WEBVTT
1
00:00:01.780 --> 00:00:02.613
(soft music)
2
00:00:02.613 --> 00:00:05.240
Is there really such a thing as an AI unicorn?
3
00:00:05.240 --> 00:00:06.200
There might not be,
4
00:00:06.200 --> 00:00:09.160
but for sure there are horses for courses.
5
00:00:09.160 --> 00:00:12.040
Find out more when we talk with Colin Lenaghan,
6
00:00:12.040 --> 00:00:14.120
Global Senior Vice President
7
00:00:14.120 --> 00:00:16.583
of Net Revenue Management at PepsiCo.
8
00:00:17.480 --> 00:00:19.470
Welcome to "Me, Myself, and AI,"
9
00:00:19.470 --> 00:00:22.710
a podcast on artificial intelligence in business.
10
00:00:22.710 --> 00:00:26.550
Each episode we introduce you to someone innovating with AI.
11
00:00:26.550 --> 00:00:29.410
I'm Sam Ransbotham, Professor of Information Systems
12
00:00:29.410 --> 00:00:31.090
at Boston College.
13
00:00:31.090 --> 00:00:32.510
I'm also the guest editor
14
00:00:32.510 --> 00:00:35.380
for the AI and Business Strategy Big Idea program
15
00:00:35.380 --> 00:00:38.170
at MIT Sloan Management Review.
16
00:00:38.170 --> 00:00:42.040
And I'm Shervin Khodabandeh, Senior Partner with BCG
17
00:00:42.040 --> 00:00:45.870
and I co-lead BCG's AI practice in North America.
18
00:00:45.870 --> 00:00:49.930
And together MIT SMR and BCG have been researching AI
19
00:00:49.930 --> 00:00:53.530
for five years, interviewing hundreds of practitioners
20
00:00:53.530 --> 00:00:56.260
and surveying thousands of companies on what it takes
21
00:00:56.260 --> 00:01:00.200
to build and to deploy and scale AI capabilities
22
00:01:00.200 --> 00:01:01.560
across the organization
23
00:01:01.560 --> 00:01:04.210
and really transform the way organizations operate.
24
00:01:06.140 --> 00:01:08.060
Today, we're talking with Colin Lenaghan.
25
00:01:08.060 --> 00:01:09.840
Colin is the Global Senior Vice President
26
00:01:09.840 --> 00:01:12.310
of Net Revenue Management at PepsiCo.
27
00:01:12.310 --> 00:01:15.240
Colin, thanks for dialing in today from the UK, welcome.
28
00:01:15.240 --> 00:01:16.220
It's my pleasure, Sam.
29
00:01:16.220 --> 00:01:18.360
It's an honor to be with you guys today.
30
00:01:18.360 --> 00:01:22.580
I'm excited to be in such esteemed company,
31
00:01:22.580 --> 00:01:24.640
really looking forward to this session.
32
00:01:24.640 --> 00:01:26.350
Yeah, we'll definitely keep the esteemed comment
33
00:01:26.350 --> 00:01:27.820
in there for sure.
34
00:01:27.820 --> 00:01:29.020
Colin, can you tell us a little bit
35
00:01:29.020 --> 00:01:31.310
about your current role at PepsiCo?
36
00:01:31.310 --> 00:01:33.590
My role, globally, very much
37
00:01:33.590 --> 00:01:36.080
is around building capabilities
38
00:01:36.080 --> 00:01:39.570
that help our business units win in the future.
39
00:01:39.570 --> 00:01:42.230
There's three major legs that I try to push,
40
00:01:42.230 --> 00:01:43.600
or that I am pushing.
41
00:01:43.600 --> 00:01:46.320
One is solidifying our foundation,
42
00:01:46.320 --> 00:01:49.390
that's like, have we got the right talent in the right roles
43
00:01:49.390 --> 00:01:52.580
at the right seniority to integrate
44
00:01:52.580 --> 00:01:54.420
all the different elements that have to come together
45
00:01:54.420 --> 00:01:55.961
for revenue management.
46
00:01:55.961 --> 00:01:59.750
The other leg then is are we building for future growth?
47
00:01:59.750 --> 00:02:02.320
So, if you take the potential of our brands
48
00:02:02.320 --> 00:02:04.960
and how we can position them with the consumer
49
00:02:04.960 --> 00:02:06.970
as it relates to pricing and promotion, et cetera,
50
00:02:06.970 --> 00:02:09.410
that's a huge unlock for PepsiCo.
51
00:02:09.410 --> 00:02:13.070
And then this whole third leg is this digital transformation
52
00:02:13.070 --> 00:02:13.917
that we're calling it.
53
00:02:13.917 --> 00:02:15.760
And that goes from,
54
00:02:15.760 --> 00:02:18.460
everything from standardizing the analytics
55
00:02:18.460 --> 00:02:21.930
that we want all our businesses to be looking at
56
00:02:21.930 --> 00:02:25.140
across the world to speed up diagnostic,
57
00:02:25.140 --> 00:02:28.080
to speed up decision-making, and solutioning,
58
00:02:28.080 --> 00:02:31.080
right through to this more advanced AI agenda,
59
00:02:31.080 --> 00:02:33.200
and we call that our AI program.
60
00:02:33.200 --> 00:02:35.390
So, it's those major three areas.
61
00:02:35.390 --> 00:02:38.130
And clearly we sort of orient it
62
00:02:38.130 --> 00:02:40.140
with what the business need is
63
00:02:40.140 --> 00:02:42.990
across the world and what are the capabilities
64
00:02:42.990 --> 00:02:45.280
that we need to help and support to deploy
65
00:02:45.280 --> 00:02:47.730
and clearly work to implement those
66
00:02:47.730 --> 00:02:50.153
and hopefully extract the value out of them.
67
00:02:51.210 --> 00:02:53.803
What is the scope of this that you're working on?
68
00:02:54.650 --> 00:02:57.540
The scope is right across the full spectrum
69
00:02:57.540 --> 00:03:01.350
of the PepsiCo portfolio, everything from beverages,
70
00:03:01.350 --> 00:03:04.500
to snacks, of course dairy.
