196
00:09:11,440 --> 00:09:14,240
Speaker 5: The most most complex one. Let's bring that one. I think.

197
00:09:14,240 --> 00:09:17,320
Speaker 4: So we spend quite a bit on advertising, and Heineken

198
00:09:17,559 --> 00:09:20,560
Speaker 4: is a largely a lot of it marketing company, and

199
00:09:20,640 --> 00:09:22,600
Speaker 4: we really care about our brands.

200
00:09:22,280 --> 00:09:25,400
Speaker 5: And products, you know, we're almost obsessed with it. Let's

201
00:09:25,440 --> 00:09:26,000
Speaker 5: take an example.

202
00:09:26,080 --> 00:09:28,600
Speaker 4: Let's say you have X million dollars as your budget

203
00:09:29,200 --> 00:09:32,280
Speaker 4: and you have two brands, let's say Heineken and Tosa Kiss.

204
00:09:32,520 --> 00:09:34,760
Speaker 4: I think the crowd audienceale will be familiar with that.

205
00:09:35,600 --> 00:09:38,959
Speaker 4: And then you have three touch points TV, YouTube, Instagram,

206
00:09:39,360 --> 00:09:43,840
Speaker 4: and I want to optimize my advertising budget between different

207
00:09:43,880 --> 00:09:46,960
Speaker 4: brand and touch points, so heiniken on Instagram.

208
00:09:46,960 --> 00:09:49,640
Speaker 5: How much should I spend? It's a very easy question to.

209
00:09:49,600 --> 00:09:52,320
Speaker 4: Ask, but to actually solve this you have to study

210
00:09:52,440 --> 00:09:55,600
Speaker 4: historically how these performed and then create a model and

211
00:09:55,640 --> 00:09:59,080
Speaker 4: then predict if I allocate my budget in this way

212
00:09:59,559 --> 00:10:02,440
Speaker 4: that's probab more optimal. Before it was more like somebody

213
00:10:02,440 --> 00:10:05,319
Speaker 4: to con gun based decisions saying okay, here goes xpillion,

214
00:10:05,360 --> 00:10:07,560
Speaker 4: here goes five million, and we say no, no, no, that's

215
00:10:07,559 --> 00:10:08,520
Speaker 4: not the right proportion.

216
00:10:08,800 --> 00:10:11,959
Speaker 3: What did the AI tell you about the accuracy of

217
00:10:12,000 --> 00:10:14,240
Speaker 3: those spending decisions?

218
00:10:14,679 --> 00:10:17,319
Speaker 4: In the past, we looked at the return on investment

219
00:10:17,360 --> 00:10:20,200
Speaker 4: from this advertising, so how much incremental volume or volume

220
00:10:20,240 --> 00:10:23,520
Speaker 4: of brre reselling or revenue are recreating and we find

221
00:10:23,520 --> 00:10:26,640
Speaker 4: out can we improve that. It's a moment to apply AI.

222
00:10:27,200 --> 00:10:29,440
Speaker 4: And when we look at that this significant improvement. In

223
00:10:29,440 --> 00:10:33,439
Speaker 4: some cases we have thirty percent uplift, thirty percent, thirty percent,

224
00:10:33,480 --> 00:10:37,000
Speaker 4: three zero not everywhere. In some places we got but

225
00:10:37,040 --> 00:10:39,439
Speaker 4: it ranges between ten to thirty percent uplift depending on

226
00:10:39,480 --> 00:10:41,520
Speaker 4: the type of AI product you're building.

227
00:10:41,559 --> 00:10:43,079
Speaker 5: And that impacts the top line.

228
00:10:43,440 --> 00:10:45,960
Speaker 4: So it's very easy to also realize that value.

229
00:10:46,000 --> 00:10:47,000
Speaker 5: People get to see.

230
00:10:46,760 --> 00:10:49,480
Speaker 3: It, so you say, oh, we we can do a

231
00:10:49,480 --> 00:10:51,640
Speaker 3: way better job if we spend x more or x

232
00:10:51,760 --> 00:10:54,440
Speaker 3: less in this particular area. Give me another example of

233
00:10:54,480 --> 00:10:54,920
Speaker 3: a if.

234
00:10:54,920 --> 00:10:57,880
Speaker 4: Another one would be we have a very big large

235
00:10:57,920 --> 00:11:01,679
Speaker 4: salesforce within Ainiken. So this sales reap. What they do

236
00:11:01,720 --> 00:11:04,480
Speaker 4: they go to the outlets, the bars and restaurants, and

237
00:11:04,559 --> 00:11:08,040
Speaker 4: they maintain that human to human relationships with these our

238
00:11:08,080 --> 00:11:11,240
Speaker 4: customers is super important to maintain that, and they go

239
00:11:11,360 --> 00:11:13,959
Speaker 4: solve the customer problems. Let's say someone is out of stock,

240
00:11:14,040 --> 00:11:17,320
Speaker 4: someone is about to churn, or their surprise mismatch, something

241
00:11:17,400 --> 00:11:19,720
Speaker 4: like this, and before they us to go like this,

242
00:11:19,880 --> 00:11:22,120
Speaker 4: let's say a sales rep on a day to day

243
00:11:22,200 --> 00:11:25,640
Speaker 4: job has to visit five places ab CD E, five

244
00:11:25,640 --> 00:11:29,760
Speaker 4: different outlets, and he used to go ABCDE. Turns out,

245
00:11:29,920 --> 00:11:32,679
Speaker 4: the model tells you on any given day, if you

246
00:11:32,760 --> 00:11:37,120
Speaker 4: optimize taking into account the traffic conditions, instead of going

247
00:11:37,160 --> 00:11:40,120
Speaker 4: from that linear route, you go to D first and

248
00:11:40,160 --> 00:11:42,000
Speaker 4: then to B, then to C, then to E and

249
00:11:42,040 --> 00:11:44,240
Speaker 4: then to A. And the reason for doing that is

250
00:11:44,600 --> 00:11:47,439
Speaker 4: the model tells you, if you visit customer D first,

251
00:11:47,440 --> 00:11:49,760
Speaker 4: he has the biggest problem that needs the most amount

252
00:11:49,800 --> 00:11:52,600
Speaker 4: of time to be solved, and that's how it optimize.

253
00:11:52,640 --> 00:11:54,280
Speaker 5: And also the sales reps now.

254
00:11:54,120 --> 00:11:57,600
Speaker 4: They are becoming so educated with some of these AI models,

255
00:11:57,840 --> 00:12:00,800
Speaker 4: they are now becoming a business advisors. So they are

256
00:12:01,080 --> 00:12:04,280
Speaker 4: no longer just solving little problems. They are having the

257
00:12:04,360 --> 00:12:06,400
Speaker 4: time to say what else can I do for you

258
00:12:06,960 --> 00:12:09,400
Speaker 4: as the customer? So that I think it was a

259
00:12:09,400 --> 00:12:12,000
Speaker 4: big change within Heineken because it impacted a lot of

260
00:12:12,000 --> 00:12:13,679
Speaker 4: people that were using that.

261
00:12:13,880 --> 00:12:16,560
Speaker 3: In an instance, it requires not just building a model

262
00:12:17,040 --> 00:12:19,600
Speaker 3: that can be smarter about how people should spend their

263
00:12:19,600 --> 00:12:21,720
Speaker 3: time and what they should say, but you have to

264
00:12:21,880 --> 00:12:27,280
Speaker 3: obviously educate your salesforce to believe in what the selling

265
00:12:27,840 --> 00:12:29,840
Speaker 3: tell me about that piece. Is that a piece that

266
00:12:29,520 --> 00:12:31,360
Speaker 3: you that you're a part of or is someone else?

267
00:12:31,480 --> 00:12:33,959
Speaker 4: Yeah, that's also part of because that is super important.

268
00:12:33,960 --> 00:12:36,680
Speaker 4: I think we can build the best models. Best algorithm's

269
00:12:36,760 --> 00:12:39,800
Speaker 4: highest accuracy doesn't mean anything if it's not used the

270
00:12:39,840 --> 00:12:42,199
Speaker 4: right way. So what we do we have within our

271
00:12:42,240 --> 00:12:46,440
Speaker 4: company a pretty big upskilling program. So bring everyone along

272
00:12:47,080 --> 00:12:50,640
Speaker 4: in common understanding, basic understanding of what AI does. Not

273
00:12:50,720 --> 00:12:54,640
Speaker 4: everyone needs to understand new networks or algorithms right, But

274
00:12:54,760 --> 00:12:58,320
Speaker 4: what we do is give them a handheld device and

275
00:12:58,360 --> 00:13:00,839
Speaker 4: an app which is driven by AI. Play with it,

276
00:13:01,200 --> 00:13:03,760
Speaker 4: have fun, see how it changes your life. And once

277
00:13:03,800 --> 00:13:07,160
Speaker 4: you start liking the product, liking the UIUX, then you

278
00:13:07,200 --> 00:13:09,680
Speaker 4: start getting more. And the insights also tell you the

279
00:13:09,679 --> 00:13:12,319
Speaker 4: story because once you start getting the value, I am

280
00:13:12,360 --> 00:13:16,840
Speaker 4: not having to pitch my models anymore. The sales reps

281
00:13:17,080 --> 00:13:19,439
Speaker 4: and the markets they are pitching on behalf of us.

282
00:13:19,960 --> 00:13:21,760
Speaker 5: And that's such a good place to.

283
00:13:21,720 --> 00:13:26,040
Speaker 3: Be, is it. It's interesting? So in that instance where

284
00:13:26,040 --> 00:13:29,760
Speaker 3: you're designing a more efficient form of interaction and fruitful

285
00:13:29,840 --> 00:13:33,000
Speaker 3: for of interaction between sales reps and customers, I could

286
00:13:33,040 --> 00:13:35,920
Speaker 3: see a version of that where it is really clear

287
00:13:36,559 --> 00:13:40,480
Speaker 3: looking up from a high level that things are working better,

288
00:13:40,720 --> 00:13:43,680
Speaker 3: but it might not be clear to the salesperson. Is

289
00:13:43,720 --> 00:13:46,680
Speaker 3: the salesperson who's now following the direction of the AI

290
00:13:47,200 --> 00:13:48,800
Speaker 3: aware that they are more efficient?

