0:00

Welcome to The Difference Engine, the

0:02

show for founders, funders and the

0:04

category curious. Don't confuse

0:06

size, don't confuse valuation with category

0:08

leadership. I'm not the only person frustrated

0:11

by this. You disagree with my

0:13

analysis. You either acquire

0:16

or you are acquired. Imitation is the

0:18

sincerest form of flattery. And it's proof

0:20

that you're living the odd. We

0:22

all know history is written by the winners. Hello

0:28

and welcome to The Difference Engine. Today

0:31

we're turning our attention towards West

0:33

London and some big and we

0:35

mean big tech investments. And

0:38

we'll be looking at the players

0:40

in the great AI category race.

0:42

What does the future look like

0:44

for them? We'll also be teaching you

0:46

how to make your ecosystem blossom

0:48

and sharing some lessons from category king

0:51

Mark Benioff. But first, Starlink has a

0:53

so-called Neobank pivoted its way to

0:55

a category lead. So

1:00

Starlink is a Neobank. Yes. And,

1:03

oh gosh, at least 10 years

1:05

ago, we were knocking out copy for companies

1:07

looking to target the bank saying, don't worry,

1:09

big high street banks, this lot will go

1:11

away and they won't worry. They've stuck around

1:14

a while. But some of

1:16

them are making some interesting moves, particularly Starlink

1:19

with its estimable CEO

1:22

who's just moving on and handing

1:24

the baton over. This is

1:26

a company that's had an idea,

1:28

a category like idea. Well, yes,

1:30

its category idea was that it's

1:33

put all its technology into a

1:35

subsidiary called Engine. And we

1:37

think that this will be the

1:39

engine of Starlink's growth in the

1:41

future. And it's not without precedent

1:43

because those familiar with

1:45

the UK shopping scene will know

1:48

Ocado. It's an online supermarket and

1:50

this online supermarket has done exactly

1:52

the same. It appears to be

1:55

increasing in demonstrator for the Ocado

1:57

smart platform with its unvogue

2:00

and highly valued combination

2:03

of artificial intelligence robotics and automation this

2:05

may be a case where the category

2:08

move and falls off the time so.

2:11

Obviously starting bank started this is the new

2:13

type of banking you know slightly better bank

2:15

with and boughton at the helm a car

2:17

when it started off lots of

2:19

people for the car there was just

2:21

the online branch of waitrose no you

2:24

had a deal with waitrose that deal

2:26

expired famously lots of people rip them

2:28

off and they then start another

2:30

deal with marks and spencer and I believe they're

2:32

also got deals with morris and say now they

2:34

are becoming what they probably always wanted to become

2:37

which is a platform but the real platform

2:39

is the technology the sundry for sure.

2:42

I'm much the same thing is happening now

2:44

with stalling it's got a

2:47

profitable i understand business yeah i would

2:49

like and by because she drove that

2:51

thing to profitability very very early on

2:53

the most of the competitors are still

2:55

bleeding money look over here look over

2:57

here i'm coming at you know with

3:00

a platform this is a much smarter

3:02

move yeah and assault banking remainer of

3:04

bought it and amp in australia so

3:06

they're running on the engine platform. And

3:09

of course this means that selling don't have to

3:11

try and get new banking licenses in other

3:13

countries and they don't have to do

3:15

the long haul of building a trusted

3:17

brand but they're still building revenues now

3:19

that's sort of important at the moment

3:21

because they're starting to hint at an

3:23

IPO. And we

3:26

both know that the value in that

3:28

is going to come from the engine

3:30

part of the business not the starting

3:32

bank part of the business is interesting

3:34

you pick up those two aspects of

3:36

banks so this is what really makes

3:38

a bank different from some of the

3:40

other pseudo that's called nokolomneo banks to

3:42

the maxi don't have licenses particular

3:44

new caram looking at you Revolut

3:46

and other people they have some operating

3:48

license they do not have banking licenses

3:50

the two things you picked out there

3:52

about the banks the stalling's working with

3:54

our brands which is very hard to

3:56

build up it certainly takes a couple

3:58

years and as you see. say

4:00

the license. So we always were saying to those banks,

4:03

the sort of defender banks, the high street

4:05

banks, you've got a couple of things that

4:07

the neo guys want. And now

4:10

it's all turned on its head where you've got a former

4:12

neo bank, Starling, going

4:14

out and grabbing the IP assets

4:16

of the brand, and the

4:18

customer base and the license.

