WEBVTT
1
00:00:00.208 --> 00:00:03.854
Say your company needs something your own
city
2
00:00:04.207 --> 00:00:08.046
does not have. A growth investor who knows
your sector,
3
00:00:08.544 --> 00:00:12.254
a head of sales who has done this, a big
customer
4
00:00:12.559 --> 00:00:15.996
in another country. It exists somewhere in
Europe.
5
00:00:16.494 --> 00:00:20.172
Can you actually use it? Here is one data
point that
6
00:00:20.188 --> 00:00:23.127
should make you ask. When scientists in
7
00:00:23.320 --> 00:00:26.452
European hubs publish together, The
8
00:00:26.452 --> 00:00:29.978
European Commission scores Europe's
network
9
00:00:30.283 --> 00:00:33.585
at 0.76. America,
10
00:00:33.682 --> 00:00:36.824
0.81. Almost level.
11
00:00:37.513 --> 00:00:41.088
When inventors in different hubs file
patents together,
12
00:00:41.745 --> 00:00:44.919
Europe drops to 0.40.
13
00:00:45.608 --> 00:00:48.895
America, 0.68.
14
00:00:49.536 --> 00:00:52.937
That does not tell us everything about how
startups scale,
15
00:00:53.630 --> 00:00:57.465
but it tells us that the connection
between Europe's hubs
16
00:00:57.884 --> 00:01:01.657
gets weaker as research moves toward
17
00:01:02.175 --> 00:01:05.783
a product. Today, islands of excellence.
18
00:01:16.463 --> 00:01:20.162
Hello and welcome everybody to episode 781
19
00:01:20.438 --> 00:01:24.213
of the Startuprad.io franchise. My name is
Joe. I'm
20
00:01:24.213 --> 00:01:27.955
coming to you from Frankfurt am Main,
Germany, and this is part 7
21
00:01:28.132 --> 00:01:31.361
of our series on the European scale-up
question.
22
00:01:31.842 --> 00:01:35.296
Here is what we are looking at today. In
the last
23
00:01:35.441 --> 00:01:39.087
18 months, the EU has published 3
24
00:01:39.087 --> 00:01:42.894
sets of numbers that belong together: a
ranking of
25
00:01:42.990 --> 00:01:46.749
Europe's most innovative regions, its
first startup and
26
00:01:46.781 --> 00:01:50.331
scale-up index for every member country,
and a
27
00:01:50.331 --> 00:01:53.413
study on how well Europe's innovation hubs
28
00:01:53.766 --> 00:01:57.520
connect with each other. Put them side by
side and you
29
00:01:57.520 --> 00:02:01.209
get a question every founder in Germany,
Austria, Switzerland, and
30
00:02:01.209 --> 00:02:04.401
beyond runs into sooner or later: when the
31
00:02:04.418 --> 00:02:07.979
capability you need sits in another hub,
what
32
00:02:07.979 --> 00:02:11.588
stands between you and using it? That is
the question for this
33
00:02:11.588 --> 00:02:15.278
episode. I'll give you the evidence, what
it shows, what it
34
00:02:15.278 --> 00:02:18.935
does not show, and in each of the 5 areas
we cover
35
00:02:19.112 --> 00:02:22.558
the decision I would make as a
36
00:02:22.558 --> 00:02:26.297
founder. Thesis. Here's the thesis
37
00:02:26.523 --> 00:02:30.034
I want to state carefully as usual. Europe
has
38
00:02:30.292 --> 00:02:33.852
outstanding innovation regions. What a
scaling
39
00:02:33.852 --> 00:02:37.274
company needs is strong capability at home
and the
40
00:02:37.291 --> 00:02:40.665
ability to reach capital, customers, and
41
00:02:40.762 --> 00:02:44.466
expertise elsewhere. The evidence suggests
that second
42
00:02:44.499 --> 00:02:47.987
part reaching beyond your home hub is
weaker in Europe
43
00:02:48.357 --> 00:02:52.110
than in the USA. And it suggests
44
00:02:52.239 --> 00:02:55.984
the gap is widest in the commercial layer,
not
45
00:02:56.097 --> 00:02:59.689
in the lab. That is a hypothesis we will
test today,
46
00:03:00.336 --> 00:03:04.086
not a law of nature. Continuity beat.