291
00:13:48,920 --> 00:13:50,359
Speaker 5: They are?

292
00:13:50,559 --> 00:13:51,600
Speaker 3: How are they aware the market?

293
00:13:51,600 --> 00:13:54,360
Speaker 4: They realize few things that they were visiting customers just

294
00:13:54,400 --> 00:13:57,040
Speaker 4: because they had to visit because it was in the schedule.

295
00:13:57,520 --> 00:13:59,959
Speaker 4: Now they go there and they find out, wait a minute,

296
00:14:00,200 --> 00:14:03,160
Speaker 4: I never tackled this big problem that was not being addressed,

297
00:14:03,600 --> 00:14:07,120
Speaker 4: and they solved it. And the customer feedback also comes

298
00:14:07,160 --> 00:14:09,920
Speaker 4: back saying we are really happy. So for all these products,

299
00:14:10,000 --> 00:14:12,000
Speaker 4: we get the feedback not just from the sales reps,

300
00:14:12,160 --> 00:14:14,760
Speaker 4: but also for the customers. Do you really like the

301
00:14:14,840 --> 00:14:17,920
Speaker 4: recommendations we are giving you? And that's the best validation

302
00:14:17,960 --> 00:14:20,440
Speaker 4: you can think of, because it's four stand or feedback.

303
00:14:20,880 --> 00:14:23,800
Speaker 3: When the AI is doing this ranking, it wants you

304
00:14:23,840 --> 00:14:27,920
Speaker 3: to focus on the customer with the biggest problem first

305
00:14:28,000 --> 00:14:29,240
Speaker 3: or is it much more complex than that.

306
00:14:29,320 --> 00:14:30,680
Speaker 5: It's a little bit more complex than that.

307
00:14:30,760 --> 00:14:33,400
Speaker 4: Yeah, but usually it's a rank pordert in terms of

308
00:14:33,400 --> 00:14:35,960
Speaker 4: which one is the biggest problem that needs the most

309
00:14:35,960 --> 00:14:36,640
Speaker 4: amount of time.

310
00:14:37,200 --> 00:14:38,160
Speaker 5: That's how it's ranked order.

311
00:14:38,240 --> 00:14:40,440
Speaker 4: But sometimes you can also override the model, right, We

312
00:14:40,520 --> 00:14:43,720
Speaker 4: also give options to people. You don't have to all

313
00:14:43,800 --> 00:14:46,080
Speaker 4: the time one hundred percent follow the recommendation. If you

314
00:14:46,080 --> 00:14:48,920
Speaker 4: have some urgent priority, you can overwrite that.

315
00:14:48,920 --> 00:14:51,800
Speaker 3: That's also possible with something like that. Is there a

316
00:14:51,920 --> 00:14:55,200
Speaker 3: next level you can go to? So you design this

317
00:14:55,280 --> 00:14:56,920
Speaker 3: system and you say, oh, I can make our sales

318
00:14:56,920 --> 00:14:59,480
Speaker 3: staff a lot more effective and the way they operate

319
00:14:59,560 --> 00:15:03,960
Speaker 3: with their customers, and then you see that it works,

320
00:15:04,320 --> 00:15:06,000
Speaker 3: and then it comes back and then you say, okay,

321
00:15:06,000 --> 00:15:09,240
Speaker 3: what's two point oh? Is there a two point oh.

322
00:15:09,040 --> 00:15:09,480
Speaker 5: It could be.

323
00:15:09,600 --> 00:15:13,120
Speaker 4: So it's always about innovation. Then you think, okay, today

324
00:15:13,160 --> 00:15:15,360
Speaker 4: we go and solve the problems that have already happened.

325
00:15:16,400 --> 00:15:19,080
Speaker 4: What if we solve the problems that are likely to happen,

326
00:15:19,680 --> 00:15:22,440
Speaker 4: that will be the next step. So this customer hasn't

327
00:15:22,480 --> 00:15:25,160
Speaker 4: been very active for a while, there's a high chance

328
00:15:25,200 --> 00:15:28,240
Speaker 4: that that customer might churn out of Heineken. So what

329
00:15:28,320 --> 00:15:31,640
Speaker 4: actions can I actually recommend to make sure And we

330
00:15:31,720 --> 00:15:33,640
Speaker 4: do this, by the way, we also gather a lot

331
00:15:33,680 --> 00:15:37,040
Speaker 4: of customer feedback and complaints and feedback, and we use

332
00:15:37,160 --> 00:15:39,600
Speaker 4: LLLMS to extract and glean information. Okay, what are the

333
00:15:39,600 --> 00:15:42,080
Speaker 4: real pain points? What's the theme and the topic that

334
00:15:42,200 --> 00:15:44,760
Speaker 4: needs to be addressed. And once you do that, then

335
00:15:44,800 --> 00:15:47,520
Speaker 4: also you can prepare ahead of time. We're already there,

336
00:15:47,560 --> 00:15:49,120
Speaker 4: by the way, when I say two point zero, we're

337
00:15:49,160 --> 00:15:52,600
Speaker 4: already testing it. You solve problems or you'll try to

338
00:15:52,600 --> 00:15:55,440
Speaker 4: solve problems before daven occur. So I think that's a

339
00:15:55,480 --> 00:15:57,280
Speaker 4: little bit of a two point zero. Then we have

340
00:15:57,320 --> 00:15:58,720
Speaker 4: to see what else we can do with it.

341
00:15:59,240 --> 00:16:03,600
Speaker 3: Tell me you've had this partnership at Heineken with IBM

342
00:16:04,240 --> 00:16:08,480
Speaker 3: for since twenty thirteen, twenty thirteen, So you came in

343
00:16:08,600 --> 00:16:12,200
Speaker 3: and there was already a strong working relationship. Tell me

344
00:16:12,240 --> 00:16:14,760
Speaker 3: about how that relationship started and what does it mean

345
00:16:14,800 --> 00:16:18,920
Speaker 3: on a practical basis. You're building all these tools, how

346
00:16:18,920 --> 00:16:20,400
Speaker 3: does the interaction with IBM work.

347
00:16:20,920 --> 00:16:22,640
Speaker 4: Yeah, so I think good to give a little bit

348
00:16:22,640 --> 00:16:26,760
Speaker 4: of context. Heineken started this digital transformation journey in twenty

349
00:16:26,840 --> 00:16:30,040
Speaker 4: twenty formally, but the tech was already there. We had

350
00:16:30,080 --> 00:16:34,120
Speaker 4: our systems, platforms, data, everything was there. So all the

351
00:16:34,200 --> 00:16:37,600
Speaker 4: IT four ID systems is where IBM was partnering with

352
00:16:37,680 --> 00:16:40,280
Speaker 4: us from the get go, from twenty thirteen, and it's

353
00:16:40,320 --> 00:16:43,080
Speaker 4: a very long standing partnership because as we found the

354
00:16:43,160 --> 00:16:46,560
Speaker 4: tech is evolving, our partnership also kept evolving because we

355
00:16:46,600 --> 00:16:49,600
Speaker 4: need to keep up up to the speed. So it

356
00:16:49,720 --> 00:16:53,520
Speaker 4: was more about IT for IT systems, cybersecurity platform, data,

357
00:16:54,280 --> 00:16:56,600
Speaker 4: incident management, service level, you name it.

358
00:16:57,000 --> 00:16:57,520
Speaker 5: All of that.

359
00:16:57,680 --> 00:17:00,600
Speaker 4: IBM was supporting us both in terms of hands and

360
00:17:00,680 --> 00:17:01,560
Speaker 4: also in terms.

361
00:17:01,360 --> 00:17:02,760
Speaker 5: Of strategy to create together.

362
00:17:03,360 --> 00:17:05,800
Speaker 4: But that also evolved, like I said, when we went

363
00:17:05,840 --> 00:17:09,600
Speaker 4: into this digital transformation journey in twenty twenty, then we

364
00:17:09,680 --> 00:17:13,960
Speaker 4: started building this digital core, which is the central nervous

365
00:17:13,960 --> 00:17:17,680
Speaker 4: system software system of finikin that's where IBM is really

366
00:17:17,720 --> 00:17:20,600
Speaker 4: partnering with us and helping us not just shape the

367
00:17:20,640 --> 00:17:22,560
Speaker 4: whole thing in terms of building it hands on, but

368
00:17:22,600 --> 00:17:25,080
Speaker 4: how do we strategize that so that it lands well.

369
00:17:25,160 --> 00:17:28,720
Speaker 4: So that's yeah, it's a long trusted partnership. I think

370
00:17:28,720 --> 00:17:30,200
Speaker 4: we are going to go a long way together.

371
00:17:30,560 --> 00:17:33,520
Speaker 3: Yeah, Heineken Space just sells out of Amsterdam.

372
00:17:34,160 --> 00:17:36,480
Speaker 5: The head office is absolutely so the IBN.

373
00:17:36,160 --> 00:17:38,359
Speaker 3: People who work with you are they on site?

374
00:17:38,640 --> 00:17:40,720
Speaker 4: There are some on site and there are some teams

375
00:17:40,720 --> 00:17:43,680
Speaker 4: in India some times spread across the globe. But for

376
00:17:43,680 --> 00:17:46,000
Speaker 4: for tech, I think the location doesn't matter. But you

377
00:17:46,040 --> 00:17:48,080
Speaker 4: still need people on site to actually talk with the

378
00:17:48,119 --> 00:17:50,960
Speaker 4: business and really understand what the problem for.