4:20

Yeah. For their own. But

4:23

it's sort of interesting. I think the

4:25

investment community gets this and gets where

4:27

the value is. This might want to

4:29

have a look at the London Stock

4:32

Exchange's website where a condo is quoted.

4:34

And if you look at it, its

4:36

classification is consumer staples, not terribly exciting.

4:38

But if you look at the banner

4:40

on the LSE's webpage, it's got a

4:43

picture of a card as demonstration website.

4:45

So it's just a massive robots. Let's

4:48

look at stock exchanges a lot. But

4:52

they could have put a very consumer

4:54

friendly smiling, a car delivery person and

4:56

the lovely electric van, but they didn't.

4:58

They went straight into the tech. That's

5:00

where value is. And I think there's

5:03

another piece here, right? So the classic

5:05

cliche about the internet.com boom, which I

5:07

lived through, we both lived through was

5:09

that the money was made by Microsoft

5:11

and Cisco. They were the guys selling

5:14

the quote unquote picks and shovels. Yeah.

5:16

They were the Levi Strauss

5:18

of the era. And as Warren said, when

5:20

all the water went out, those guys still

5:22

had a business. And it's

5:24

amazing how category design, it does

5:27

need to often pivot. It's

5:30

not the first time. So you think about

5:32

slack, which started off as a

5:34

video game chat app from from

5:37

memory. Minecraft, which was

5:39

called cave game, I think, hence the mind

5:41

only occurred to me that mine and cave.

5:44

And they added survival to the building skills,

5:46

which made it much more competitive. And of

5:48

course, AWS, what did that go out of

5:52

Amazon's ability to build huge infrastructure for

5:54

this global which in turn came out

5:56

of the ambition to be the bookstore

5:58

for the world, right? started off

6:00

as a online bookshop. So your first

6:02

shot may not be where you end

6:04

up. We haven't been party to the

6:06

internal conversations at Staling but it looks

6:08

a lot like Ocado and Ocado we

6:10

know because those guys are also Westland

6:12

and were very intentional in the way

6:15

that they went after the opportunity and

6:17

they took some short-term pain particularly around

6:19

the waitrose loss to

6:21

get the long-term gain of being the

6:23

leader in the category that they set

6:25

out to do. It's fascinating that that's

6:28

essentially going for value from a B2C

6:30

proposition to a B2B proposition.