71
00:03:04.500 --> 00:03:06.660
This is a classic capability that can help
72
00:03:06.660 --> 00:03:09.990
to solve some of the problems that we're facing.
73
00:03:09.990 --> 00:03:11.640
So I think the categories,
74
00:03:11.640 --> 00:03:13.640
we're almost agnostic on the category.
75
00:03:13.640 --> 00:03:16.840
If there's a really clear use case of where this capability
76
00:03:16.840 --> 00:03:19.210
can help us address a business problem,
77
00:03:19.210 --> 00:03:22.000
it really then can go across wherever we want.
78
00:03:22.000 --> 00:03:25.201
And I would see this applying as much to Quaker in the US
79
00:03:25.201 --> 00:03:30.110
as our Quaker business in China, or our beverage business
80
00:03:30.110 --> 00:03:33.840
in Latin America to the UK.
81
00:03:33.840 --> 00:03:35.170
It's everywhere.
82
00:03:35.170 --> 00:03:37.260
I think you've been at PepsiCo for 23 years
83
00:03:37.260 --> 00:03:39.120
if I remember right?
84
00:03:39.120 --> 00:03:41.420
I'm 23 odd years at PepsiCo,
85
00:03:41.420 --> 00:03:43.980
and I actually started what you might call
86
00:03:43.980 --> 00:03:47.610
a net revenue monitoring transformation in the UK eons ago
87
00:03:47.610 --> 00:03:49.013
it seems like now.
88
00:03:49.013 --> 00:03:52.530
And at that time we were building capabilities on systems
89
00:03:52.530 --> 00:03:55.780
but very much on the low, very much linear.
90
00:03:55.780 --> 00:03:58.560
I would say literally in the last 18 months,
91
00:03:58.560 --> 00:04:03.040
what we've had to adapt to has been very rapid indeed.
92
00:04:03.040 --> 00:04:06.740
PepsiCo clearly sees advanced analytics,
93
00:04:06.740 --> 00:04:09.640
AI, as a core to how we're going to have to operate
94
00:04:09.640 --> 00:04:11.030
in the future.
95
00:04:11.030 --> 00:04:14.460
So Colin, if you can, give us one specific example
96
00:04:14.460 --> 00:04:16.810
of a place you've used artificial intelligence.
97
00:04:16.810 --> 00:04:19.880
What's one project you're excited about?
98
00:04:19.880 --> 00:04:22.832
I think the example is very much related to
99
00:04:22.832 --> 00:04:27.450
pricing around how do we take very high level,
100
00:04:27.450 --> 00:04:31.130
very average pricing insights and transform that
101
00:04:31.130 --> 00:04:33.800
from you know, if I could give the example,
102
00:04:33.800 --> 00:04:37.210
operating with 60 elasticities that help you understand
103
00:04:37.210 --> 00:04:39.610
where your pricing opportunity is to 40,000.
104
00:04:39.610 --> 00:04:43.170
I mean, that's the sort of scale that you're getting to.
105
00:04:43.170 --> 00:04:45.560
And then the decomposition of that elasticity
106
00:04:45.560 --> 00:04:50.291
around what it drags and draws from across your portfolio,
107
00:04:50.291 --> 00:04:54.870
across the portfolio in one retailer versus another retailer
108
00:04:54.870 --> 00:04:58.960
really is sort of mind-blowing around what that can do.
109
00:04:58.960 --> 00:05:02.570
And that's a real life example of a product
110
00:05:02.570 --> 00:05:04.610
that we're scaling up as we speak.
111
00:05:04.610 --> 00:05:07.650
Capabilities like that give us insight
112
00:05:07.650 --> 00:05:10.480
around what the right bets to make are
113
00:05:10.480 --> 00:05:13.840
and how we can journey towards an end state
114
00:05:13.840 --> 00:05:16.040
around what that portfolio price architecture
115
00:05:16.040 --> 00:05:17.420
could look like.
116
00:05:17.420 --> 00:05:20.140
I think if we didn't have this type of capability,
117
00:05:20.140 --> 00:05:22.270
you would be flying quite blind.
118
00:05:22.270 --> 00:05:24.480
I think you'd be spending a lot of time with consumers
119
00:05:24.480 --> 00:05:26.150
trying to get answers out of them
120
00:05:26.150 --> 00:05:28.400
that they're maybe not equipped to give.
121
00:05:28.400 --> 00:05:30.790
And I think you'd be taking bets,
122
00:05:30.790 --> 00:05:32.260
and you'd be taking some risks,
123
00:05:32.260 --> 00:05:36.600
because as we know, pricing is the most powerful lever
124
00:05:36.600 --> 00:05:39.340
in a PnL without doubt, stronger than volumes,
125
00:05:39.340 --> 00:05:41.150
stronger than cogs, whatever.
126
00:05:41.150 --> 00:05:43.490
You get it right, it's amazing.
127
00:05:43.490 --> 00:05:45.330
If you get it wrong, it's devastating. (laughs)
128
00:05:45.330 --> 00:05:48.372
So across all the things we're trying to do,
129
00:05:48.372 --> 00:05:50.740
this capability on pricing for me
130
00:05:50.740 --> 00:05:53.300
is by far the most exciting.
131
00:05:53.300 --> 00:05:57.270
If we can get that deployed, let's say the 15 markets
132
00:05:57.270 --> 00:05:59.520
in a short period of time,
133
00:05:59.520 --> 00:06:02.120
I think that becomes really powerful
134
00:06:02.120 --> 00:06:04.360
in our ability to understand the marketplace
135
00:06:04.360 --> 00:06:07.760
and to candidly make sure we're giving the consumers
136
00:06:07.760 --> 00:06:08.593
what they want.
137
00:06:08.593 --> 00:06:09.610
Yeah, it makes sense.
138
00:06:09.610 --> 00:06:13.230
But going from something like 60 to 40,000,
139
00:06:13.230 --> 00:06:14.510
that seems daunting,
140
00:06:14.510 --> 00:06:16.140
but there must be a pretty big carrot out there
141
00:06:16.140 --> 00:06:18.340
for you to want to do that.