379
00:17:51,080 --> 00:17:52,720
Speaker 5: Those interactions are also very important.

380
00:17:53,040 --> 00:17:54,639
Speaker 3: When you said you wanted to when you got there,

381
00:17:54,640 --> 00:17:57,560
Speaker 3: you wanted to build a digital nervous system, what does

382
00:17:57,560 --> 00:17:58,000
Speaker 3: that mean?

383
00:17:58,440 --> 00:17:59,720
Speaker 5: Maybe good to give you an example.

384
00:18:00,200 --> 00:18:04,919
Speaker 4: Let's say iPhone, right, it's a central platform, but you

385
00:18:04,960 --> 00:18:08,440
Speaker 4: can download thousands of apps there and all of them

386
00:18:08,800 --> 00:18:11,600
Speaker 4: once you download seamlessly integrates with the system and you

387
00:18:11,600 --> 00:18:14,439
Speaker 4: don't see any difference. This is the same thing. So

388
00:18:14,480 --> 00:18:16,639
Speaker 4: what we want to build within Heineken is a central

389
00:18:16,680 --> 00:18:19,159
Speaker 4: software system which the old school way of saying it

390
00:18:19,200 --> 00:18:22,879
Speaker 4: is the ERP enterprise resource planning. It removes the fragmentation

391
00:18:22,920 --> 00:18:25,760
Speaker 4: of different platforms, it brings it all together. It makes

392
00:18:25,760 --> 00:18:29,240
Speaker 4: sure all the business applications within supply chain, commerce, finance

393
00:18:29,440 --> 00:18:33,600
Speaker 4: HR all in one place and coordinates them everything orchestrates them.

394
00:18:34,080 --> 00:18:37,040
Speaker 4: The benefit of doing that is two one across the

395
00:18:37,119 --> 00:18:39,919
Speaker 4: value channing of one way of doing business because everything

396
00:18:39,960 --> 00:18:44,560
Speaker 4: is standardized, but we also have multiple markets globally. Across

397
00:18:44,600 --> 00:18:47,440
Speaker 4: the multiple markets also it becomes one way of doing business,

398
00:18:47,640 --> 00:18:50,840
Speaker 4: so it's both ways. And once you standardize it, we

399
00:18:50,920 --> 00:18:55,360
Speaker 4: can embed new apps which will seamlessly integrate, and then

400
00:18:55,400 --> 00:18:58,200
Speaker 4: it just keeps scaling further. Can we scaled very quick

401
00:18:58,520 --> 00:19:01,359
Speaker 4: and having that digital core really help us scale because

402
00:19:01,400 --> 00:19:04,840
Speaker 4: the value from AI and insights is not just building

403
00:19:04,840 --> 00:19:07,359
Speaker 4: one product in one place. It's how quickly can you

404
00:19:07,400 --> 00:19:07,840
Speaker 4: scale it?

405
00:19:08,240 --> 00:19:10,000
Speaker 3: And are you still building it or is it an

406
00:19:10,040 --> 00:19:10,760
Speaker 3: ongoing thing or.

407
00:19:10,680 --> 00:19:13,160
Speaker 4: It's ongoing thing because there are nuances in markets, There

408
00:19:13,320 --> 00:19:17,359
Speaker 4: are nuances in tax systems and currency systems, so it

409
00:19:17,400 --> 00:19:19,440
Speaker 4: takes a little bit of as much as we want

410
00:19:19,440 --> 00:19:21,480
Speaker 4: to standardize, you also have to bake in some of

411
00:19:21,520 --> 00:19:25,440
Speaker 4: the nuances otherwise people cannot use it. So those sort

412
00:19:25,440 --> 00:19:27,560
Speaker 4: of outliers we have to also bake in.

413
00:19:28,240 --> 00:19:31,720
Speaker 3: You must learn something when you suddenly suddenly, but when

414
00:19:31,760 --> 00:19:33,800
Speaker 3: you standardize a bunch of things that have not been

415
00:19:33,840 --> 00:19:37,520
Speaker 3: standardized before, presumably you have a basis for comparisons you

416
00:19:37,560 --> 00:19:38,479
Speaker 3: couldn't make before.

417
00:19:39,840 --> 00:19:40,399
Speaker 5: That's correct.

418
00:19:40,440 --> 00:19:43,439
Speaker 4: So we also get a lot of external inspiration. So

419
00:19:43,520 --> 00:19:46,560
Speaker 4: sometimes these large projects we don't start by over whom,

420
00:19:47,000 --> 00:19:50,320
Speaker 4: so we get inspiration from partners like IBM or someone else.

421
00:19:51,040 --> 00:19:52,840
Speaker 4: How have they done it in somewhere else and where

422
00:19:52,880 --> 00:19:56,560
Speaker 4: it's really working. So then you get those ideas, the learnings,

423
00:19:56,840 --> 00:19:59,639
Speaker 4: and you start building that way, and while doing that,

424
00:19:59,680 --> 00:20:01,600
Speaker 4: you feel out that, wait a minute, we might have

425
00:20:01,640 --> 00:20:04,280
Speaker 4: done something different and maybe it's even better than what

426
00:20:04,359 --> 00:20:05,000
Speaker 4: others have done.

427
00:20:05,440 --> 00:20:07,280
Speaker 5: Yeah, so it also creates creativity.

428
00:20:07,520 --> 00:20:11,639
Speaker 3: Yeah, I'm just curious whether there was an insight that

429
00:20:11,720 --> 00:20:16,040
Speaker 3: you learned from that process that comes to mind A

430
00:20:16,080 --> 00:20:16,600
Speaker 3: big one.

431
00:20:16,680 --> 00:20:20,040
Speaker 4: I think we don't look at tech for the sake

432
00:20:20,080 --> 00:20:22,600
Speaker 4: of tech and embedding it, you know, just just as

433
00:20:22,920 --> 00:20:25,640
Speaker 4: one would say it's one single core, one single platform,

434
00:20:25,680 --> 00:20:29,080
Speaker 4: everything coordinated. What's the big deal. The big deal is

435
00:20:29,080 --> 00:20:31,560
Speaker 4: bringing people along to actually believe that there is a

436
00:20:31,720 --> 00:20:34,719
Speaker 4: benefit of doing one way of business, and that actually

437
00:20:34,760 --> 00:20:37,240
Speaker 4: means the entire company, not just the leadership team.

438
00:20:37,520 --> 00:20:38,320
Speaker 5: So to bring.

439
00:20:38,160 --> 00:20:41,679
Speaker 4: Everyone on board and say, tell us how this platform

440
00:20:41,720 --> 00:20:43,760
Speaker 4: should look like, what are the components we should build?

441
00:20:44,160 --> 00:20:46,440
Speaker 4: It's a pretty big task. Yeah, that's where the chain

442
00:20:46,520 --> 00:20:47,280
Speaker 4: management comes in.

443
00:20:47,359 --> 00:20:50,840
Speaker 3: Yeah, what was hard about that? Did you have bumps

444
00:20:51,080 --> 00:20:51,960
Speaker 3: or we did?

445
00:20:52,080 --> 00:20:55,359
Speaker 4: I think it's about convincing people the benefit of doing this.

446
00:20:55,920 --> 00:20:58,000
Speaker 4: Why do we say, if you standardize something, we can

447
00:20:58,040 --> 00:21:00,640
Speaker 4: go at high speed in scaling. It's not very easy

448
00:21:00,640 --> 00:21:03,199
Speaker 4: to visualize that at first. But what you do is

449
00:21:03,240 --> 00:21:05,960
Speaker 4: you show some proof of concepts. And that's I won't

450
00:21:05,960 --> 00:21:08,240
Speaker 4: call it a trick. It's almost bread and butter of

451
00:21:08,359 --> 00:21:11,320
Speaker 4: what we do. Show a small proof of concept, show

452
00:21:11,359 --> 00:21:13,919
Speaker 4: that it works, show that we can scale, and then

453
00:21:13,960 --> 00:21:16,120
Speaker 4: automatically if people start having the faith, and then well

454
00:21:16,160 --> 00:21:17,400
Speaker 4: say okay, I see it.

455
00:21:17,520 --> 00:21:18,679
Speaker 5: Yeah it makes sense.

456
00:21:19,000 --> 00:21:22,080
Speaker 3: Sergery, at least half of what you've talked about is

457
00:21:22,119 --> 00:21:24,080
Speaker 3: not about the check itself, but about being a kind

458
00:21:24,080 --> 00:21:27,320
Speaker 3: of evangelist for the check. It is half the right percentage.

459
00:21:27,440 --> 00:21:30,920
Speaker 3: How much of your time is spent convincing an organization

460
00:21:31,440 --> 00:21:33,480
Speaker 3: and people in the organization to see the value in

461
00:21:33,480 --> 00:21:35,480
Speaker 3: what you're doing as opposed to building the thing that

462
00:21:35,600 --> 00:21:36,280
Speaker 3: has value.

463
00:21:36,440 --> 00:21:38,879
Speaker 4: Yeah, I think that proportion changed over time. When I

464
00:21:38,920 --> 00:21:41,280
Speaker 4: first joined, I was very much into the products itself.

465
00:21:41,680 --> 00:21:44,000
Speaker 4: I was to review codes myself, let me check what's

466
00:21:44,040 --> 00:21:46,480
Speaker 4: going on. And over time, of course, then you focus

467
00:21:46,560 --> 00:21:49,480
Speaker 4: on somewhere else. You realize, like I said, best codes,

468
00:21:49,520 --> 00:21:52,280
Speaker 4: best models, I use this, if not use the right way,

469
00:21:52,359 --> 00:21:54,800
Speaker 4: then I said, okay, now my time is to actually

470
00:21:54,840 --> 00:21:57,840
Speaker 4: inspire and show people the value of it. What I

471
00:21:57,920 --> 00:22:02,760
Speaker 4: realized is explaining in very simple language really goes a

472
00:22:02,800 --> 00:22:06,639
Speaker 4: long way because you take away that that anxiety that

473
00:22:06,960 --> 00:22:09,359
Speaker 4: a new product is coming in and we humans a

474
00:22:09,400 --> 00:22:12,040
Speaker 4: little bit have this. I don't know if it's the

475
00:22:12,119 --> 00:22:14,560
Speaker 4: right thing to say, but it's inertia of rest. We

476
00:22:14,760 --> 00:22:17,919
Speaker 4: like status, poke, we don't sometimes like change that stops our.