6:32

So going from selling to

6:35

you and I getting our

6:37

groceries to the really

6:40

deep infrastructure specialists who are

6:42

putting together the next generation

6:44

of support for

6:46

e-commerce and Staling Bank. I

6:48

think from selling to banking

6:50

consumers like us to the banking infrastructure

6:52

specialists. I mean it can't hurt that

6:55

consumers are aware of the brand. But

6:57

if it puts a nice pattern, we

7:00

put on our own name on the technology,

7:02

Staling Bank, Ocado and now we're going to

7:04

go do the B2B, the heavy lift the

7:06

sales where the money's really made at the

7:08

back end. We're doing that with a brand

7:10

that people know. So it's sort of interesting

7:13

case study in that respect but we do

7:15

wish the guys at Staling all

7:17

the very very best. I'm personally a user

7:20

as I am of Ocado and I really

7:22

like the way that these guys have sort

7:24

of pivoted to where they

7:26

always wanted to be which is a

7:28

very strong B2B category. Yeah we hope

7:30

that bodes really well for any future

7:33

IPO. So from both of us, good

7:35

luck Staling Bank. More

7:37

on the blog. I've

7:46

got a question for Paul here. Paul, what

7:49

is a flat? Oh at

7:51

the risk of appearing prudish, I'm afraid

7:53

I don't know. Well okay, let me

7:55

tell you. It's Frankfurt, London, Amsterdam and

7:57

Paris and it's used to describe Who's

8:00

who in terms of data center investment

8:02

there and apparently good old uk is

8:05

leaving it why because we've

8:07

apparently attained leadership in public cloud

8:09

infrastructure i hear so i know

8:11

what you're talking about now right

8:13

so microsoft in a slow news

8:15

week, decided to announce that it's

8:17

going to invest in very hopefully for the

8:19

uk government two point five billion in

8:22

a data center district very close to here

8:24

this is a local news story we're talking

8:26

about probably. Two miles that way

8:28

and actin yeah you might

8:31

think oh what what news is

8:33

that because in that particular district

8:35

we've already got to equinix data

8:37

sites our options cult of

8:40

air as well and advantage.

8:43

I just said it's going to put another data

8:45

center and so what's special about microsoft pumping some

8:47

money into the world i think the key is

8:49

northwest one of the key is a right it's

8:51

all very well putting a dumb ass

8:53

data center anywhere you want to

8:55

flap or not. But if you're

8:57

talking and they are talking about

8:59

twenty thousand of these super rare

9:02

ai GPUs by twenty twenty six

9:04

that's a hell of a lot of money. It is

9:06

it is a hell of a lot of money and again

9:09

it might just be a look

9:12

at us i've got a new data center and

9:14

it's a bit special because we've got the ai

9:16

GPUs in there. But it's a little bit more

9:18

than that they're already showing that they intend to

9:20

build an ecosystem around this and a lot of

9:22

this investment money is going to do

9:24

that precisely. Yeah i think the

9:26

other thing about any sort of data center investment

9:28

whilst it may not bring a lot of jobs

9:30

so it might not help mr sunak the government

9:33

in that way it does bring expertise

9:35

and it brings experts who want

9:37

to be close physically close to

9:40

where the computer. Yeah but you know it's

9:42

part of this investment and they are wrapping

9:45

an ecosystem around this great bit of real

9:47

estate i mean that they're going to apparently

9:49

train one million people with

9:51

the skills you need to work in ai.

9:53

I think that's something i take with it

9:55

with a pinch of salt but there's no

9:57

doubt that microsoft is absolutely killing it. terms

10:00

of AI, the co-pilot language that they use

10:02

is being adopted left and right. But it

10:04

all starts with spades in the ground,

10:06

shovels in the earth, whatever you want to call it, and

10:09

building data centers. Yeah. But the issue

10:11

is, this is a classic category play.

10:13

It's not just another data center. This

10:16

moves everything forward. In terms

10:18

of, in terms of the fact it's not

10:20

just done processing

10:22

being attached to dump pipes, this is enabling

10:24

an entirely new paradigm to be created. Right.

10:26

And you know, you can get that wrong.

10:29

I don't know if you probably will remember

10:31

when the dark fiber sort of

10:33

gold rush was on and everyone was digging holes

10:35

in streets and filling them with optical cables. Well,

10:38

they never got lit up. So that

10:40

never took off. But I think you'd be a very

10:42

brave person to bet against what's going to happen with

10:44

AI and 2.5 billion is not

10:46

an inconsiderable sum of money. Yeah. But

10:48

this whole district is absolutely full of processing

10:51

power. The key thing is, are

10:53

the other providers going to start following Microsoft

10:55

and actually start equipping their data centers to

10:57

be a bit more, you can't

10:59

be a bit more AI specific, but having

11:02

more AI capacity. Or just fold under, right?

11:04

Like just use the compute power that's there.