142
00:06:18.340 --> 00:06:20.040
I think once you begin to see---
143
00:06:20.040 --> 00:06:22.680
and of course the 40,000 means you take up certain pack,
144
00:06:22.680 --> 00:06:24.910
to a certain customer, to a certain geography, right?
145
00:06:24.910 --> 00:06:28.290
It's just a full deaveraging almost of your pricing.
146
00:06:28.290 --> 00:06:32.360
So that's how you get that level of granularity.
147
00:06:32.360 --> 00:06:35.140
I mean my mind has now moved on to this is now what
148
00:06:35.140 --> 00:06:36.320
is just going to happen everywhere,
149
00:06:36.320 --> 00:06:38.730
because why wouldn't we have this?
150
00:06:38.730 --> 00:06:42.840
If my partners in the data analytics team can tell me
151
00:06:42.840 --> 00:06:45.550
that we are industrializing this as we speak,
152
00:06:45.550 --> 00:06:46.930
you're going to be able to spin this off
153
00:06:46.930 --> 00:06:50.054
and build it in weeks versus months.
154
00:06:50.054 --> 00:06:52.950
It becomes a bit of a no-brainer
155
00:06:52.950 --> 00:06:55.630
as to why we shouldn't be considering that.
156
00:06:55.630 --> 00:06:57.063
I wanted to ask Colin,
157
00:06:57.063 --> 00:06:59.730
what's the best part of your job?
158
00:06:59.730 --> 00:07:01.300
The best part of my job,
159
00:07:01.300 --> 00:07:04.610
it's almost a little bit why we're doing this podcast.
160
00:07:04.610 --> 00:07:07.570
I honestly learn every day, and that's not a cliche
161
00:07:07.570 --> 00:07:08.630
or thing that I'm just saying.
162
00:07:08.630 --> 00:07:09.657
I mean, I keep saying to people,
163
00:07:09.657 --> 00:07:12.157
"I still get the butterflies every Monday morning,
164
00:07:12.157 --> 00:07:15.132
where I think, crikey, what's going to happen this week,
165
00:07:15.132 --> 00:07:16.127
(laughing)
166
00:07:16.127 --> 00:07:17.780
and what am I going to have to learn?"
167
00:07:17.780 --> 00:07:21.200
And also, candidly, we're building a capability
168
00:07:21.200 --> 00:07:24.180
with PepsiCo that's relatively young.
169
00:07:24.180 --> 00:07:27.890
And I take great hope that there's a legacy there
170
00:07:27.890 --> 00:07:28.890
that I could be proud of
171
00:07:28.890 --> 00:07:31.110
and then I've made a small contribution
172
00:07:31.110 --> 00:07:32.211
to helping the business.
173
00:07:32.211 --> 00:07:34.520
That's what really excites me about this role
174
00:07:34.520 --> 00:07:36.040
and as ever with PepsiCo.
175
00:07:36.040 --> 00:07:37.400
I'm going to give you a chance to redirect though,
176
00:07:37.400 --> 00:07:39.350
from crikey, what do I have to learn,
177
00:07:39.350 --> 00:07:41.540
to crikey, what do I get to learn?
178
00:07:41.540 --> 00:07:43.570
No, honestly, it's absolutely true.
179
00:07:43.570 --> 00:07:44.403
It's absolutely true,
180
00:07:44.403 --> 00:07:46.227
because these things are coming at me now, right?
181
00:07:46.227 --> 00:07:49.270
I have to embrace them. I have to learn.
182
00:07:49.270 --> 00:07:51.340
I have to be super uncomfortable.
183
00:07:51.340 --> 00:07:53.627
And honestly, I just go with it saying,
184
00:07:53.627 --> 00:07:55.210
"Look, this is making me uncomfortable."
185
00:07:55.210 --> 00:07:59.090
And that's probably a good sign rather than a bad sign.
186
00:07:59.090 --> 00:08:01.200
I am going through a personal journey with AI,
187
00:08:01.200 --> 00:08:05.170
and it's real, and it's eventful, shall we say.
188
00:08:05.170 --> 00:08:06.608
So how are you doing that learning?
189
00:08:06.608 --> 00:08:09.041
I'm sure you're a little further on the maturity curve
190
00:08:09.041 --> 00:08:10.610
than you're admitting here.
191
00:08:10.610 --> 00:08:13.540
I think you're being a bit modest with your background.
192
00:08:13.540 --> 00:08:15.040
I think there's a couple of levels here.
193
00:08:15.040 --> 00:08:18.730
One, PepsiCo's very much an organization and a culture
194
00:08:18.730 --> 00:08:22.020
that learns by doing, so I think we're very targeted
195
00:08:22.020 --> 00:08:24.390
on key use cases where we see value
196
00:08:24.390 --> 00:08:26.140
for this type of capability.
197
00:08:26.140 --> 00:08:29.890
I think we're operating very collaboratively with agility
198
00:08:29.890 --> 00:08:32.420
across how we get those use cases developed,
199
00:08:32.420 --> 00:08:35.170
how we prove them out and then how we begin to scale them.
200
00:08:35.170 --> 00:08:37.540
But then broader, I would say PepsiCo
201
00:08:37.540 --> 00:08:41.820
is making quite an investment in just bringing literacy
202
00:08:41.820 --> 00:08:44.580
of advanced analytics up across the broader community
203
00:08:44.580 --> 00:08:46.325
starting with the senior management.
204
00:08:46.325 --> 00:08:48.250
This is clearly something
205
00:08:48.250 --> 00:08:49.760
that has to be driven from the top.
206
00:08:49.760 --> 00:08:51.750
It needs cultural change.
207
00:08:51.750 --> 00:08:53.620
And so we are starting to do a lot of work
208
00:08:53.620 --> 00:08:57.420
in elevating that literacy, really getting the understanding
209
00:08:57.420 --> 00:09:00.405
of what this is, what do I need to know
210
00:09:00.405 --> 00:09:04.110
that is enough for me to embrace and leverage the capability
211
00:09:04.110 --> 00:09:05.940
versus be the expert, right?
212
00:09:05.940 --> 00:09:08.150
I mean, obviously we're hiring a lot of amazing people
213
00:09:08.150 --> 00:09:09.880
to help us do that and partnering
214
00:09:09.880 --> 00:09:12.420
with a lot of great parties outside of PepsiCo
215
00:09:12.420 --> 00:09:13.810
to help us do that.