57
00:03:41.419 --> 00:03:45.094
Where does this sit in our series? In
58
00:03:45.174 --> 00:03:48.866
parts 1 to 5, we looked at the bottlenecks
one at a
59
00:03:48.866 --> 00:03:52.541
time. Fragmentation as a hidden growth
tax, capital that
60
00:03:52.733 --> 00:03:56.489
breaks between funding rounds, public
money that funds supply but
61
00:03:56.489 --> 00:04:00.164
rarely becomes a customer, and experienced
operators
62
00:04:00.501 --> 00:04:04.128
who do not recycle into a new company. In
part 6,
63
00:04:04.273 --> 00:04:07.788
Capital Gravity, we looked at companies
moving their parent
64
00:04:07.820 --> 00:04:11.635
company, their CEO, and their capital
relationships to the United
65
00:04:11.651 --> 00:04:15.402
States. Today's idea is that several of
66
00:04:15.498 --> 00:04:18.993
these bottlenecks may also be problems of
access
67
00:04:19.041 --> 00:04:22.776
between hubs, not all of them, but
several. And
68
00:04:23.112 --> 00:04:26.622
one correction on the record. In part 5,
we cited
69
00:04:26.895 --> 00:04:30.309
the figure that 56% of startups founded by
unicorn
70
00:04:30.421 --> 00:04:33.755
alumni are started in the same city as the
unicorn.
71
00:04:34.108 --> 00:04:37.767
That figure comes from Dealroom and
Accel's 2022 data,
72
00:04:38.040 --> 00:04:41.815
and it covers Europe and Israel together.
We had put it next to
73
00:04:41.815 --> 00:04:45.590
their 2026 numbers. The figure is real.
The year
74
00:04:45.590 --> 00:04:49.334
and the region matter. Europe already has
very strong
75
00:04:49.848 --> 00:04:53.591
innovation regions. The EU's Regional
Innovation
76
00:04:53.687 --> 00:04:56.723
Ranking for 2025 covers
77
00:04:56.981 --> 00:05:00.595
241 regions. 38 are rated
78
00:05:00.949 --> 00:05:04.192
Innovation Leaders and 69 Strong
Innovators
79
00:05:04.770 --> 00:05:08.572
Between 2018 and 2025, performance
80
00:05:08.700 --> 00:05:12.325
went up in 233
81
00:05:12.951 --> 00:05:16.063
of the altogether 239 regions
82
00:05:16.641 --> 00:05:20.395
that could be compared. Number 1 is
actually
83
00:05:20.443 --> 00:05:23.683
Stockholm, then Copenhagen, London,
84
00:05:24.116 --> 00:05:27.261
Zurich, and Upper Bavaria around Munich.
85
00:05:28.111 --> 00:05:31.623
One thing to be precise about: this ranks
regional innovation
86
00:05:31.655 --> 00:05:35.422
systems, research, skills, companies
investing in
87
00:05:35.422 --> 00:05:38.756
innovation. It is not a ranking of startup
cities,
88
00:05:39.189 --> 00:05:42.555
but 2 out of the 5 are in our home market,
89
00:05:42.732 --> 00:05:46.210
Zurich and Munich. Now look at the
national
90
00:05:46.242 --> 00:05:49.673
picture. On the EU's new Startup and
Scale-up
91
00:05:49.769 --> 00:05:53.263
Index, Germany ranks 14th of
92
00:05:53.520 --> 00:05:57.367
27, at roughly the EU average. That is not
a
93
00:05:57.367 --> 00:06:01.118
contradiction. Big federal countries
always have strong
94
00:06:01.294 --> 00:06:05.140
and weak regions. What it tells you is
simply that a region can
95
00:06:05.156 --> 00:06:08.586
be excellent while the national conditions
around it are
96
00:06:08.746 --> 00:06:12.143
average. They measure different things.
And Europe
97
00:06:12.271 --> 00:06:16.053
overall is moving. The EU Startup and
Scale-Up Index
98
00:06:16.117 --> 00:06:19.610
improved by 13.5 percentage points between
99
00:06:19.610 --> 00:06:23.039
2020 and 2025. That is an index
100
00:06:23.039 --> 00:06:26.500
score, not 13.5% more startups.
101
00:06:26.869 --> 00:06:29.974
The Commission also said the process is
102
00:06:30.314 --> 00:06:33.920
uneven between countries, and that
late-stage money is still
103
00:06:34.033 --> 00:06:37.009
constrained. So the strong regions exist.