11:07

Go for the whole Microsoft story. Maybe

11:09

build some federated AI type technologies on

11:11

top of Azure. That's really the game

11:13

here. It's like, you want to be

11:15

the person with, yes, the most capacity,

11:17

but also the most convincing story that

11:19

you're going to lead. And you know,

11:21

what you can do is you can

11:23

frighten your competition into submission. You can't

11:26

do that in every category. This one

11:28

happens to involve, you know, hard stuff

11:30

like actual bare metal. But if

11:32

you get it right, you just demonstrate to the

11:34

market that there's nobody else going to lead this

11:36

category. And by default, you

11:38

become the category leader. Still a lot

11:40

to play, but I do think

11:42

it's a fascinating move and very, very good

11:45

for West London. Yeah. Classic category play. Congratulations

11:47

to Microsoft on their vision and commitment on

11:49

this one. So red

11:52

lights on cranes in West London, Microsoft chucking some

11:54

cash down. What's our lesson here from a category

11:56

point of view? Do it seriously. Build yourself an

11:58

ecosystem. This isn't just about building. an

12:00

AI specific data center. It's about building

12:03

a whole ecosystem around it and that

12:05

really throws the gauntlet down not just

12:07

to other data center suppliers but to

12:09

everybody who wants to drive the next

12:12

paradigm in computing based around AI. You've

12:21

got to learn to earn.

12:25

This week we're going to give you a

12:27

few pro tips on what an ecosystem is.

12:29

Once again an ecosystem is a key component

12:31

of the category design process alongside blueprints and

12:34

points of view. So why should

12:36

you care about ecosystems? Well it's one of

12:38

the most important things that

12:40

you can do to help build value

12:43

for the customers that your category has

12:45

been created to serve but

12:47

you know this ecosystem has to deliver

12:49

value to those people who are with

12:51

you on the journey in delivering it

12:53

as well. Right and it's

12:55

fair to say in this world of digital

12:57

products that virtually nothing is vertically

13:00

integrated. I you're not mining a product

13:02

from the ground, refining it, manufacturing something

13:05

with it and selling it. It's always

13:07

almost always a mishmash of various contributors

13:09

and so the ecosystem is that ecosystem

13:12

that puts together the offer that allows

13:14

you to dominate your category. Yeah it's

13:16

a mutual life support network if you

13:19

want to think of it that way

13:22

but we also think that actually

13:24

the customers themselves particularly they're really

13:26

innovative customers are a vital component

13:28

of the ecosystem because it's them

13:31

that helps your category stick. Right

13:33

and sidebar on analysts because we've

13:36

had a annoying another annoying conversation

13:38

about analysts is like the analysts

13:41

ain't that fundamental to your ecosystem what

13:43

you said about customers much more important.

13:45

Oh totally I mean I you know

13:47

I think the analysts support

13:49

the premise of the ecosystem but ultimately

13:51

they're not a critical component part of

13:53

it. Alright let's get into it so

13:56

the ecosystem is a word

13:58

for a natural environment. I mean like it's a way chosen

14:00

word. We don't talk about a network, we

14:02

talk about an ecosystem. Do you want to

14:04

talk a little bit about how apt that

14:06

is? Yeah, I mean it's about being mutually

14:08

supportive but it's also critical that it evolves

14:10

over time. Just like an actual

14:13

world, you think? Exactly. And one change will

14:15

percolate through that whole ecosystem.

14:18

But let's just try and

14:20

dive down into a little bit more. Ecosystems

14:23

are built of organisations or

14:25

organisms in the natural world

14:27

that have a vested interest

14:29

in each other's success. So

14:32

often those organisations that

14:34

are going to be in your ecosystem

14:36

are partners and if you

14:39

think about partners, particularly if you're building

14:41

a tech ecosystem,

14:44

there's really sort of two types of

14:46

partners to think about. The one I

14:48

think of as transactional partnerships, you know,

14:50

their channels are distribution ultimately.

14:53

And the thing about channels is

14:55

that not many people know this.

14:59

About 75% of world trade

15:01

is conducted through partnerships. And

15:03

that's about $5 trillion

15:07

in businesses and governments spending

15:09

on tech each year. And I

15:12

say three-quarters of that will go through

15:14

partners. Then there's another slightly

15:17

more nebulous but just as important

15:19

type of partner which are the

15:21

non-transactional partnerships. And they're companies and

15:23

organisations and even individuals who add

15:26

value during or after

15:28

the transaction but don't collect any

15:31

money on behalf of a vendor.