216
00:09:13.810 --> 00:09:16.070
I would say that's a big initiative right now
217
00:09:16.070 --> 00:09:19.280
as we start to build more and more use cases
218
00:09:19.280 --> 00:09:22.450
and start to build more products as we call them,
219
00:09:22.450 --> 00:09:24.750
then how we operate with that,
220
00:09:24.750 --> 00:09:26.810
and candidly, how we learned to live with that
221
00:09:26.810 --> 00:09:28.320
is a big opportunity.
222
00:09:28.320 --> 00:09:29.680
And I would say we're all learning.
223
00:09:29.680 --> 00:09:32.609
I can speak for myself for sure.
224
00:09:32.609 --> 00:09:34.020
The problem with all learning
225
00:09:34.020 --> 00:09:35.283
is that everybody has to learn.
226
00:09:35.283 --> 00:09:37.680
I mean, you mentioned that the senior management,
227
00:09:37.680 --> 00:09:39.480
how do you get, for example, those people
228
00:09:39.480 --> 00:09:42.840
to learn something that they too didn't grow up knowing?
229
00:09:42.840 --> 00:09:44.615
I think there's a lot of factors
230
00:09:44.615 --> 00:09:45.548
that come to that,
231
00:09:45.548 --> 00:09:48.020
one being very overt about it
232
00:09:48.020 --> 00:09:50.600
and taking time to make it important.
233
00:09:50.600 --> 00:09:53.878
The tone from the top, I think is very important in PepsiCo.
234
00:09:53.878 --> 00:09:55.700
We're all in on this one.
235
00:09:55.700 --> 00:09:58.168
I think this is one where it's kind of table stakes.
236
00:09:58.168 --> 00:09:59.330
(laughs) There's no debate.
237
00:09:59.330 --> 00:10:01.930
You have to be really leaning in heavily here
238
00:10:01.930 --> 00:10:03.580
if you're going to want to compete
239
00:10:03.580 --> 00:10:07.510
in the future CPG industry and broader.
240
00:10:07.510 --> 00:10:09.830
So, I think it's a combination of factors,
241
00:10:09.830 --> 00:10:12.660
and I don't foresee it being an easy exercise
242
00:10:12.660 --> 00:10:14.993
or something that we're just going to walk up
243
00:10:14.993 --> 00:10:15.826
and do a course
244
00:10:15.826 --> 00:10:18.690
and hey, we're all sort of literate.
245
00:10:18.690 --> 00:10:22.240
I feel this is going to be such a cultural process,
246
00:10:22.240 --> 00:10:25.980
even how we're working and doing things is much more fluid,
247
00:10:25.980 --> 00:10:30.458
much more a process and an evolution.
248
00:10:30.458 --> 00:10:34.010
You have to learn to really go in a very fluid way here,
249
00:10:34.010 --> 00:10:36.050
and there's a lot of zigzag left and right.
250
00:10:36.050 --> 00:10:38.320
There's a lot of two steps forward, one step back
251
00:10:38.320 --> 00:10:41.030
as you're experimenting and trying to get this capability
252
00:10:41.030 --> 00:10:42.530
to do what you want it to do.
253
00:10:42.530 --> 00:10:44.040
So I have to say personally,
254
00:10:44.040 --> 00:10:46.785
I feel quite uncomfortable at times.
255
00:10:46.785 --> 00:10:51.080
I feel I'm not in control, but then I've just got to look at
256
00:10:51.080 --> 00:10:54.590
and sort of engage with the team that we're working with,
257
00:10:54.590 --> 00:10:56.600
and they kind of know what they're doing,
258
00:10:56.600 --> 00:10:59.010
so (laughs) you have to have a lot of trust
259
00:10:59.010 --> 00:11:00.070
in this thing around that.
260
00:11:00.070 --> 00:11:02.070
It's something that we're building for the long term.
261
00:11:02.070 --> 00:11:04.960
And I think that long-term perspective
262
00:11:04.960 --> 00:11:06.510
is also very important.
263
00:11:06.510 --> 00:11:10.320
I think another point related to how we're going to learn,
264
00:11:10.320 --> 00:11:13.810
I think seeing this not as a very short-term fix,
265
00:11:13.810 --> 00:11:16.953
that it solves an immediate problem in 2021.
266
00:11:16.953 --> 00:11:20.260
This is going to just be capability that morphs
267
00:11:20.260 --> 00:11:23.320
into helping us solve lots of problems in the longer term.
268
00:11:23.320 --> 00:11:25.050
And so I think that's why we view it
269
00:11:25.050 --> 00:11:27.030
as a very strategic capability,
270
00:11:27.030 --> 00:11:29.420
helping us to solve strategic problems
271
00:11:29.420 --> 00:11:32.330
that hopefully over time, we improve, we enrich,
272
00:11:32.330 --> 00:11:34.880
we get better, and we strengthen the capability
273
00:11:34.880 --> 00:11:37.400
and hopefully the culture at the same time.
274
00:11:37.400 --> 00:11:39.470
Colin, I want to build on a point you made
275
00:11:39.470 --> 00:11:43.120
about the process not being set in stone
276
00:11:43.120 --> 00:11:46.126
and sort of the process itself is ever changing,
277
00:11:46.126 --> 00:11:49.210
and sort of the new way of working between all the parties
278
00:11:49.210 --> 00:11:50.450
that are involved.
279
00:11:50.450 --> 00:11:53.180
What do you think the role of yourself as a leader
280
00:11:53.180 --> 00:11:55.200
and other senior management is
281
00:11:55.200 --> 00:12:00.200
in giving comfort and stability in a situation
282
00:12:00.570 --> 00:12:02.070
like as you said yourself,
283
00:12:02.070 --> 00:12:05.330
like you yourself aren't always comfortable with it.
284
00:12:05.330 --> 00:12:07.850
How do you guys deal with that kind of a,
285
00:12:07.850 --> 00:12:09.473
I would say a culture shift?