477
00:22:18,640 --> 00:22:20,320
Speaker 4: So every time you build a new product that will

478
00:22:20,400 --> 00:22:22,879
Speaker 4: change our way of working. Yeah, there's inherent little bit

479
00:22:22,880 --> 00:22:26,000
Speaker 4: of anxiety. Yeah, take that away. Yeah, a job becomes

480
00:22:26,000 --> 00:22:26,640
Speaker 4: a lot easier.

481
00:22:26,800 --> 00:22:29,159
Speaker 3: Are you a good evangelist?

482
00:22:29,480 --> 00:22:30,479
Speaker 5: So far it's working?

483
00:22:30,640 --> 00:22:33,760
Speaker 4: I think I can do better, for sure, because it's

484
00:22:33,800 --> 00:22:37,760
Speaker 4: about understanding what's the reason people sometimes might be reluctant

485
00:22:38,160 --> 00:22:41,760
Speaker 4: to actually onboard or adopt a new technology. And once

486
00:22:41,760 --> 00:22:45,240
Speaker 4: you sort of understand that, then that anxiety goes away,

487
00:22:45,280 --> 00:22:46,240
Speaker 4: it becomes easier.

488
00:22:46,640 --> 00:22:48,520
Speaker 3: How many people work for a hidekinde.

489
00:22:48,600 --> 00:22:50,760
Speaker 5: About eighty five thousand, ninety thousand.

490
00:22:50,680 --> 00:22:55,160
Speaker 3: Globally, so you have essentially a city pretty much. And

491
00:22:55,480 --> 00:22:57,440
Speaker 3: if you look at that universe of eighty five thousand,

492
00:22:58,000 --> 00:23:00,719
Speaker 3: is there anyone in that universe who is not what

493
00:23:00,760 --> 00:23:01,200
Speaker 3: you're doing?

494
00:23:02,560 --> 00:23:05,200
Speaker 4: The way we do it is we prioritize based on

495
00:23:05,280 --> 00:23:09,080
Speaker 4: the size of the market and the potential opportunity. Yes,

496
00:23:09,200 --> 00:23:12,560
Speaker 4: if I had infinite resources, I would go everywhere within

497
00:23:12,560 --> 00:23:15,119
Speaker 4: the high Naken company and do everything, But we cannot.

498
00:23:15,400 --> 00:23:16,840
Speaker 5: We don't have infinite resources.

499
00:23:17,000 --> 00:23:18,960
Speaker 4: So we say, let's be a little bit picky and

500
00:23:19,000 --> 00:23:21,880
Speaker 4: choosy where the biggest opportunities are. But it's a matter

501
00:23:21,920 --> 00:23:24,560
Speaker 4: of time. Right today we touch upon the big market's

502
00:23:24,600 --> 00:23:28,080
Speaker 4: biggest scope. Over time, it's going to be pervasive through

503
00:23:28,080 --> 00:23:31,600
Speaker 4: the company. But the appetite is already there, so people

504
00:23:31,640 --> 00:23:34,280
Speaker 4: are really even if they have not really embedded some product,

505
00:23:34,720 --> 00:23:37,840
Speaker 4: they're asking for it, which is a fantastic place to be.

506
00:23:38,400 --> 00:23:41,359
Speaker 3: Yeah, yeah, what's been your biggest disappointment?

507
00:23:41,440 --> 00:23:41,840
Speaker 5: So far.

508
00:23:42,400 --> 00:23:44,640
Speaker 4: So far, it's been very fulfilling, I must say, but

509
00:23:44,720 --> 00:23:47,159
Speaker 4: I think what I would look back. Can we do

510
00:23:47,240 --> 00:23:49,240
Speaker 4: things a little bit quicker? Can we go a little

511
00:23:49,240 --> 00:23:52,400
Speaker 4: bit at high speed? And that's why this whole concept

512
00:23:52,400 --> 00:23:55,239
Speaker 4: of digital backbone. Can we standardized everything? If we can

513
00:23:55,320 --> 00:23:58,119
Speaker 4: speed that up, if we can really scale quickly, I

514
00:23:58,119 --> 00:24:00,159
Speaker 4: think that will be the best. Because today have a

515
00:24:00,240 --> 00:24:03,520
Speaker 4: very good problem. People are asking for products. Sometimes I say, yeah,

516
00:24:03,600 --> 00:24:07,040
Speaker 4: I need to put it on a timeline in a roadmap,

517
00:24:07,119 --> 00:24:08,960
Speaker 4: because I cannot just cater to it immediately.

518
00:24:09,400 --> 00:24:12,320
Speaker 3: Presumably that's one of the things that the people you're

519
00:24:12,320 --> 00:24:14,280
Speaker 3: working with at IBM can tell you. They can give

520
00:24:14,280 --> 00:24:18,639
Speaker 3: you a sense of how quickly others have adopted some

521
00:24:18,680 --> 00:24:19,440
Speaker 3: of these technology.

522
00:24:19,480 --> 00:24:21,400
Speaker 4: That's correct, and that's actually one of the bench funds

523
00:24:21,440 --> 00:24:24,200
Speaker 4: that you're referring to. We see I'll be losing pace,

524
00:24:24,240 --> 00:24:26,679
Speaker 4: and which other things can we go forward? And in

525
00:24:26,720 --> 00:24:28,720
Speaker 4: some case when you look at the digital core and

526
00:24:28,760 --> 00:24:32,040
Speaker 4: the backbone, maybe specific areas we can speed up because

527
00:24:32,080 --> 00:24:34,399
Speaker 4: those are the areas that are maximum potential value, and

528
00:24:34,440 --> 00:24:36,240
Speaker 4: some of them we can deprioritize a little bit.

529
00:24:36,359 --> 00:24:38,760
Speaker 5: Yeah, that we do all the time. Yeah, just a

530
00:24:38,760 --> 00:24:39,880
Speaker 5: pragmatic approach.

531
00:24:39,600 --> 00:24:43,879
Speaker 3: To I'm curious about a highly specific question, which is,

532
00:24:44,560 --> 00:24:49,800
Speaker 3: so here you have a legacy brewer bas in the Netherlands,

533
00:24:49,840 --> 00:24:53,840
Speaker 3: one hundred and sixty years old. If I were to say,

534
00:24:54,240 --> 00:24:56,480
Speaker 3: I want you to take an entirely new job, I

535
00:24:56,520 --> 00:24:58,800
Speaker 3: want you to do what you're doing, but I want

536
00:24:58,840 --> 00:25:02,159
Speaker 3: you to do it for an American company in a

537
00:25:02,240 --> 00:25:07,479
Speaker 3: completely different industry that's thirty years old, maybe a company

538
00:25:07,480 --> 00:25:11,919
Speaker 3: that makes the vacuum cleaners thirty years old in America.

539
00:25:11,960 --> 00:25:13,760
Speaker 3: How much of what you're doing. I guess what I'm

540
00:25:13,760 --> 00:25:16,159
Speaker 3: trying to say is, are there's things that are particular

541
00:25:16,200 --> 00:25:20,360
Speaker 3: to Heineken that have made your job sort of challenging

542
00:25:20,400 --> 00:25:23,520
Speaker 3: or interesting or that just wouldn't be an issue in

543
00:25:23,560 --> 00:25:24,680
Speaker 3: another environment.

544
00:25:25,240 --> 00:25:28,520
Speaker 4: So it's a good question, and thanks for the enticing offer,

545
00:25:28,600 --> 00:25:30,840
Speaker 4: but I will pollite.

546
00:25:31,040 --> 00:25:33,760
Speaker 3: I tried to make it as an average exactly.

547
00:25:33,400 --> 00:25:35,880
Speaker 4: But I won't polite to reject offer. But I'll tell

548
00:25:35,880 --> 00:25:36,560
Speaker 4: you why i'll reject.

549
00:25:36,600 --> 00:25:39,320
Speaker 3: You're going to be in Nebraska. They're making just one

550
00:25:39,400 --> 00:25:40,359
Speaker 3: kind of vacuum cleaner.

551
00:25:40,520 --> 00:25:43,159
Speaker 4: What I went to school in Iowas, Yes, exactly, So

552
00:25:43,240 --> 00:25:46,520
Speaker 4: I'm quite a bit. I think there's a cultural difference

553
00:25:46,720 --> 00:25:50,240
Speaker 4: we're all very passionate about our brands and products, and

554
00:25:50,280 --> 00:25:52,119
Speaker 4: there's a lot of it is connection based in the

555
00:25:52,160 --> 00:25:55,280
Speaker 4: sense we create these connections with our customers sometimes consumers,

556
00:25:55,800 --> 00:25:58,240
Speaker 4: and it's all about maintaining that. And once you get

557
00:25:58,240 --> 00:26:00,000
Speaker 4: a feel of it, you feel part of the family.

558
00:26:01,000 --> 00:26:03,000
Speaker 4: That's a very good feeling to have. And the fact

559
00:26:03,000 --> 00:26:04,639
Speaker 4: that today where I am, if I look back, I

560
00:26:04,720 --> 00:26:08,520
Speaker 4: probably will happy to say very fortunate to have probably

561
00:26:08,520 --> 00:26:09,800
Speaker 4: one of the best jobs in the world in the

562
00:26:09,840 --> 00:26:12,200
Speaker 4: current times. And there is no end to innovation, by

563
00:26:12,200 --> 00:26:14,760
Speaker 4: the way, and even within Heineken, yes it's a traditional

564
00:26:14,760 --> 00:26:17,600
Speaker 4: company who is stopping innovation. There's a lot more too,

565
00:26:18,240 --> 00:26:20,359
Speaker 4: so I'll be very busy for the next few years.