15:33

And there could be a myriad

15:35

of technology, strategic, analytical or business

15:38

alliances in there. The

15:40

reality is that a lot of

15:42

organisations are both types at once.

15:45

They may be adding a whole lot of value that

15:48

may not have money instantly collected

15:51

from it but their partnership

15:53

will add huge strategic value to their

15:55

particular offer as part of the ecosystem.

15:58

Right. And I think it is worth saying also that as

16:00

you begin your category journey, there may be people

16:03

in the ecosystem that drop off. Most

16:06

more than likely are these 71% distribution folks

16:08

that you cannot get to today. You're not

16:10

big enough, you're not important enough, you haven't

16:12

made the case for the category, no one's

16:15

seeing it the way you're seeing it because

16:17

you're designing the category and they will need

16:19

to be co-opted. But it could be that

16:21

that's some time in the future but

16:23

you should conceive of it as early as

16:25

possible. And your first partners in your ecosystem

16:28

are going to be the most important ones

16:30

because their success as part of the ecosystem

16:32

will encourage other people to join. Now

16:35

to the point you've just made, when

16:37

you're building an ecosystem at

16:39

the beginning this is all part of

16:42

your category blueprint which

16:44

we've talked about at length and

16:47

the people are going to be with you at

16:49

the beginning of the journey and not the people

16:52

doing business as usual. Because remember this

16:54

is all part of doing business not as

16:57

usual. So let me just capture this, so you're saying

16:59

that some of the people in the ecosystem early doors

17:02

in the category could be very

17:04

left of center, what's

17:06

the relationship but they've got a joint

17:08

enterprise if you will with your category

17:10

because they stand to gain a lot

17:12

from it even though that may not

17:14

be obvious even to them at

17:16

the start. They are not wedded to an

17:19

existing ecosystem, they're not wedded to the status

17:21

quo or even if they are they've got

17:23

a profound reason to move things forward. So

17:25

that's what you're looking for, you're looking for

17:27

the organization or the individual

17:29

that wants to drive change. And it

17:32

may be two steps to get where

17:34

you want. One engagement I'm thinking of

17:36

in particular was where our

17:38

client desperately wanted to have somebody in their

17:40

ecosystem that was a global SI, they

17:43

know that the distribution that you talked about

17:45

before will get them to lots of enterprise deals

17:47

that they couldn't get on their own. But they

17:49

couldn't go straight at them, we had to make

17:52

a step stone move with somebody

17:54

else with the deliberate intention of

17:56

getting some attention on their products

17:58

so that the big. guys could say, maybe

18:01

there's something in this. But I believe the phrase is

18:03

a Sprat to catch a mackerel. I thought it was

18:05

a Sprat to catch a whale, but same thing. Well,

18:08

you're just more ambitious than these some times. Yeah, you

18:10

can tell. But I think the point

18:12

here is that what you're

18:14

looking for with that strategy is to get

18:16

a real sense of fear of missing out

18:19

of FOMO. Yes. Amongst the bigger players who

18:21

will suddenly move

18:23

with you. And that's when your ecosystem development

18:25

becomes exponential. Contrast all of that with the

18:27

ambition that too many people have these days,

18:30

which is these are my distribution partners. These

18:32

are the guys that I need to get

18:34

to. And if I don't get to them,

18:36

I won't have achieved anything. This is much

18:38

bigger thinking, right? Yeah. And when we're talking

18:41

about bigger thinking, those early ecosystem partners has

18:44

to be prepared to go with you on that category

18:46

journey. You know, they have to

18:48

be prepared to take part in validating strikes,

18:50

for instance. They will be an important part

18:52

of that and in vandalizing that category in

18:55

their sales, marketing and any other influencing activity