286
00:12:10.580 --> 00:12:14.740
Like many large organizations, proving stuff out
287
00:12:14.740 --> 00:12:19.630
and very overtly socializing that clearly builds confidence,
288
00:12:19.630 --> 00:12:21.320
builds trust, and helps you to move
289
00:12:21.320 --> 00:12:22.950
from one step to the next.
290
00:12:22.950 --> 00:12:25.360
Another aspect of what we're trying to do with AI
291
00:12:25.360 --> 00:12:28.185
is the whole promotional optimization piece.
292
00:12:28.185 --> 00:12:30.820
If you take the results that we're starting to see
293
00:12:30.820 --> 00:12:32.520
from these capabilities,
294
00:12:32.520 --> 00:12:36.200
it immediately inspires trust and credibility.
295
00:12:36.200 --> 00:12:39.290
And therefore that's a really important process for us,
296
00:12:39.290 --> 00:12:42.382
because you're starting small, you're building confidence.
297
00:12:42.382 --> 00:12:45.790
And I think that's going to be important because
298
00:12:45.790 --> 00:12:48.810
if you really took on quite a large, big bang approach
299
00:12:48.810 --> 00:12:51.223
with this thing, I just think it's overwhelming.
300
00:12:52.175 --> 00:12:57.175
I'm acclimatized or acclimated as you guys in the US say,
301
00:12:57.280 --> 00:13:01.165
but I think we're very much into prove things out,
302
00:13:01.165 --> 00:13:04.080
embrace of course, and be very aggressive
303
00:13:04.080 --> 00:13:06.840
in what we're trying to do, but make sure we take the steps
304
00:13:06.840 --> 00:13:09.200
to take the organization with us, to build trust
305
00:13:09.200 --> 00:13:11.870
and evolve the culture in that way.
306
00:13:11.870 --> 00:13:15.140
This sort of process that's not necessarily
307
00:13:15.140 --> 00:13:17.470
always set in stone, you know,
308
00:13:17.470 --> 00:13:20.450
adaptability is part and parcel
309
00:13:20.450 --> 00:13:24.080
of getting these things to scale, right?
310
00:13:24.080 --> 00:13:26.800
For sure, this capability needs to trickle down
311
00:13:26.800 --> 00:13:29.260
into the hands and minds of people
312
00:13:29.260 --> 00:13:31.660
who do the day-to-day operations.
313
00:13:31.660 --> 00:13:33.630
That's where the real hard work is.
314
00:13:33.630 --> 00:13:36.050
So as we put task forces together to get,
315
00:13:36.050 --> 00:13:37.920
for example, an MVP up and running,
316
00:13:37.920 --> 00:13:39.950
that's quite exciting, right, because everyone's there.
317
00:13:39.950 --> 00:13:42.787
It's a project, but then you go, "Okay, so, wow,
318
00:13:42.787 --> 00:13:44.720
that MVP worked really well."
319
00:13:44.720 --> 00:13:45.770
Now we're scaling this thing.
320
00:13:45.770 --> 00:13:48.780
We're taking it broad-based, maybe to a category
321
00:13:48.780 --> 00:13:52.740
within a market or a customer or whatever the case may be.
322
00:13:52.740 --> 00:13:55.380
You very rapidly get from a very excited project team
323
00:13:55.380 --> 00:13:58.100
to saying, "Hey, I got to use this thing."
324
00:13:58.100 --> 00:14:01.660
And then I think you're into the classic potential areas
325
00:14:01.660 --> 00:14:05.160
where you run the risk of it becoming tactical
326
00:14:05.160 --> 00:14:07.460
versus it being incredibly strategic.
327
00:14:07.460 --> 00:14:09.360
Do people really understand their role,
328
00:14:09.360 --> 00:14:11.290
how they're supposed to embrace it?
329
00:14:11.290 --> 00:14:14.490
Those, I think for me and my agenda
330
00:14:14.490 --> 00:14:16.500
that I have within PepsiCo,
331
00:14:16.500 --> 00:14:19.510
those are areas that are very top-of-my-mind right now
332
00:14:19.510 --> 00:14:22.170
by okay, how are we really going to operationalize this
333
00:14:22.170 --> 00:14:23.600
day to day?
334
00:14:23.600 --> 00:14:25.200
Well, let's build on that.
335
00:14:25.200 --> 00:14:27.380
You mentioned more of the strategic nature of this.
336
00:14:27.380 --> 00:14:29.940
Some of these things seem like small changes.
337
00:14:29.940 --> 00:14:32.620
How do you keep this collection of small changes
338
00:14:32.620 --> 00:14:33.640
from ending up in a place
339
00:14:33.640 --> 00:14:37.240
that isn't where your overall strategic focus should be?
340
00:14:37.240 --> 00:14:40.210
I guess, how do you balance that set of small changes
341
00:14:40.210 --> 00:14:43.870
and making incremental improvements with not just ending up
342
00:14:43.870 --> 00:14:45.490
at a place that's slightly better,
343
00:14:45.490 --> 00:14:46.403
but wanting to end up at a place
344
00:14:46.403 --> 00:14:48.800
that is strategically better?
345
00:14:48.800 --> 00:14:49.790
Yeah, that's a great question,
346
00:14:49.790 --> 00:14:52.020
because you can very much in the day-to-day
347
00:14:52.020 --> 00:14:53.950
of these projects and what you're trying to do,
348
00:14:53.950 --> 00:14:55.970
get lost in that, right?
349
00:14:55.970 --> 00:14:57.440
I think the use cases though
350
00:14:57.440 --> 00:14:59.710
that we're trying to drive at and in PepsiCo
351
00:14:59.710 --> 00:15:03.570
in my space, those are very much around big levers
352
00:15:03.570 --> 00:15:06.210
for how we accelerate the top-line growth
353
00:15:06.210 --> 00:15:08.120
and improve the quality of our growth.
354
00:15:08.120 --> 00:15:11.403
So pricing is super strategic.
355
00:15:11.403 --> 00:15:14.709
That's not going to go tactical anytime soon.