566
00:26:21,359 --> 00:26:23,439
Speaker 3: What are the Dutch like? This is one of the

567
00:26:23,440 --> 00:26:27,720
Speaker 3: oldest and most successful commercial cultures in the world, A

568
00:26:27,760 --> 00:26:31,000
Speaker 3: tiny country that's been solidly successful, that's for it. I'm

569
00:26:31,040 --> 00:26:34,560
Speaker 3: curious about innovating in that kind of environment. How is

570
00:26:34,600 --> 00:26:37,359
Speaker 3: that different from innovating in a huge country like the

571
00:26:37,440 --> 00:26:40,600
Speaker 3: United States or in a different kind of national culture.

572
00:26:41,200 --> 00:26:43,960
Speaker 4: I think it's a question of opportunity because within Netherlands,

573
00:26:43,960 --> 00:26:47,280
Speaker 4: by the way, Netherlands has one of the most highest

574
00:26:47,320 --> 00:26:51,280
Speaker 4: number of startups within Europe, if not the highest. So

575
00:26:51,359 --> 00:26:54,760
Speaker 4: there is this culture of innovation that's already embedded in there.

576
00:26:54,800 --> 00:26:57,840
Speaker 4: It's happening all the time. Companies like Phillips, ASML some

577
00:26:57,920 --> 00:27:01,640
Speaker 4: of the very big players already there. So it could

578
00:27:01,640 --> 00:27:04,399
Speaker 4: be a little bit different. I think in Netherlands we

579
00:27:04,440 --> 00:27:06,080
Speaker 4: want to make sure what we are doing really is

580
00:27:06,119 --> 00:27:08,320
Speaker 4: going to work, so there's a little bit of discussion alignment.

581
00:27:08,840 --> 00:27:11,960
Speaker 4: It's more structured, but also agile in a way we

582
00:27:12,040 --> 00:27:15,000
Speaker 4: do things, and us was more like let's do let's

583
00:27:15,000 --> 00:27:18,680
Speaker 4: go quick, experiment, learn fail. So I think there's pros

584
00:27:18,680 --> 00:27:22,120
Speaker 4: and cons on both sides, but so far it's quite good.

585
00:27:22,480 --> 00:27:25,240
Speaker 3: Give me a sense of what your what's a day

586
00:27:25,240 --> 00:27:28,960
Speaker 3: in a life like for you? What does it look

587
00:27:29,080 --> 00:27:31,040
Speaker 3: like to have the job that you have in a

588
00:27:31,160 --> 00:27:31,760
Speaker 3: place a cande.

589
00:27:32,080 --> 00:27:34,280
Speaker 4: First of all, it starts with the calendar and the

590
00:27:34,359 --> 00:27:37,600
Speaker 4: number of meetings I have, which is usually filled for

591
00:27:37,600 --> 00:27:41,240
Speaker 4: forty hours longer in the week. So that's the starting point,

592
00:27:41,520 --> 00:27:43,200
Speaker 4: and then I have to pick and choose which meetings

593
00:27:43,240 --> 00:27:45,679
Speaker 4: I need to prepare for what, and usually these meetings

594
00:27:45,680 --> 00:27:48,480
Speaker 4: are mostly about where are we with the product.

595
00:27:48,119 --> 00:27:48,960
Speaker 5: What are the challenges?

596
00:27:49,040 --> 00:27:51,840
Speaker 4: How can I help and solve it, and then sometimes

597
00:27:51,880 --> 00:27:55,560
Speaker 4: also pitching new products or convincing something, and also sometimes

598
00:27:55,600 --> 00:27:58,760
Speaker 4: chain management. I'd also do sessions where I present internally

599
00:27:58,880 --> 00:28:01,800
Speaker 4: quite quite often go to different places, because it always

600
00:28:01,840 --> 00:28:04,280
Speaker 4: helps to be in front of the audience when you're

601
00:28:04,280 --> 00:28:08,000
Speaker 4: presenting something. We also started something recently which we call

602
00:28:08,080 --> 00:28:11,239
Speaker 4: AI boot Camp, which is you use JANEAI as an

603
00:28:11,280 --> 00:28:14,600
Speaker 4: interface for all these big AI models and people can

604
00:28:14,640 --> 00:28:17,960
Speaker 4: interact in a very fun way. That's our new way

605
00:28:17,960 --> 00:28:21,639
Speaker 4: of really convincing the rest of the company that hey,

606
00:28:21,680 --> 00:28:23,840
Speaker 4: this is fun to play with and let's go.

607
00:28:24,520 --> 00:28:27,280
Speaker 3: So yeah, it's how many people would you cycle through

608
00:28:28,280 --> 00:28:29,800
Speaker 3: that kind of book camp anyone?

609
00:28:30,000 --> 00:28:32,040
Speaker 4: Usually we keep it a small group just to make

610
00:28:32,080 --> 00:28:36,320
Speaker 4: sure everyone is doing something hands on and nobody's just listening.

611
00:28:36,800 --> 00:28:39,320
Speaker 4: So usually a twenty to thirty people max. And then

612
00:28:39,480 --> 00:28:41,560
Speaker 4: go from one place to other and it's all hands on.

613
00:28:41,720 --> 00:28:44,000
Speaker 4: You cannot sit and watch. You have to participate.

614
00:28:44,240 --> 00:28:46,840
Speaker 3: Are you directly involved at all in the design or

615
00:28:46,880 --> 00:28:49,120
Speaker 3: creation of any of these tools?

616
00:28:49,400 --> 00:28:51,720
Speaker 4: So I review it and then I used to review

617
00:28:51,720 --> 00:28:54,920
Speaker 4: also the codes before, and now I'm mostly like trying

618
00:28:54,960 --> 00:28:57,280
Speaker 4: to get the feedback from the people that are using it,

619
00:28:57,520 --> 00:28:59,440
Speaker 4: because that's my best validation point.

620
00:28:59,600 --> 00:29:02,400
Speaker 5: I high net promoter score on these products.

621
00:29:02,560 --> 00:29:04,600
Speaker 4: I know the job is well done, but I do

622
00:29:04,680 --> 00:29:06,920
Speaker 4: check accuracy of model. Some of the basic things you'll

623
00:29:07,000 --> 00:29:09,720
Speaker 4: check in AI. Is the model drifting over time? What's

624
00:29:09,720 --> 00:29:10,400
Speaker 4: the accuracy?

625
00:29:10,480 --> 00:29:10,640
Speaker 5: You know?

626
00:29:10,760 --> 00:29:13,440
Speaker 4: How is it hosted on a platform? These things I check.

627
00:29:13,800 --> 00:29:15,920
Speaker 4: But we also have mechanisms on those, so it's not

628
00:29:16,000 --> 00:29:17,600
Speaker 4: like every time you have a DiPT deep and look

629
00:29:17,600 --> 00:29:18,160
Speaker 4: into everything.

630
00:29:18,480 --> 00:29:18,720
Speaker 5: Yeah.

631
00:29:18,720 --> 00:29:21,160
Speaker 4: So once you have these mechanisms in place, then these

632
00:29:21,240 --> 00:29:22,480
Speaker 4: sort of tasks become easier.

633
00:29:23,280 --> 00:29:25,280
Speaker 3: You've used the phrase that you want to make honey

634
00:29:25,360 --> 00:29:28,200
Speaker 3: in the Best Connected Brewer. What does that phrase mean?

635
00:29:28,880 --> 00:29:29,080
Speaker 5: Yeah?

636
00:29:29,120 --> 00:29:30,880
Speaker 4: So I think it started with the ambition in twenty

637
00:29:30,880 --> 00:29:33,720
Speaker 4: twenty when we said we're going to digital transform. Remember

638
00:29:33,720 --> 00:29:36,440
Speaker 4: the pendulum I was talking about from gut page to

639
00:29:36,520 --> 00:29:39,240
Speaker 4: all the way to data driven And in today's world,

640
00:29:39,240 --> 00:29:41,719
Speaker 4: when you think of digital transmission, there are a few components,

641
00:29:41,800 --> 00:29:44,920
Speaker 4: cybersecurity being one of them, the digital core, like I

642
00:29:44,960 --> 00:29:47,760
Speaker 4: was saying, is one of them. Simplification and automation of

643
00:29:47,800 --> 00:29:50,520
Speaker 4: systems is one of them. Our breweries, how can we simplify?

644
00:29:50,760 --> 00:29:52,960
Speaker 4: And then comes data and AI, which is the really

645
00:29:53,000 --> 00:29:55,520
Speaker 4: one of the biggest components. And when you think of

646
00:29:55,560 --> 00:29:58,520
Speaker 4: best connected Brewer, the idea is we have been serving

647
00:29:58,560 --> 00:30:01,720
Speaker 4: our consumers and customers for one hundred and sixty two years.

648
00:30:02,360 --> 00:30:06,160
Speaker 4: What's different If you leverage tech into Bay's world. I

649
00:30:06,200 --> 00:30:09,400
Speaker 4: think you can really enhance the experience the customers have.

650
00:30:09,920 --> 00:30:12,360
Speaker 4: The example I was giving you earlier about the salesforce

651
00:30:12,400 --> 00:30:15,960
Speaker 4: going in different places and optimizing the rout that's a

652
00:30:16,000 --> 00:30:18,720
Speaker 4: good example why the relation is maintained just simply by

653
00:30:18,760 --> 00:30:21,680
Speaker 4: data driven insights. So if you can connect all the

654
00:30:21,720 --> 00:30:26,240
Speaker 4: different applications, all the platforms, remove fragmentation, scale very quick,

655
00:30:26,720 --> 00:30:30,680
Speaker 4: make sure your company is CyberSecure, things are simple and automated.