18:57

they're doing. So you have to choose your

18:59

ecosystem partners very, very carefully. But it is

19:02

very frightening to me how many times you

19:04

express that you want somebody to join your

19:06

ecosystem. They're not there now. And

19:08

there's this sort of trepidation, fear, I

19:11

would say, of going to just ask

19:14

before the question gets asked, shall we try and

19:16

partner with somebody? Often it's like,

19:18

well, they will never partner with us. Like you've

19:20

given up before you've even started. And that an

19:22

ambitious view of the ecosystem is just doomed

19:25

to failure. Well, I think you should ask yourself a

19:27

very tough question if you're not prepared to go out

19:29

to a potential partner, explain what it is you want

19:31

to do on your category journey. Probably

19:33

means your point of view isn't

19:36

hard and fast enough at that stage. And

19:39

it really, really should be. You should be

19:41

totally confident in taking that point of view

19:43

out to market. It may just be that

19:45

you may get rejected from people who are

19:47

in existing ecosystems simply because they cannot move

19:50

fast enough or it's not in their interest

19:52

to move. They might absolutely get what it

19:54

is you're trying to do, but it's just

19:56

not the right time. Right. So just summarising

19:59

some of this. conversation about ecosystem

20:01

then. So what we're saying is be

20:03

extremely ambitious in your ecosystem. Yeah,

20:05

make sure they're aligned with your point of view.

20:08

Go for the randos that the random sort

20:10

of folks you think may not go along

20:12

with it. Ask as many people as possible.

20:15

Because it's part of building the category,

20:17

right? And the ecosystem partners are

20:20

amplifying your category. That's the big prize. That's

20:22

what you want. That's what you want. And

20:24

they're not going to do that if they're

20:26

massively invested in the current status quo, or

20:28

I'm not going to do it immediately. So

20:31

don't take rejection as a rejection per se

20:33

of your category

20:35

idea. They will be rejecting

20:37

a partnership with you for loads of other

20:39

reasons that are not to do with your

20:41

category vision. So it's all about

20:44

sourcing those sort of left of center,

20:46

stupidly ambitious, equally

20:48

bought into your POV as you are type

20:51

partners. That's the ideal ecosystem. Yeah, they're your

20:53

partners in the new world you're going to

20:55

create. What

21:04

does the future? Let's look

21:06

into our crystal ball. There's

21:09

a lot of confusion about AI. Let's look once again

21:11

into our crystal ball, and see what

21:13

we can predict is going to happen as

21:15

we emerge from this Cambrian explosion. So I'm

21:17

going to defer to somebody who knows a

21:19

lot more than me on this topic. What's

21:22

the Cambrian explosion, Jonathan? Well,

21:25

I do have to go back to my university

21:27

days on this as an undergraduate. And I'm

21:29

sure that from those days as a

21:32

geology student, I learned that the Cambrian

21:34

explosion was one of the very earliest

21:36

florings of life on Earth. Well,

21:38

it's a trick question because I wasn't talking

21:40

about the Cambrian explosion. I was talking about

21:43

the Cambrian explosion of AI companies

21:45

that everyone's expecting is going to

21:47

happen in the new year. So

21:49

we've got LF Alpha, 500

21:52

tasty million euros. That's got our friends, Cavalry

21:55

Ventures invested in that and Schwartz Group, who

21:57

are the guys behind one of the big

21:59

German which

24:01

is going to be this Cambrian explosion of AI,

24:04

how do you take that and use it to your advantage?

24:06

What are you driving at here in terms of the naming? Yeah,

24:09

so obviously your category name, the fight for

24:11

the language is on. We talked

24:13

in another section about the use

24:15

of the word copilot. Yeah. Is

24:18

that a smart thing to do at this point? Well, it

24:20

seems to be emerging as certainly

24:23

a term that's language around AI, but

24:25

also one that's owned by Microsoft. Yeah.

24:27

So do you want to go there?