356
00:15:14.709 --> 00:15:18.400
Promotional investment and using promotions
357
00:15:18.400 --> 00:15:21.387
to underpin category strategy and what we're trying to do
358
00:15:21.387 --> 00:15:25.570
with the joint value creation with our retailers,
359
00:15:25.570 --> 00:15:26.910
that is super strategic,
360
00:15:26.910 --> 00:15:29.370
and that's not going to go away anytime soon.
361
00:15:29.370 --> 00:15:33.418
And ultimately when you think about from the user,
362
00:15:33.418 --> 00:15:37.360
is it always going to be efficient for these capabilities
363
00:15:37.360 --> 00:15:39.350
to be in verticals and in silos
364
00:15:39.350 --> 00:15:41.550
or at some point, are they all going to come together?
365
00:15:41.550 --> 00:15:46.280
So the revenue monitor in the field has all these integrated
366
00:15:46.280 --> 00:15:49.240
at his or her fingertips that allows them
367
00:15:49.240 --> 00:15:52.290
to plan strategically, leveraging this capability.
368
00:15:52.290 --> 00:15:53.740
And that's a little bit of the vision
369
00:15:53.740 --> 00:15:55.540
of where we might want to be going
370
00:15:55.540 --> 00:15:56.860
in terms of these capabilities.
371
00:15:56.860 --> 00:15:59.460
So this is very long term.
372
00:15:59.460 --> 00:16:01.660
I think it's a little bit of, you know, the Irish saying
373
00:16:01.660 --> 00:16:03.730
that we have here around "horses for courses",
374
00:16:03.730 --> 00:16:05.570
and I think we're quite pragmatic about that.
375
00:16:05.570 --> 00:16:07.810
But when you see the potential of it
376
00:16:07.810 --> 00:16:11.580
and what I'm fully expecting is that more and more parties
377
00:16:11.580 --> 00:16:15.410
begin to embrace this, it then becomes almost an expectation
378
00:16:15.410 --> 00:16:17.070
on both sides that we're going to have to operate
379
00:16:17.070 --> 00:16:18.870
with these types of capabilities.
380
00:16:18.870 --> 00:16:21.910
But if I take how we're trying to set these things up,
381
00:16:21.910 --> 00:16:23.630
what we want them to achieve,
382
00:16:23.630 --> 00:16:25.790
could there be some longer-term vision
383
00:16:25.790 --> 00:16:27.980
of a more integrated solution underpinned
384
00:16:27.980 --> 00:16:31.157
with this connectivity and AI, which it can do.
385
00:16:31.157 --> 00:16:33.781
That keeps certainly what I'm trying to drive,
386
00:16:33.781 --> 00:16:36.000
super strategic, I think.
387
00:16:36.000 --> 00:16:38.230
As you think about talent,
388
00:16:38.230 --> 00:16:39.310
I'm assuming those are probably
389
00:16:39.310 --> 00:16:41.480
the three most important dimensions, technical,
390
00:16:41.480 --> 00:16:44.200
organizational, and commercial.
391
00:16:44.200 --> 00:16:48.130
You're finding that there is sort of unicorns out there
392
00:16:48.130 --> 00:16:50.370
that have enough of all three
393
00:16:50.370 --> 00:16:53.470
or you're finding that it's required for everybody
394
00:16:53.470 --> 00:16:56.780
to have some of all three, but some would spike in some
395
00:16:56.780 --> 00:17:01.070
or others, or could it work where you've got talent
396
00:17:01.070 --> 00:17:04.030
that's commercially and organizationally savvy on one side,
397
00:17:04.030 --> 00:17:07.010
and then talent that's technically savvy on the other,
398
00:17:07.010 --> 00:17:09.830
and combined they sort of cover everything.
399
00:17:09.830 --> 00:17:10.740
I'm trying to get a sense
400
00:17:10.740 --> 00:17:14.300
of how much an individual contributor has to partake
401
00:17:14.300 --> 00:17:17.290
of all those three main components.
402
00:17:17.290 --> 00:17:18.123
It's a great question.
403
00:17:18.123 --> 00:17:20.770
I think it also depends where you sit in the organization.
404
00:17:20.770 --> 00:17:23.480
So if you take globally, candidly,
405
00:17:23.480 --> 00:17:27.130
I need more distinction in the capability set.
406
00:17:27.130 --> 00:17:30.300
So I need a classic strategy type individual
407
00:17:30.300 --> 00:17:32.120
to drive that strategy agenda.
408
00:17:32.120 --> 00:17:34.150
I need a capability expert,
409
00:17:34.150 --> 00:17:36.280
and then I need a digital tech expert,
410
00:17:36.280 --> 00:17:38.460
because we're building capabilities.
411
00:17:38.460 --> 00:17:40.033
The more you go down in the organization,
412
00:17:40.033 --> 00:17:43.840
the more you see the need for the blend to come together
413
00:17:43.840 --> 00:17:45.800
in the operating markets.
414
00:17:45.800 --> 00:17:48.694
And you don't see too many unicorns,
415
00:17:48.694 --> 00:17:52.670
and I'm not sure you need to have the unicorn,
416
00:17:52.670 --> 00:17:56.210
because ultimately in the business,
417
00:17:56.210 --> 00:17:58.380
it's around how do you use this to commercialize
418
00:17:58.380 --> 00:18:00.490
and drive the business impact?
419
00:18:00.490 --> 00:18:01.323
Well, that's good to hear,
420
00:18:01.323 --> 00:18:03.070
because I think it's tough to find those people
421
00:18:03.070 --> 00:18:04.000
who can do everything.
422
00:18:04.000 --> 00:18:06.350
And it sounds great on paper to find someone
423
00:18:06.350 --> 00:18:07.300
who does all those things,
424
00:18:07.300 --> 00:18:11.070
but the reality is we call them unicorns for a reason.
425
00:18:11.070 --> 00:18:13.003
It seems like Colin's saying
426
00:18:13.003 --> 00:18:14.068
you could build a unicorn.
427
00:18:14.068 --> 00:18:14.901
(laughing)
428
00:18:14.901 --> 00:18:15.820
by having the parts.
429
00:18:15.820 --> 00:18:17.850
Yeah, you've got the parts.
430
00:18:17.850 --> 00:18:19.948
Assemble them internally together
431
00:18:19.948 --> 00:18:22.710
and create the organizational unicorn
432
00:18:22.710 --> 00:18:24.330
rather than individual unicorn.