104
00:06:37.840 --> 00:06:41.535
The question is what it costs a company in
one of them
105
00:06:41.844 --> 00:06:45.552
to use what the others have. The
106
00:06:45.552 --> 00:06:48.863
framework: local capability and outside
access.
107
00:06:51.420 --> 00:06:53.935
Here is a model for you today. 2 ideas.
108
00:06:54.696 --> 00:06:58.047
Ecosystem density is how much startup
capability
109
00:06:58.416 --> 00:07:01.654
sits inside one single hub. Think
founders,
110
00:07:01.991 --> 00:07:05.181
angels, universities, seed funds, early
customers.
111
00:07:06.159 --> 00:07:09.670
Ecosystem connectivity is how easily a
company can
112
00:07:09.926 --> 00:07:13.389
use what sits outside its home hub— an
113
00:07:13.389 --> 00:07:16.595
investor in Stockholm, an industrial
customer in Munich,
114
00:07:17.461 --> 00:07:20.893
an engineer in Paris, a company Starting
out
115
00:07:21.150 --> 00:07:23.863
mostly needs the first. A company
116
00:07:23.912 --> 00:07:27.621
scaling needs both, because almost no
single
117
00:07:27.621 --> 00:07:31.442
city has every investor, executive, and
customer
118
00:07:31.763 --> 00:07:35.360
you will need between 50 and 500
employees. The
119
00:07:35.376 --> 00:07:39.213
balance depends on your sector and your
stage. That is the
120
00:07:39.213 --> 00:07:42.858
lens for the rest of this episode. The
evidence—
121
00:07:43.083 --> 00:07:46.663
what it shows and what it does not. The
best
122
00:07:46.775 --> 00:07:49.984
evidence we have comes from a study of the
123
00:07:49.984 --> 00:07:53.160
Commission's research department published
on June
124
00:07:53.192 --> 00:07:56.610
25th. The researchers mapped collaboration
125
00:07:57.444 --> 00:08:00.941
between 168 European and
126
00:08:01.054 --> 00:08:04.391
144 American urban areas
127
00:08:04.808 --> 00:08:08.289
from 2000 to 2023. Europe
128
00:08:08.498 --> 00:08:12.124
here includes the UK, Switzerland, and
Norway,
129
00:08:12.525 --> 00:08:16.136
so it's not just an EU story. For patents,
they
130
00:08:16.136 --> 00:08:19.892
looked at where the investors on
international patent applications are
131
00:08:19.908 --> 00:08:23.375
based. Then they built a benchmark— the
links
132
00:08:23.680 --> 00:08:27.339
you would expect between hubs given how
big each
133
00:08:27.435 --> 00:08:31.127
hub is. The index measures how closely the
134
00:08:31.127 --> 00:08:34.834
real links follow that benchmark. 1 is a
perfect fit,
135
00:08:35.011 --> 00:08:38.781
0 means no relationship at all, between
the real network
136
00:08:39.085 --> 00:08:42.372
and the expected one. It is not a
percentage of links
137
00:08:42.372 --> 00:08:46.059
achieved. Here is the most important
138
00:08:46.059 --> 00:08:49.650
comparison in this episode. When European
scientists in
139
00:08:49.650 --> 00:08:53.080
different hubs publish together, Europe's
network score is
140
00:08:53.305 --> 00:08:56.174
0.76 against America's
141
00:08:56.255 --> 00:08:59.541
0.81. Close. When European
142
00:08:59.541 --> 00:09:03.132
investors in different hubs patent
together, Europe
143
00:09:03.132 --> 00:09:06.212
falls to 0.4 against America's
144
00:09:06.742 --> 00:09:10.319
0.68. And Europe's hubs
145
00:09:10.752 --> 00:09:14.345
are physically closer to each other than
America's. So
146
00:09:14.586 --> 00:09:18.147
physical distance alone does not explain
it. What
147
00:09:18.147 --> 00:09:21.837
separates Munich from Zurich is not the
train ride.
148
00:09:21.997 --> 00:09:25.703
It is a border, a currency, a tax system,
an
149
00:09:25.831 --> 00:09:29.649
employment law, and separate patent and
company registries.