656
00:30:31,360 --> 00:30:34,680
Speaker 4: That's what we call the best connected brower. That's the ambition.

657
00:30:34,800 --> 00:30:38,280
Speaker 3: Actually, how do you measure this success of what you're doing?

658
00:30:38,760 --> 00:30:44,200
Speaker 3: Is do you expect that your efforts will have a

659
00:30:44,280 --> 00:30:46,880
Speaker 3: measurable and tangible effect on the bottom line of the company?

660
00:30:46,960 --> 00:30:52,240
Speaker 3: And can you actually figure out what the impact of

661
00:30:52,280 --> 00:30:52,960
Speaker 3: your efforts is?

662
00:30:53,440 --> 00:30:53,960
Speaker 5: Yeah, we do.

663
00:30:54,080 --> 00:30:57,000
Speaker 4: I think that is super important to measure because the

664
00:30:57,080 --> 00:31:00,120
Speaker 4: first one I was referring to proof of value and

665
00:31:00,160 --> 00:31:03,000
Speaker 4: I'm embedding some model, does it really work? So we

666
00:31:03,080 --> 00:31:06,000
Speaker 4: do a B testing, which is basically you keep aside

667
00:31:06,000 --> 00:31:08,560
Speaker 4: some sample and you actually launch the product on a

668
00:31:08,600 --> 00:31:10,640
Speaker 4: different sample, and you see the difference between the two.

669
00:31:11,120 --> 00:31:14,000
Speaker 4: The assumption is those that had the product and those

670
00:31:14,000 --> 00:31:16,440
Speaker 4: that didn't have the product, both of them went through

671
00:31:16,440 --> 00:31:20,200
Speaker 4: the same experiences because of market seasonal, etc. That's one

672
00:31:20,240 --> 00:31:22,440
Speaker 4: good way of doing it. And if you cannot have

673
00:31:22,520 --> 00:31:25,120
Speaker 4: the luxury sometimes of doing a B testing because everyone

674
00:31:25,200 --> 00:31:27,520
Speaker 4: is having high appetite, give me the product I don't

675
00:31:27,520 --> 00:31:30,040
Speaker 4: want to sit aside. Then you do some sort of

676
00:31:30,880 --> 00:31:33,040
Speaker 4: causal models like we say, so you kind of look

677
00:31:33,040 --> 00:31:36,040
Speaker 4: at what would have happened if the model was not there,

678
00:31:36,920 --> 00:31:39,560
Speaker 4: and then you predict that and since the model was there,

679
00:31:39,640 --> 00:31:42,120
Speaker 4: something else happened. The difference between the two is the

680
00:31:42,120 --> 00:31:45,080
Speaker 4: incremental value the model is creating. A B testing is

681
00:31:45,080 --> 00:31:47,880
Speaker 4: more accurate the causal models. The other one, like you said,

682
00:31:47,920 --> 00:31:50,760
Speaker 4: which called time series model, a little bit less accurate,

683
00:31:50,760 --> 00:31:53,560
Speaker 4: but directionally both give you the sense that yes, it's working.

684
00:31:53,960 --> 00:31:55,800
Speaker 3: What happens if you do a B testing or a

685
00:31:55,840 --> 00:31:58,600
Speaker 3: new idea and you don't see a difference.

686
00:31:59,480 --> 00:32:01,280
Speaker 4: In that case, we will move on to something else

687
00:32:01,320 --> 00:32:04,800
Speaker 4: because it means it's already optimal. Then we said, good,

688
00:32:05,240 --> 00:32:07,680
Speaker 4: check that now let's move on to something else. But

689
00:32:07,760 --> 00:32:10,080
Speaker 4: we need to just make sure that the process is

690
00:32:10,120 --> 00:32:11,320
Speaker 4: still running optimally.

691
00:32:11,360 --> 00:32:12,600
Speaker 5: So time to time, you keep.

692
00:32:12,480 --> 00:32:15,440
Speaker 4: Doing every testing anyway, every six months or whatever the

693
00:32:15,440 --> 00:32:17,960
Speaker 4: timeframe is. Yeah, just to make sure that it's still

694
00:32:18,040 --> 00:32:18,760
Speaker 4: still relevant.

695
00:32:18,840 --> 00:32:21,040
Speaker 3: But what if this is we're getting out on a

696
00:32:21,080 --> 00:32:23,680
Speaker 3: little bit of a digression here, but it's something I've

697
00:32:23,720 --> 00:32:27,040
Speaker 3: often thought about. What if the value that is being

698
00:32:27,080 --> 00:32:30,440
Speaker 3: created is not measurable? So I'll give you a dumb example.

699
00:32:30,680 --> 00:32:34,960
Speaker 3: When you were talking earlier about the salesman and giving

700
00:32:34,960 --> 00:32:38,200
Speaker 3: them a new, you know, better instructions about how to

701
00:32:38,200 --> 00:32:42,440
Speaker 3: basically spend their day, what if you tested that, discovered

702
00:32:42,440 --> 00:32:45,400
Speaker 3: it it didn't have any effect on the bottom line.

703
00:32:45,640 --> 00:32:48,680
Speaker 3: But in fact, what was happening was that the salesmen

704
00:32:48,720 --> 00:32:51,520
Speaker 3: were a lot happier with their jobs. And we're satisfied

705
00:32:52,400 --> 00:32:55,160
Speaker 3: and we're excited to come to work. Do you measure

706
00:32:55,320 --> 00:32:57,040
Speaker 3: something like that? Something intolerable?

707
00:32:57,160 --> 00:32:57,640
Speaker 5: Measure?

708
00:32:58,120 --> 00:33:00,720
Speaker 4: One way is NPS four, which I said promoter score.

709
00:33:00,760 --> 00:33:02,640
Speaker 4: Are you really happy with the product? Has it changed

710
00:33:02,640 --> 00:33:05,920
Speaker 4: your life? That gives you a good indication then, And

711
00:33:06,320 --> 00:33:08,160
Speaker 4: by the way, it's a numeric output, so it gives

712
00:33:08,200 --> 00:33:12,480
Speaker 4: you a score between minus hundred to plus hundred, and

713
00:33:12,560 --> 00:33:15,400
Speaker 4: sometimes it's not even tangible. Let's say we do something

714
00:33:15,440 --> 00:33:18,040
Speaker 4: for corporate affairs because they want to get external signals

715
00:33:18,040 --> 00:33:21,640
Speaker 4: of consumer insights and then just lean some information. Maybe

716
00:33:21,640 --> 00:33:24,200
Speaker 4: we act on it, maybe we don't, but this is

717
00:33:24,200 --> 00:33:26,160
Speaker 4: for a good cause. Sometimes you just want to study

718
00:33:26,160 --> 00:33:28,680
Speaker 4: the market. There's no immediate value if you don't create

719
00:33:28,720 --> 00:33:31,000
Speaker 4: a product out of it, or something to do with

720
00:33:31,120 --> 00:33:34,400
Speaker 4: legal If there's a reputational risk for Heineken, can I

721
00:33:34,440 --> 00:33:36,800
Speaker 4: extract some insight that will prevent us or create the

722
00:33:36,800 --> 00:33:41,200
Speaker 4: best briefing or summary or external briefing that using AI

723
00:33:41,280 --> 00:33:45,880
Speaker 4: that will help us protect ourselves. That's also reputational damage.

724
00:33:46,240 --> 00:33:49,440
Speaker 3: Last question before we go to questions. I'm curious when

725
00:33:49,440 --> 00:33:53,680
Speaker 3: you look at the very beginning you talked about this

726
00:33:53,920 --> 00:33:58,080
Speaker 3: linear value chain we're in that along that chain are

727
00:33:58,120 --> 00:34:01,320
Speaker 3: you having the most and where are you having the

728
00:34:01,440 --> 00:34:03,760
Speaker 3: least impact. I'm more interested in the second half of that.

729
00:34:04,040 --> 00:34:07,880
Speaker 4: Yeah, I think we covered few things, but one area

730
00:34:07,920 --> 00:34:10,000
Speaker 4: I think we can do more is really.

731
00:34:09,800 --> 00:34:11,440
Speaker 5: Understanding consumer sentiments.

732
00:34:12,280 --> 00:34:15,279
Speaker 4: And the reason for that is Heineken is people go

733
00:34:15,360 --> 00:34:17,520
Speaker 4: to the bars and outlets and you're not really leading

734
00:34:17,600 --> 00:34:20,680
Speaker 4: your first hand data there right, you're enjoying a beer,

735
00:34:20,719 --> 00:34:23,239
Speaker 4: then you walk away. I don't know exactly what you did.

736
00:34:23,640 --> 00:34:25,799
Speaker 4: I can get some aggregated data to make some sense

737
00:34:25,840 --> 00:34:29,000
Speaker 4: out of it. But if we can really get consumer

738
00:34:29,040 --> 00:34:32,040
Speaker 4: insights as to what the consumers like and dislike, what

739
00:34:32,160 --> 00:34:34,640
Speaker 4: sort of ad you like? How should I design my

740
00:34:34,719 --> 00:34:39,319
Speaker 4: Hanneken campaign so it resonates with a cluster of individuals.

741
00:34:39,800 --> 00:34:41,400
Speaker 4: That would be a little bit of holy grail as

742
00:34:41,440 --> 00:34:43,560
Speaker 4: the next step, like you were talking about two point zero,

743
00:34:44,160 --> 00:34:47,040
Speaker 4: and to get consumer insights first party data.

744
00:34:46,880 --> 00:34:47,800
Speaker 5: It's not super easy.