433
00:18:24.330 --> 00:18:25.230
That's right.
434
00:18:25.230 --> 00:18:28.650
Is there a specific example of an area
435
00:18:28.650 --> 00:18:33.350
where you guys put a solution in place
436
00:18:33.350 --> 00:18:37.070
and the organization went through that learning
437
00:18:37.070 --> 00:18:39.670
and that understanding that you can talk about?
438
00:18:39.670 --> 00:18:42.540
Yeah, we're live in a particular market
439
00:18:42.540 --> 00:18:45.550
with one of those products that we have,
440
00:18:45.550 --> 00:18:47.480
and this capability is very much
441
00:18:47.480 --> 00:18:52.480
around how we predict a different shape
442
00:18:52.510 --> 00:18:56.030
of promotional calendar against the shared objectives
443
00:18:56.030 --> 00:18:58.450
between a retailer and ourselves.
444
00:18:58.450 --> 00:19:01.040
So it's very much a strategic capability.
445
00:19:01.040 --> 00:19:03.270
So for example, we might want to agree
446
00:19:03.270 --> 00:19:06.380
that given the profile of the shoppers
447
00:19:06.380 --> 00:19:08.810
and the agenda of the retailer,
448
00:19:08.810 --> 00:19:11.500
they want to accelerate their top-line growth.
449
00:19:11.500 --> 00:19:13.100
In another retailer that could be the need
450
00:19:13.100 --> 00:19:15.520
to manage their margins a bit more.
451
00:19:15.520 --> 00:19:18.740
And that's taking a full category view
452
00:19:18.740 --> 00:19:21.580
of how this capability can help them to do that.
453
00:19:21.580 --> 00:19:24.267
And it's not what you might refer to
454
00:19:24.267 --> 00:19:27.130
as common, backward looking tools.
455
00:19:27.130 --> 00:19:30.620
This is very much around forward-looking capability.
456
00:19:30.620 --> 00:19:34.870
Against this objective we and the retailer have set,
457
00:19:34.870 --> 00:19:38.440
this is the shape of the calendar that needs to change
458
00:19:38.440 --> 00:19:39.780
for us to achieve that.
459
00:19:39.780 --> 00:19:42.270
I remember we went through a great process
460
00:19:42.270 --> 00:19:45.331
through the MVP of, you know, we probably took a month
461
00:19:45.331 --> 00:19:49.085
in virtual rooms or whatever,
462
00:19:49.085 --> 00:19:52.120
bringing all sorts of all their third-party data
463
00:19:52.120 --> 00:19:54.430
like panel data, like ROI data
464
00:19:54.430 --> 00:19:56.820
to validate this crazy hypothesis
465
00:19:56.820 --> 00:19:58.980
that this algorithm was suggesting.
466
00:19:58.980 --> 00:20:00.310
Like, there's no way that would work
467
00:20:00.310 --> 00:20:03.390
on this particular part of the category.
468
00:20:03.390 --> 00:20:05.390
And we managed to get confidence
469
00:20:05.390 --> 00:20:07.240
that what the capability was suggesting
470
00:20:07.240 --> 00:20:09.430
for execution could work.
471
00:20:09.430 --> 00:20:11.930
And low and behold, when we put it into the market,
472
00:20:11.930 --> 00:20:14.408
it was in the stores, hundreds and hundreds of stores.
473
00:20:14.408 --> 00:20:15.550
It worked.
474
00:20:15.550 --> 00:20:17.950
And this is a little bit about what we talked about before,
475
00:20:17.950 --> 00:20:19.940
around proving that use case site,
476
00:20:19.940 --> 00:20:23.340
showing how it can predict different scenarios
477
00:20:23.340 --> 00:20:26.397
that maybe we would be comfortable with.
478
00:20:26.397 --> 00:20:28.720
And that then builds confidence for us to keep this
479
00:20:28.720 --> 00:20:30.920
and use this as a more strategic capability.
480
00:20:32.030 --> 00:20:33.950
Well Colin, many thanks for your time today.
481
00:20:33.950 --> 00:20:36.220
You've brought out a lot of things
482
00:20:36.220 --> 00:20:39.240
that many managers are going to feel like are urgent
483
00:20:39.240 --> 00:20:42.410
and are important really, for everyone,
484
00:20:42.410 --> 00:20:44.580
not just revenue management, not just snacks,
485
00:20:44.580 --> 00:20:46.960
not just sodas, but I think managers everywhere.
486
00:20:46.960 --> 00:20:49.230
Thanks for taking the time to talk with us today.
487
00:20:49.230 --> 00:20:50.537
Thank you very much.
488
00:20:50.537 --> 00:20:53.120
(bright music)
489
00:20:56.290 --> 00:20:58.140
Shervin, Colin covered quite a few points.
490
00:20:58.140 --> 00:21:00.220
What struck you as particularly interesting?
491
00:21:00.220 --> 00:21:02.880
I liked his comment about horses for courses,
492
00:21:02.880 --> 00:21:07.114
talking about how the new reality that it creates
493
00:21:07.114 --> 00:21:09.580
in terms of how people will use it,
494
00:21:09.580 --> 00:21:14.504
how the big network of retailers and the whole ecosystem
495
00:21:14.504 --> 00:21:16.990
is adapting to these new ways of working
496
00:21:16.990 --> 00:21:20.450
and how even internally within PepsiCo,
497
00:21:20.450 --> 00:21:23.250
teams are doing things in different ways
498
00:21:23.250 --> 00:21:28.250
in more innovative, less static, more dynamic,
499
00:21:28.360 --> 00:21:29.820
and test-and-learn ways.
500
00:21:29.820 --> 00:21:33.690
I liked how he talked about this is going to be our future.
501
00:21:33.690 --> 00:21:36.870
We all know that, and it's not going to be a one-time thing.
502
00:21:36.870 --> 00:21:38.610
It's horses for courses,
503
00:21:38.610 --> 00:21:41.610
and that's how the new course is going to be.