150
00:09:31.448 --> 00:09:34.998
That last point matters. Boston to San
Francisco is a
151
00:09:35.047 --> 00:09:38.645
flight across the continent, but it
happens inside one currency,
152
00:09:39.304 --> 00:09:42.903
one securities regulator, one patent
office, and one
153
00:09:42.935 --> 00:09:46.389
capital market. Physical distance is small
in Europe.
154
00:09:46.791 --> 00:09:50.405
Institutional distance is large. The study
155
00:09:50.470 --> 00:09:53.747
also finds that European collaboration
clusters
156
00:09:54.117 --> 00:09:57.923
inside national borders and that the gap
with America is
157
00:09:58.180 --> 00:10:01.440
largest in complex technologies— think AI,
158
00:10:01.713 --> 00:10:05.486
biotech, and quantum computing. Now, what
this idea
159
00:10:05.486 --> 00:10:09.195
does not show: it measures invention
networks,
160
00:10:09.420 --> 00:10:13.049
not the whole business. An investor can
fund a startup in another
161
00:10:13.049 --> 00:10:16.629
country without any shared patent. A
company
162
00:10:16.886 --> 00:10:20.322
can license a technology or hire someone
remotely
163
00:10:20.659 --> 00:10:24.320
without a shared patent. And a part of the
American picture
164
00:10:24.770 --> 00:10:28.431
may be that More US companies have teams
in several hubs
165
00:10:28.848 --> 00:10:32.445
under one parent company, so the investors
in different cities
166
00:10:32.782 --> 00:10:36.555
show up as links. The study names possible
167
00:10:36.571 --> 00:10:40.393
causes— national rules, language,
different legal and of
168
00:10:40.409 --> 00:10:44.246
course tax frameworks— but it does not
provide which one matters
169
00:10:44.343 --> 00:10:48.084
most. So I will not tell you this number
proves where
170
00:10:48.277 --> 00:10:51.776
every European bridge breaks. What I will
say
171
00:10:52.242 --> 00:10:55.885
is this: the link between European hubs
172
00:10:56.238 --> 00:11:00.025
gets weaker as you move from research
towards commercial
173
00:11:00.025 --> 00:11:03.299
invention that is worth testing in the
174
00:11:03.380 --> 00:11:06.734
parts of company the patent data cannot
see.
175
00:11:08.114 --> 00:11:11.918
Now the other side in its strongest
176
00:11:12.030 --> 00:11:15.866
form: maybe Europe's real problem is
177
00:11:16.058 --> 00:11:19.557
too few hubs have enough growth capital,
enough large
178
00:11:19.733 --> 00:11:23.215
customers, and enough experienced leaders.
Connecting them more
179
00:11:23.519 --> 00:11:27.129
will not create resources that are missing
everywhere.
180
00:11:28.075 --> 00:11:31.620
And a founder should choose the best
partner for the
181
00:11:31.620 --> 00:11:35.422
company wherever it is—in your own city,
in
182
00:11:35.470 --> 00:11:39.095
another European hub, or in the US. I
think that
183
00:11:39.095 --> 00:11:42.576
objection is partly right. If the resource
does
184
00:11:42.768 --> 00:11:46.394
not exist anywhere in Europe, no bridge
will help you. And the
185
00:11:46.442 --> 00:11:50.053
evidence today does not settle whether
Europe needs a
186
00:11:50.053 --> 00:11:53.763
few much stronger hubs or just better
links
187
00:11:53.763 --> 00:11:57.537
between many. Probably both, depending a
little bit on the
188
00:11:57.537 --> 00:12:01.247
sector. But here is where I part ways with
it.
189
00:12:01.793 --> 00:12:05.455
When a resource does exist in Europe, the
extra costs of
190
00:12:05.567 --> 00:12:09.132
reaching it across a border is real and is
paid by the
191
00:12:09.165 --> 00:12:13.003
founder. The EU's own research on venture
192
00:12:13.115 --> 00:12:16.857
and growth capital published in October
2025 describes
193
00:12:17.210 --> 00:12:20.646
Europe's market as still fragmented and
lists
194
00:12:21.625 --> 00:12:25.367
legal, tax, and market barriers to
investing
195
00:12:25.415 --> 00:12:28.835
across borders and to raising money from
pension funds and
196
00:12:28.851 --> 00:12:32.640
insurers. More money behind the same
borders would make
197
00:12:32.704 --> 00:12:36.044
every hub richer. It would not
automatically
198
00:12:36.397 --> 00:12:40.154
make them easier to reach. So the use for
gold is not
199
00:12:40.170 --> 00:12:43.568
connected. activity for its own sake, but
it's building valuable
200
00:12:43.696 --> 00:12:47.413
capabilities and making them cheaper to
reach. We should
201
00:12:47.606 --> 00:12:51.243
judge the links by what they let a company
do.