745
00:34:48,200 --> 00:34:51,120
Speaker 4: So what we are trying to do is create digital

746
00:34:51,160 --> 00:34:54,680
Speaker 4: twins of consumers. So at an aggregate level, they give

747
00:34:54,719 --> 00:34:57,279
Speaker 4: you a sense of Also with agent Kei, which is

748
00:34:57,280 --> 00:35:00,000
Speaker 4: also you hear a lot about to get a sense

749
00:35:00,160 --> 00:35:03,200
Speaker 4: of how consumers might react to certain campaign or certain product,

750
00:35:03,719 --> 00:35:05,640
Speaker 4: and that should give us quite a bit of insights

751
00:35:05,640 --> 00:35:07,759
Speaker 4: that right now we don't have access to. I think

752
00:35:07,800 --> 00:35:09,359
Speaker 4: that's one of the areas we could really do.

753
00:35:09,280 --> 00:35:10,319
Speaker 5: A lot more. Yeah.

754
00:35:10,560 --> 00:35:12,960
Speaker 3: Yeah, So if I said last question, if I sat

755
00:35:13,080 --> 00:35:16,399
Speaker 3: down with you, it's twenty twenty six. Now, we did

756
00:35:16,400 --> 00:35:19,319
Speaker 3: this over five years from now twenty thirty one, we're

757
00:35:19,320 --> 00:35:22,680
Speaker 3: sitting in this chair, tell me what the kind of

758
00:35:23,600 --> 00:35:25,280
Speaker 3: what's going to be the next big score.

759
00:35:26,120 --> 00:35:28,480
Speaker 4: I think one area will be how we make our

760
00:35:28,520 --> 00:35:31,799
Speaker 4: lives as employees are finding a lot easier. So the repetitive,

761
00:35:32,360 --> 00:35:35,719
Speaker 4: boring task, manual task. Can we automate those things and

762
00:35:35,840 --> 00:35:38,160
Speaker 4: just use the time to do something more creative and

763
00:35:38,239 --> 00:35:41,520
Speaker 4: think big about the business itself. That will be one area,

764
00:35:41,560 --> 00:35:44,160
Speaker 4: most on the productivity side. But the other area would

765
00:35:44,200 --> 00:35:46,960
Speaker 4: be Indeed, when we look at gen Z and this

766
00:35:47,080 --> 00:35:50,319
Speaker 4: is fact, I'm not saying something my own opinion, there's

767
00:35:50,360 --> 00:35:54,319
Speaker 4: a trend of distinct trend of alcohol as a beverage.

768
00:35:54,560 --> 00:35:57,080
Speaker 4: The consumption is on a decline. So then what's the

769
00:35:57,120 --> 00:36:01,680
Speaker 4: next best thing for the new generation? What will resonate?

770
00:36:02,320 --> 00:36:04,279
Speaker 4: Those are the pockets we need to find, and I

771
00:36:04,280 --> 00:36:07,120
Speaker 4: think that's where we will transition very quickly over the

772
00:36:07,120 --> 00:36:07,719
Speaker 4: next five years.

773
00:36:07,760 --> 00:36:08,920
Speaker 5: And if you get there, I think that will be

774
00:36:08,920 --> 00:36:09,520
Speaker 5: big success.

775
00:36:10,120 --> 00:36:16,399
Speaker 3: So you think that your specific department responsibility can help

776
00:36:16,440 --> 00:36:18,880
Speaker 3: the company in discovering what the answer to that question

777
00:36:18,960 --> 00:36:19,360
Speaker 3: is about.

778
00:36:19,600 --> 00:36:21,880
Speaker 4: Definitely, that's the ambition, that's what we're trying to do.

779
00:36:22,120 --> 00:36:23,880
Speaker 4: That's what we are really trying to get this one

780
00:36:23,960 --> 00:36:26,080
Speaker 4: over insights. I think that's the last mile. That's the

781
00:36:26,120 --> 00:36:27,600
Speaker 4: one part that is left.

782
00:36:28,320 --> 00:36:31,000
Speaker 3: Sergey, this has been fascinating. Thank you so much. I

783
00:36:31,040 --> 00:36:33,320
Speaker 3: should say thank you for a question, sir. My uncle

784
00:36:33,680 --> 00:36:37,080
Speaker 3: was a Heineken salesman in Jamaica. He was the local

785
00:36:37,120 --> 00:36:40,960
Speaker 3: distributor and I have so many child memories of going

786
00:36:41,000 --> 00:36:43,359
Speaker 3: to Jamaica and he would show up in his Heineken truck.

787
00:36:44,160 --> 00:36:50,680
Speaker 3: So we're resonating deep in my mimory with this conversation.

788
00:36:50,760 --> 00:36:52,480
Speaker 3: He would come and he would have a Heineken right

789
00:36:52,520 --> 00:36:54,520
Speaker 3: there on the on the table and would drink it

790
00:36:54,560 --> 00:36:57,000
Speaker 3: at the end of the day. But we have we

791
00:36:57,080 --> 00:37:00,440
Speaker 3: have a few moments for questions. They're all the screen

792
00:37:00,480 --> 00:37:02,319
Speaker 3: and I don't have my glasses. Can you read them?

793
00:37:02,440 --> 00:37:04,640
Speaker 5: Yeah, I can, I can read them. Then they should

794
00:37:04,680 --> 00:37:05,879
Speaker 5: be going order to the first one.

795
00:37:06,040 --> 00:37:08,799
Speaker 3: Yeah no, no, no, no no, no, Rookie air, never

796
00:37:08,880 --> 00:37:12,400
Speaker 3: do that. Okay, read the first four and pick the

797
00:37:12,400 --> 00:37:13,360
Speaker 3: one you want to answer.

798
00:37:14,800 --> 00:37:15,959
Speaker 5: Okay, go tip.

799
00:37:16,440 --> 00:37:18,120
Speaker 4: But I gave it away already, so I'm going to

800
00:37:18,200 --> 00:37:21,120
Speaker 4: now do what I said. No, I think the first

801
00:37:21,120 --> 00:37:23,399
Speaker 4: one is quite relevant. So it's a question for both

802
00:37:23,400 --> 00:37:26,480
Speaker 4: of us. If you were advising a twenty year old,

803
00:37:26,880 --> 00:37:29,840
Speaker 4: what three skills. Would you tell them to start developing

804
00:37:30,000 --> 00:37:33,480
Speaker 4: right now to stay relevant in an AI driven world?

805
00:37:34,400 --> 00:37:37,799
Speaker 3: Oh well, well you don't you have a twenty year

806
00:37:37,800 --> 00:37:39,160
Speaker 3: old or a near twenty year old. You have a

807
00:37:39,200 --> 00:37:39,919
Speaker 3: fifteen year old?

808
00:37:39,960 --> 00:37:41,319
Speaker 5: You told me I have a fifteen year old?

809
00:37:41,400 --> 00:37:42,600
Speaker 3: All right, what do you tell your son?

810
00:37:42,760 --> 00:37:44,120
Speaker 5: I thought you were going to answer them this first,

811
00:37:44,120 --> 00:37:45,480
Speaker 5: but my.

812
00:37:45,520 --> 00:37:48,040
Speaker 3: Kids are two and four. I tell them to put

813
00:37:48,080 --> 00:37:52,239
Speaker 3: away their toys. You this is more, you start more?

814
00:37:52,800 --> 00:37:55,000
Speaker 3: Do you have a is your fifteen year old son

815
00:37:55,080 --> 00:37:55,479
Speaker 3: or daughter?

816
00:37:55,600 --> 00:37:56,000
Speaker 5: His son?

817
00:37:56,239 --> 00:38:00,440
Speaker 4: And he's already thinker with AI. He's doing his own Python, etc.

818
00:38:00,719 --> 00:38:03,400
Speaker 4: Which I couldn't imagine when I was fifteen. So I

819
00:38:03,440 --> 00:38:06,640
Speaker 4: think I'll give a high level answer to be actually

820
00:38:06,680 --> 00:38:10,480
Speaker 4: successful depending on whether your hands on within AI, building

821
00:38:10,480 --> 00:38:13,160
Speaker 4: models yourself or not. There are three things I think

822
00:38:13,200 --> 00:38:17,160
Speaker 4: is super important. One is having that tech background, having

823
00:38:17,160 --> 00:38:19,319
Speaker 4: a common understanding of what AI really is.

824
00:38:19,400 --> 00:38:20,280
Speaker 5: It always helps.

825
00:38:20,520 --> 00:38:23,239
Speaker 4: Not everyone needs to have the details and algorithms and

826
00:38:23,280 --> 00:38:27,120
Speaker 4: how models work, not needed, but having that basic understanding

827
00:38:27,160 --> 00:38:29,520
Speaker 4: always is good. Then you know exactly how to gauge

828
00:38:29,640 --> 00:38:31,759
Speaker 4: what AI is really doing. And I think the other

829
00:38:31,800 --> 00:38:35,000
Speaker 4: thing is if you are in a corporate setting and

830
00:38:35,040 --> 00:38:37,839
Speaker 4: you are doing something for the business, work backwards from

831
00:38:37,840 --> 00:38:41,839
Speaker 4: the business and understand whatever you're building should actually touch

832
00:38:41,880 --> 00:38:44,279
Speaker 4: the business and make it beneficial for them. It's not

833
00:38:44,440 --> 00:38:46,680
Speaker 4: AI and modeling for the sake of it. That's for

834
00:38:46,680 --> 00:38:49,520
Speaker 4: a separate research and development. If you're in a corporate world,

835
00:38:49,719 --> 00:38:52,680
Speaker 4: try to build something beneficial for business. And I third

836
00:38:52,680 --> 00:38:54,719
Speaker 4: one I think which myself I learned quite a bit

837
00:38:54,760 --> 00:38:56,280
Speaker 4: in my in last six years.

838
00:38:56,920 --> 00:38:59,759
Speaker 5: It's communication. Talking about AI.