24:29

What's clever about that is that it's

24:31

starting to describe how you

24:33

will use this stuff. Because everybody's very

24:36

happy to talk about AI as though

24:38

it's some form of magic that if

24:40

you apply to things, but people have

24:42

to understand this year and

24:45

going forward what they're actually going to

24:47

do with it. What is it going

24:49

to enable? So I think unless you

24:51

can describe what something is going to

24:53

enable, don't go near it. That's the

24:55

key to categorization. The word AI will sort

24:58

of disappear off into the rear view mirror,

25:01

but the sorts of sub segments of

25:03

the use of that technology, which may

25:05

have AI plus a modifying words with

25:07

it, that is sort of more directionally

25:09

where you want to be thinking. And

25:12

the example that comes to mind again is

25:14

the IntelliSense guys out of the UK. And

25:17

what they're doing in terms of, their

25:20

AI is all about mineral extraction,

25:23

et cetera. So it

25:25

could be that we find a bunch of

25:27

these AIs, the word

25:30

AI disappears, but it's more to do with

25:32

what the category enables

25:34

the end user to do. Right, in

25:36

that case it's mineral extraction. So how

25:39

do you go from the AI that

25:41

enables the mineral extraction, there is some

25:43

modifier in the middle of

25:45

that, which describes the impact of this

25:47

new approach to data processing, which is

25:50

essentially what AI is. Absolutely, and what's

25:52

very hard to see, but it's almost

25:54

inevitable is with this amount of funding,

25:56

chasing a growing number, but as we

25:58

see it, a reasonably small

26:01

number of companies, we've listed just sort

26:03

of seven or eight of them, not

26:05

everybody's gonna win. So the

26:07

time to think about how you differentiate yourself is

26:09

now, because you don't wanna

26:11

be player three, player four,

26:14

and head to head with somebody

26:16

like Microsoft. You get killed. So

26:19

half a billion sounds super impressive today, but

26:22

after the Cambrian explosion settles down a little

26:24

bit, it's time to be different. The

26:26

risk of misusing a geology word, there'll

26:28

be a strata, which is sort of platformy. We've

26:30

talked about that before. That may be owned by

26:33

the big guys, and it may revert to zero

26:35

as a profit machine. It might just be like

26:37

the internet, you know, like you wanna be

26:40

a telco at this point, right? So the

26:42

big AI sort of evolution may settle down,

26:44

and very quickly you'll have to be the

26:46

specialist in your niche, but to your geology,

26:48

environmental stuff. So you'll have to figure out

26:50

what problem it is that you solve, solve

26:53

it very differently, and prove your difference. I

27:02

wanted to talk a little bit about big lumps of data, box,

27:06

that's what they're called, box. Mark Benioff

27:09

has published another tome fairly recently. It's

27:12

called Trailblazer for people that haven't, you

27:14

know, in any way been aware of

27:17

what this now does. We give all the

27:19

team copies of Behind the Clouds, which is

27:21

the story of how salesforce.com went from idea

27:23

to billion. Now that seems very passe, because

27:25

a billion was a long time ago for

27:27

this guy. Yeah, but I think, you know,

27:29

as you'd expect, there's a bit of a

27:32

category story here, and you know, we go

27:34

on quite a lot about Eddie Yoon and

27:36

Linda Deke's seminal piece in Harvard

27:38

Business Review. I mean, it's 10 years old, but

27:40

it's still good stuff. And you know, what we

27:42

got out of that was that category leaders get

27:45

most of the value out

27:47

of the ecosystems they build around them.

27:50

And funnily enough, even on that books,

27:52

10 years old, you know, I see

27:54

UK and European entrepreneurs in organizations like

27:56

Boardwave, et cetera, now only catching

27:58

on to how long it... takes to fully flesh

28:01

out a category and the beauty of the

28:03

long-distance strategy that are building a new category.