203
00:12:55.537 --> 00:12:59.046
So let us get practical. Think of 5 areas
where your
204
00:12:59.062 --> 00:13:02.603
company may need something from outside
your home hub:
205
00:13:03.003 --> 00:13:06.624
capital, people, customers, know-how, and
your company
206
00:13:06.624 --> 00:13:10.391
structure itself. For each, here's what we
know, what
207
00:13:10.439 --> 00:13:14.126
we do not know, and the decision I would
take.
208
00:13:15.328 --> 00:13:18.631
Keep in mind, this is no legal or
financial recommendation. Yeah,
209
00:13:19.592 --> 00:13:23.087
take it at your own risk and do your
homework. Capital.
210
00:13:23.584 --> 00:13:27.127
What we know: the EU's own capital
research finds
211
00:13:27.207 --> 00:13:30.622
barriers to cross-border investments and
to fundraising from large
212
00:13:30.670 --> 00:13:34.453
institutions. What we do not know from
public data is how
213
00:13:34.662 --> 00:13:38.111
Often a good round fails because of a
border.
214
00:13:38.592 --> 00:13:42.426
If I were a founder, before your next
round, look beyond
215
00:13:42.426 --> 00:13:45.715
your home network for investors who
understand your sector and
216
00:13:45.892 --> 00:13:49.582
can back your next stage. Ask
217
00:13:49.582 --> 00:13:53.368
every fund 2 questions: does your mandate
let
218
00:13:53.368 --> 00:13:56.769
you invest in my country, and how much do
you keep in
219
00:13:56.865 --> 00:14:00.700
reserve for follow-on rounds? A fund that
loves you
220
00:14:00.732 --> 00:14:04.291
but cannot follow you is not your lead.
People—
221
00:14:06.112 --> 00:14:09.720
what we know: Dealroom reported for Europe
and Israel in
222
00:14:09.752 --> 00:14:13.198
2022 that 56% of startups founded by
223
00:14:13.278 --> 00:14:16.901
unicorn alumni were started in the same
city as the
224
00:14:16.901 --> 00:14:20.475
unicorn. What that does not tell us is how
225
00:14:20.539 --> 00:14:24.368
far those founders' experience reaches. A
founder can stay in
226
00:14:24.480 --> 00:14:28.278
Berlin and still hire in Stockholm. Raise
in London and sell
227
00:14:28.278 --> 00:14:32.051
in France. Local founding and wide reach
can go
228
00:14:32.051 --> 00:14:35.776
together. If I were a founder: when you
hire
229
00:14:35.969 --> 00:14:39.453
your first leaders for the 50 to 500
employee
230
00:14:39.453 --> 00:14:43.098
stage, start from the capability gap, not
the postcode.
231
00:14:43.852 --> 00:14:47.449
Search beyond your home hub when the skill
is not there, and
232
00:14:47.529 --> 00:14:51.142
check early what it takes to employ
someone and give
233
00:14:51.239 --> 00:14:54.517
them equity in the countries you are
likely to hire from.
234
00:15:00.590 --> 00:15:03.738
Hey guys, thanks for sticking with us.
Let's keep going.
235
00:15:04.916 --> 00:15:08.517
A quick word on how this show works.
Startup
236
00:15:08.517 --> 00:15:12.352
Radio is funded by partners, and the
partners we like best are
237
00:15:12.352 --> 00:15:16.077
the ones who make things easier to reach
across borders. Funds
238
00:15:16.222 --> 00:15:19.763
that invest across Germany, Austria, and
Switzerland, corporates that buy
239
00:15:19.812 --> 00:15:22.968
from startups in other countries, regions
that want
240
00:15:22.968 --> 00:15:26.578
founders in Berlin, Munich, Vienna, and
Zurich to know
241
00:15:27.420 --> 00:15:30.703
what they offer. If that is you, your
listeners
242
00:15:31.204 --> 00:15:34.923
are the founders and investors we— you are
trying to
243
00:15:34.939 --> 00:15:38.151
reach. Find out how we work together at
startuprad.io/become-a-partner.