504
00:21:41.610 --> 00:21:43.560
Definitely, horses for courses makes sense,
505
00:21:43.560 --> 00:21:45.510
because there isn't just one horse
506
00:21:45.510 --> 00:21:47.920
that's excellent at every racetrack.
507
00:21:47.920 --> 00:21:50.280
Similarly, I'm sure that companies would love
508
00:21:50.280 --> 00:21:53.930
to find an AI unicorn who was somehow magically good
509
00:21:53.930 --> 00:21:57.250
at everything AI-related, but it just seems unlikely.
510
00:21:57.250 --> 00:21:58.700
You know, they're not going to find someone
511
00:21:58.700 --> 00:22:02.190
who are ML coding experts, who can talk fluently
512
00:22:02.190 --> 00:22:04.720
with business managers and can communicate
513
00:22:04.720 --> 00:22:08.060
and present perfectly, and still somehow willing to work
514
00:22:08.060 --> 00:22:09.423
at entry-level salary.
515
00:22:10.320 --> 00:22:11.840
Instead, what Colin was talking about,
516
00:22:11.840 --> 00:22:12.840
and like others are saying,
517
00:22:12.840 --> 00:22:15.500
that success is a team composition
518
00:22:15.500 --> 00:22:17.864
where different people bring different skills.
519
00:22:17.864 --> 00:22:20.000
And even if some companies
520
00:22:20.000 --> 00:22:23.270
somehow could find a magical AI unicorn today,
521
00:22:23.270 --> 00:22:25.070
even that unicorn would have to learn
522
00:22:25.070 --> 00:22:27.035
as technologies change so quickly.
523
00:22:27.035 --> 00:22:29.140
Yeah, the other thing I really liked
524
00:22:29.140 --> 00:22:31.560
that he brought into the dialogue,
525
00:22:31.560 --> 00:22:34.240
well first of all, the best part of my job
526
00:22:34.240 --> 00:22:36.547
is waking up every morning and saying,
527
00:22:36.547 --> 00:22:37.847
"What am I going learn today?
528
00:22:37.847 --> 00:22:39.057
And I'm looking forward to it,
529
00:22:39.057 --> 00:22:42.660
having been at PepsiCo for 24, 25 years."
530
00:22:42.660 --> 00:22:46.150
And I thought that's really, really encouraging
531
00:22:46.150 --> 00:22:48.629
and energizing coming from a leader.
532
00:22:48.629 --> 00:22:49.780
That was a bit of a contrast too.
533
00:22:49.780 --> 00:22:51.270
In some of our other episodes,
534
00:22:51.270 --> 00:22:54.150
like with Walmart and Mehrotra,
535
00:22:54.150 --> 00:22:56.520
like these people have gone through multiple jobs
536
00:22:56.520 --> 00:22:59.740
and relatively quickly, moving from one technology space
537
00:22:59.740 --> 00:23:03.010
to another technology space, applying job learnings
538
00:23:03.010 --> 00:23:05.530
from different areas to their current role.
539
00:23:05.530 --> 00:23:07.370
Colin was a very different story.
540
00:23:07.370 --> 00:23:10.501
His own story about how he got to be in his role
541
00:23:10.501 --> 00:23:12.017
was quite different.
542
00:23:12.017 --> 00:23:14.640
He's learned while he's been there,
543
00:23:14.640 --> 00:23:16.940
not brought learning from other places.
544
00:23:16.940 --> 00:23:19.864
Yep, and I think that's actually quite valuable
545
00:23:19.864 --> 00:23:23.480
in a place like PepsiCo, and it fits quite well
546
00:23:23.480 --> 00:23:26.980
to the point he made about what it takes
547
00:23:26.980 --> 00:23:29.080
to get these things to scale.
548
00:23:29.080 --> 00:23:32.070
It is about culture change, organizational change,
549
00:23:32.070 --> 00:23:34.920
and also commercial focus for value.
550
00:23:34.920 --> 00:23:38.426
And having been in an organization long enough
551
00:23:38.426 --> 00:23:41.080
to know what it takes to create that change,
552
00:23:41.080 --> 00:23:42.480
I think really works for him.
553
00:23:42.480 --> 00:23:44.510
I could have imagined if he came from outside
554
00:23:44.510 --> 00:23:48.390
with series of very, very successful stents
555
00:23:48.390 --> 00:23:50.520
in highly digital companies,
556
00:23:50.520 --> 00:23:53.201
it might've been a very different trajectory for them.
557
00:23:53.201 --> 00:23:54.659
So the other part there too,
558
00:23:54.659 --> 00:23:55.600
is that with that ecosystem,
559
00:23:55.600 --> 00:23:57.460
Colin mentioned the butterflies that he got every day.
560
00:23:57.460 --> 00:23:59.130
But there's a lot of butterflies
561
00:23:59.130 --> 00:24:00.380
in a lot of different people there
562
00:24:00.380 --> 00:24:03.240
and getting that trust, so people are comfortable
563
00:24:03.240 --> 00:24:07.120
with those butterflies, hard job and not a technical job.
564
00:24:07.120 --> 00:24:08.848
No, it is a hard job.
565
00:24:08.848 --> 00:24:09.840
(bright music)
566
00:24:09.840 --> 00:24:11.000
Thanks for joining us.
567
00:24:11.000 --> 00:24:13.670
Next time, in our last episode of season two,
568
00:24:13.670 --> 00:24:16.390
we'll talk with Elizabeth Renieris, Founding Director
569
00:24:16.390 --> 00:24:19.210
of the IBM Notre Dame Technology Ethics Lab.
570
00:24:19.210 --> 00:24:20.043
Please join us.
571
00:24:21.100 --> 00:24:23.880
Thanks for listening to "Me, Myself, and AI."
572
00:24:23.880 --> 00:24:25.500
If you're enjoying the show,
573
00:24:25.500 --> 00:24:27.310
take a minute to write us a review.
574
00:24:27.310 --> 00:24:30.470
If you send us a screenshot, we'll send you a collection
575
00:24:30.470 --> 00:24:34.160
of MIT SMR's best articles on artificial intelligence,
576
00:24:34.160 --> 00:24:36.060
free for a limited time.
577
00:24:36.060 --> 00:24:40.823
Send your review screenshot to SMRfeedback@mit.edu.