28:06

I think you're absolutely right. I think you

28:08

don't have to look much further than Salesforce

28:10

for an object less than delivering value out

28:13

of category design. I mean, just look at

28:15

the numbers. They're

28:17

non-GAAP-REVs for FY 2024. They're

28:20

projecting 34.8 billion. And

28:25

FY 2023, that was 31.4, 22, just over 26 billion. And

28:32

he couldn't help himself. 10 years ago, it was 4.1 billion. I

28:34

mean, I think it was very eloquent. It was just a tweet

28:36

with just those numbers from the

28:39

man himself. And I think what he's saying

28:41

is interpret it as you will. We

28:43

have a category story and we have built

28:46

category that just pays

28:48

off in numbers. You know, the stock price

28:50

has increased over five times in 10 years.

28:53

I think, you know, with my M&A hats

28:56

on here, the thing I think is amazing

28:58

is the increase in margin. And

29:00

you can see the increase in margin

29:03

just going up and up and up

29:05

as their leadership in CRM became indisputable.

29:07

You know, in 2020, the margin was

29:10

just over 16%. It's

29:12

now over 30%. Right. And

29:15

so margin, let's remember this is returning money back

29:17

to the investors. Unlike everybody in the zero interest

29:19

rate period where they were just, you know, please

29:22

invest in me. I will make you some margin

29:24

at some point. Yeah. This guy's

29:26

a margin machine. Absolutely. And you

29:28

know, there wasn't a blip during COVID-19.

29:31

Maybe not. I mean, like he's bought as

29:33

well as built a lot of stuff. So

29:35

you think about Heroku that was inspired. You

29:38

think about Tableau, you could

29:40

say overpaid for it demand where? Exact

29:42

target. Don't know. But there's

29:44

some crackers in there and they get bigger and bigger

29:47

over time. So you look at, you know, 2020, it

29:49

was Slack and that was 27.7 billion, you know, Tableau

29:53

2019 15.78, MuleSoft 6.5 in 2018. I

29:58

don't know about the MuleSoft. He

30:03

has built and brought his way into

30:05

success but he was always clear in

30:07

his vision from the start that he's going to be a

30:09

category that he's going to own and

30:11

what he's done is fleshed out around it. He's

30:13

almost built the ecosystem for what used to be

30:16

called CRM. He's definitely a guy to keep an

30:18

eye on. I like the eloquence of instead of

30:20

talking about things in terms of

30:22

Flurry Language just posting the numbers, at

30:25

this time when everybody's hurting, I think he's

30:27

really telling people the boss

30:30

of Category. I think so. These things didn't

30:32

happen by accident as you say. It's a

30:34

reminder that the hard yards in Category Design

30:36

can really deliver if you get it right.

30:40

I guess my view is that if you're

30:42

looking to find some sort of inspiration for

30:44

a Category journey in 2024, look

30:47

no further than Trailblazer. Get that

30:49

on your New Year's reading

30:51

list. One of my favourite quotes from

30:53

that book is, and I have to read this, I can't

30:56

remember it all, he sort of neatly

30:58

summarised the aim of Categorisation, sort of

31:00

inter alia and something else. He said,

31:02

the future isn't about learning to be

31:04

better at doing what we already do,

31:06

it's about how far we can stretch

31:08

the boundaries of our imagination. Yeah,

31:11

and I think we always

31:13

refer back to, it's ten years old now, it's

31:15

an old book, Play Viggo, right? And

31:18

Benioff's on the dust jacket of that,

31:20

talking about the importance of Category. And

31:22

lest you forget, this guy figured it out

31:24

a long time ago and everyone else is still wringing their

31:26

hands about whether or not it's a good thing to do,

31:30

actually the proof is in the pudding and the proof

31:32

is in the numbers on this one. It's a good

31:34

one to read any time, but

31:36

the sooner the better. Yes, as far as

31:38

we're concerned, Mark Benioff is Category. We salute

31:40

you, Mark. Thank

31:44

you for listening. If you want to learn

31:47

more about what you've heard today, then you

31:49

can go and read our blog at bekategorical.com.

31:52

If you're ready to make that bold

31:54

Category play, then get in touch. You

31:56

can find all our contact details in

31:58

the show notes. Remember, don't

32:00

be better, be different.