244
00:15:42.450 --> 00:15:45.796
Now back to the 5 areas. Customers.
245
00:15:46.535 --> 00:15:50.278
A German startup should not have to become
international
246
00:15:50.326 --> 00:15:54.133
just because its next big customer is
French. In practice,
247
00:15:54.502 --> 00:15:57.923
it often does. New contract habits, new
248
00:15:57.956 --> 00:16:01.698
procurement rules, and maybe new language
and a new sales
249
00:16:01.891 --> 00:16:05.682
team. How big is that friction? The IMF
estimates
250
00:16:05.730 --> 00:16:09.248
that barriers inside Europe work like a
tariff of around
251
00:16:09.376 --> 00:16:11.816
44% on goods and
252
00:16:12.153 --> 00:16:15.523
110% on services. The IMF
253
00:16:15.523 --> 00:16:19.279
itself calls that an upper bound, and
other
254
00:16:19.375 --> 00:16:22.890
economists dispute it. It is a model of
trade
255
00:16:22.986 --> 00:16:26.822
costs, not a bill— not a bill your company
pays, but
256
00:16:26.854 --> 00:16:30.241
it tells you the friction is not
imaginary.
257
00:16:31.011 --> 00:16:34.558
Here's a decision I would make in your
second market:
258
00:16:35.024 --> 00:16:38.781
choose it where you have the strongest
evidence of customer demand
259
00:16:39.054 --> 00:16:42.522
and a credible way to sell, not where the
train is shortest.
260
00:16:43.790 --> 00:16:47.596
A nearby market like Austria or
Switzerland can be a cheap place
261
00:16:47.708 --> 00:16:51.529
to learn, to win a pilot customer and test
your sales process.
262
00:16:52.059 --> 00:16:55.415
But do not mistake Germany and Austria and
Switzerland for scale.
263
00:16:55.817 --> 00:16:58.899
Together, they're still smaller than the
market an American
264
00:16:58.964 --> 00:17:02.496
competitor gets at home. You use the
corridor to
265
00:17:02.576 --> 00:17:05.194
learn, then go where the demand is.
266
00:17:06.591 --> 00:17:10.091
Know-how. This is where Europe could have
an advantage.
267
00:17:11.247 --> 00:17:14.426
No city has to be best at everything.
268
00:17:15.406 --> 00:17:19.243
Munich, Zurich, Eindhoven, Grenoble,
Stockholm, and Cambridge
269
00:17:19.757 --> 00:17:23.450
can each specialize, but specialization
only helps
270
00:17:23.546 --> 00:17:26.324
if companies can reach it, and the patent
data
271
00:17:26.517 --> 00:17:30.017
suggests I would suggest this is where
Europe is weakest. If I were a
272
00:17:30.129 --> 00:17:33.822
founder in deep tech, identify the
capability you
273
00:17:34.031 --> 00:17:37.307
are missing, then find the best partner to
supply it,
274
00:17:37.901 --> 00:17:41.321
whether that is your own university or a
team in another
275
00:17:41.353 --> 00:17:44.853
country. And when you do partner across a
border,
276
00:17:45.527 --> 00:17:49.220
sort out who owns the resulting IP before
the
277
00:17:49.220 --> 00:17:52.351
work starts, not after. Investors care
278
00:17:52.351 --> 00:17:55.958
about clean ownership. Of course, your
company
279
00:17:55.958 --> 00:17:59.758
structure. In Part 6, Capital Gravity, we
describe
280
00:17:59.823 --> 00:18:03.599
companies moving their parent company CEOs
and listing
281
00:18:03.777 --> 00:18:07.449
plans to America as they grow. Sometimes
282
00:18:07.676 --> 00:18:11.397
that follows customers, capital, or the
executives they need.
283
00:18:11.819 --> 00:18:15.337
Sometimes it is a missing European option.
From the outside,
284
00:18:15.596 --> 00:18:18.823
you often cannot tell which. If I were a
founder
285
00:18:19.283 --> 00:18:22.345
Once a year, ask yourself the Capital
Gravity questions.
286
00:18:23.531 --> 00:18:26.994
Where is our parent company? Where does
your
287
00:18:26.994 --> 00:18:30.713
CEO work? Where are your senior commercial
hires?