839
00:38:59,840 --> 00:39:02,840
Speaker 4: If you used a lot of tech jargon and mathematics,

840
00:39:03,320 --> 00:39:07,080
Speaker 4: sometimes people lose you. It's about how you really narrate

841
00:39:07,120 --> 00:39:09,440
Speaker 4: the story in a very simple way so people can

842
00:39:09,520 --> 00:39:11,839
Speaker 4: relate to it. I think if the combination of these

843
00:39:11,880 --> 00:39:14,480
Speaker 4: three has worked very well for me, so I can

844
00:39:14,800 --> 00:39:17,279
Speaker 4: say that anything you want to add.

845
00:39:17,480 --> 00:39:22,000
Speaker 3: It's funny because I I met this guy who's the

846
00:39:22,040 --> 00:39:25,480
Speaker 3: headmaster at a Jesuit school in Manhattan. We've been chatting

847
00:39:26,000 --> 00:39:28,120
Speaker 3: and I want to do a little program at his school,

848
00:39:28,760 --> 00:39:33,719
Speaker 3: and it's all about asking questions because we're now into

849
00:39:33,760 --> 00:39:36,000
Speaker 3: the era of asking questions right.

850
00:39:36,080 --> 00:39:36,560
Speaker 5: That's correct.

851
00:39:36,600 --> 00:39:38,560
Speaker 3: AI is this incredibly good tool, but you have to

852
00:39:39,280 --> 00:39:42,680
Speaker 3: ask the right questions. But this is not just true

853
00:39:42,680 --> 00:39:45,439
Speaker 3: of AI, but it's also true of the world we're

854
00:39:45,480 --> 00:39:49,480
Speaker 3: living in is a world that's so interconnected and everything

855
00:39:49,480 --> 00:39:53,640
Speaker 3: involves so many different people that your distinguishing feature in

856
00:39:53,680 --> 00:39:56,279
Speaker 3: many context is not where they're not the answers you have,

857
00:39:56,640 --> 00:39:58,239
Speaker 3: but the quality of the questions that you ask.

858
00:39:58,640 --> 00:40:00,759
Speaker 5: That's fantastic what you said. I fully avery.

859
00:40:00,800 --> 00:40:03,759
Speaker 4: I think it's about asking the right questions that really

860
00:40:03,800 --> 00:40:06,920
Speaker 4: tells you, you know, you're looking for that that unique

861
00:40:06,920 --> 00:40:07,960
Speaker 4: thing that that you're missing.

862
00:40:08,080 --> 00:40:09,839
Speaker 5: Yeah, I fully agree.

863
00:40:10,080 --> 00:40:14,120
Speaker 3: Maybe I'll advite you to this class and like you

864
00:40:14,200 --> 00:40:17,440
Speaker 3: have that kind of time on your hands, if you

865
00:40:17,480 --> 00:40:19,879
Speaker 3: brought if you brought up you know, Heineken for all

866
00:40:19,880 --> 00:40:23,080
Speaker 3: the kids in school, that would that would really.

867
00:40:22,920 --> 00:40:26,200
Speaker 5: Yeah, surely we have to build a special product for that.

868
00:40:26,360 --> 00:40:29,360
Speaker 5: Well let's see, all right, next question, let's.

869
00:40:29,160 --> 00:40:32,400
Speaker 4: Go to this one or Malcolm, is there a particular

870
00:40:32,480 --> 00:40:35,920
Speaker 4: AI capability you are each excited to explore?

871
00:40:37,239 --> 00:40:38,240
Speaker 3: That's for you, my friend.

872
00:40:39,719 --> 00:40:42,239
Speaker 4: I think in the in the short term, I'm really

873
00:40:42,239 --> 00:40:44,920
Speaker 4: looking forward to Agent K. I is hearing a lot

874
00:40:44,960 --> 00:40:48,400
Speaker 4: of noise and hype and there are a lot of

875
00:40:48,560 --> 00:40:51,000
Speaker 4: feedback that I'm getting from a lot of companies. Have

876
00:40:51,160 --> 00:40:54,320
Speaker 4: you really embedded Agent KI within your systems? There is

877
00:40:54,320 --> 00:40:56,640
Speaker 4: a very mixed feedback. Some say yes, some say no.

878
00:40:57,160 --> 00:40:59,000
Speaker 4: I think the potential of agent k I when we

879
00:40:59,000 --> 00:41:02,279
Speaker 4: look at this task we do day to day. Let

880
00:41:02,280 --> 00:41:05,640
Speaker 4: me gives an example, invoice management or transactional finance or

881
00:41:05,760 --> 00:41:08,879
Speaker 4: very repetitive task. If you can really automate augment those

882
00:41:08,880 --> 00:41:11,080
Speaker 4: things with agent KI, I think it's going to be

883
00:41:11,120 --> 00:41:13,400
Speaker 4: a game changer. If you free up thirty percent of

884
00:41:13,400 --> 00:41:16,960
Speaker 4: our time just by embedding these things, then I can

885
00:41:16,960 --> 00:41:20,680
Speaker 4: really think big. Everyone can think big. What's next? Then

886
00:41:20,719 --> 00:41:23,240
Speaker 4: the creativity comes in. Otherwise all day you are stuck

887
00:41:23,280 --> 00:41:26,000
Speaker 4: with the repetitive task. So I think that's what I'm

888
00:41:26,000 --> 00:41:28,320
Speaker 4: really looking forward to. And this is very short term.

889
00:41:28,400 --> 00:41:29,640
Speaker 4: Within the next few years.

890
00:41:30,000 --> 00:41:33,360
Speaker 3: Yeah, we have I think, what time for one more question,

891
00:41:34,760 --> 00:41:38,560
Speaker 3: Sergey go for let's see this is it's got to

892
00:41:38,640 --> 00:41:40,439
Speaker 3: be the last one always has to be the best one.

893
00:41:40,480 --> 00:41:45,600
Speaker 4: Certain let for people hearing the phrase for the first time,

894
00:41:45,719 --> 00:41:48,880
Speaker 4: what is the real example that shows Heineken being the

895
00:41:48,920 --> 00:41:52,600
Speaker 4: best connected broer? Basically you're asking for a proof point.

896
00:41:53,040 --> 00:41:55,640
Speaker 4: Are we really becoming the best connective brower when we

897
00:41:55,680 --> 00:41:58,640
Speaker 4: look at our markets. Heinek in Mexico is a very

898
00:41:58,640 --> 00:42:01,880
Speaker 4: good example across how value chain if you work backwards

899
00:42:01,880 --> 00:42:06,680
Speaker 4: from consumers, customers and so on. We have advertising optimization

900
00:42:06,760 --> 00:42:10,640
Speaker 4: for consumers. For customers, we have next best action. Actually

901
00:42:10,719 --> 00:42:14,319
Speaker 4: for customers we are pricing and promotion optimized. For the salesforce,

902
00:42:14,360 --> 00:42:16,880
Speaker 4: we have next best action. For the breweries, we have

903
00:42:16,960 --> 00:42:19,680
Speaker 4: connected brewery. We are getting signals from these machines and

904
00:42:19,680 --> 00:42:23,120
Speaker 4: optimizing them. I think it covers a significant portion of

905
00:42:23,120 --> 00:42:26,560
Speaker 4: a value chain that's fully automated end to end. So

906
00:42:26,600 --> 00:42:28,480
Speaker 4: that would be a good example where we really saw

907
00:42:28,520 --> 00:42:31,160
Speaker 4: the benefit of taking it to the next level when

908
00:42:31,160 --> 00:42:34,440
Speaker 4: it comes to automation. So Mexico Heineken Mexico is a

909
00:42:34,440 --> 00:42:35,040
Speaker 4: good example.

910
00:42:35,320 --> 00:42:38,000
Speaker 3: Thank you so much for joining us. Thank you to

911
00:42:38,080 --> 00:42:40,240
Speaker 3: all of you who came to listen.

912
00:42:40,640 --> 00:42:43,480
Speaker 5: Thanks for it. Asper, Thank you very much.

913
00:42:43,880 --> 00:42:48,000
Speaker 3: Yeah, that's it for the first episode of season seven

914
00:42:48,400 --> 00:42:51,759
Speaker 3: of Smart Talks with IBM, But stay tuned. There's so

915
00:42:51,880 --> 00:42:54,360
Speaker 3: much more to come this season as we die further

916
00:42:54,400 --> 00:42:59,360
Speaker 3: into how AI and Quantum Computing are creating smarter business.

917
00:43:00,600 --> 00:43:03,680
Speaker 3: Smart Talks with IBM is produced by Matt Ramano, Amy

918
00:43:03,680 --> 00:43:08,680
Speaker 3: Gains McQuaid, and Jake Harper. Engineering by Ninabird Lawrence, Mastering

919
00:43:08,719 --> 00:43:12,960
Speaker 3: by Sarah Buguer music by Gramoscope, Strategy by Cassidy Meyer

920
00:43:13,280 --> 00:43:17,920
Speaker 3: and Sophia Derlin. Special thanks to Sergei Ghosh and Michelle Ganji.

921
00:43:17,960 --> 00:43:21,440
Speaker 3: Post from the Heineken Company. Smart Talks with IBM is

922
00:43:21,440 --> 00:43:25,399
Speaker 3: a production of Pushkin Industries and Ruby Studio at iHeartMedia.

923
00:43:26,200 --> 00:43:30,120
Speaker 3: To find more Pushkin podcasts, listen on the iHeartRadio app,

924
00:43:30,360 --> 00:43:35,560
Speaker 3: Apple Podcasts, or wherever you listen to podcasts. I'm Malcolm Gladwell.

925
00:43:35,920 --> 00:43:39,440
Speaker 3: This is a paid advertisement from IBM. The conversations on

926
00:43:39,480 --> 00:43:50,840
Speaker 3: this podcast don't necessarily represent IBM's positions, strategies, or opinions.