288
00:18:31.627 --> 00:18:35.362
Where do we expect to list? If the answers
are moving
289
00:18:35.474 --> 00:18:38.568
west, write down why. If the reason
290
00:18:39.466 --> 00:18:43.298
is that something exists in Europe but
costs too much to reach,
291
00:18:43.955 --> 00:18:46.891
that is exactly the friction this episode
is about.
293
00:18:53.266 --> 00:18:56.429
There's a good idea for making Europe's
spread-out map work.
294
00:18:57.520 --> 00:19:01.105
Dealroom has argued for treating the hubs
within a 4-hour
295
00:19:01.105 --> 00:19:04.868
train ride of London as one supercluster,
296
00:19:05.368 --> 00:19:09.176
which it calls the New Palo Alto. Global
Venturing
297
00:19:09.291 --> 00:19:12.926
reported that Dealroom Also sketched a
second one around Munich
298
00:19:13.344 --> 00:19:16.798
linking Berlin, Milan, Vienna, and the
Swiss
299
00:19:16.798 --> 00:19:20.574
cantons. Those are proposals. Drawing a
circle on a map
300
00:19:20.896 --> 00:19:24.318
does not mean a company can use what is
inside
301
00:19:25.057 --> 00:19:28.560
it on the same terms. For our listeners,
the real
302
00:19:28.672 --> 00:19:32.400
test case sits across Munich, Vienna, and
Zurich,
303
00:19:32.930 --> 00:19:36.497
2 of Europe's top 5 innovation regions, 3
countries,
304
00:19:36.722 --> 00:19:40.365
And Switzerland outside the EU. The
network test.
305
00:19:41.091 --> 00:19:44.670
So instead of asking every region to build
its own
306
00:19:44.670 --> 00:19:48.150
Silicon Valley, here's a test you can run
on your own company.
307
00:19:48.975 --> 00:19:52.822
Let's call it the network test. For the
next
308
00:19:52.870 --> 00:19:56.375
capability your company needs, find the
best option at home
309
00:19:56.666 --> 00:20:00.381
and the best option elsewhere. Can you
actually use the outside
310
00:20:00.381 --> 00:20:04.012
option? What extra time costs or
311
00:20:04.076 --> 00:20:07.623
eligibility rules come with it, and what
would
312
00:20:07.687 --> 00:20:11.186
change for your company if that friction
disappeared?
313
00:20:11.539 --> 00:20:15.118
A few rules so the test means something:
only count
314
00:20:15.230 --> 00:20:18.761
things your company actually needed. A
specialist fund
315
00:20:18.761 --> 00:20:21.425
saying no to a company that does not fit
316
00:20:22.436 --> 00:20:26.111
is not a border problem. Moving the whole
company
317
00:20:26.464 --> 00:20:30.234
should not be the only way to get access.
And count
318
00:20:30.298 --> 00:20:33.391
the cases where it worked too. To be
clear,
319
00:20:34.160 --> 00:20:37.878
this is a test we are proposing, not
320
00:20:37.990 --> 00:20:41.483
one we have validated. So here is why I
321
00:20:41.611 --> 00:20:45.425
ask, if you have run into this around
322
00:20:45.762 --> 00:20:49.159
a hire, a customer, or partnership that
worked or
323
00:20:49.159 --> 00:20:52.733
failed because it sat in another hub, send
it to
324
00:20:52.829 --> 00:20:56.576
partnerships@startuprad.io. I'll use real
cases to
325
00:20:56.608 --> 00:20:59.239
test this framework in a later episode of
this series.
326
00:21:01.208 --> 00:21:04.953
Investors, briefly.
327
00:21:05.548 --> 00:21:09.291
if I were running a fund in Germany,
Austria, or Switzerland, I
328
00:21:09.323 --> 00:21:13.100
would build co-investment relationships
with funds in other hubs
329
00:21:13.325 --> 00:21:17.102
early so you have partners for later
rounds you cannot
330
00:21:17.118 --> 00:21:20.300
carry alone. If you are a limited partner
331
00:21:20.747 --> 00:21:24.451
ask your funds which follow-on rounds they
had to pass
332
00:21:24.531 --> 00:21:28.220
on because of a border. If you are a
corporate, the
333
00:21:28.220 --> 00:21:31.603
cheapest improvement you can make is a
buying process that a
334
00:21:31.603 --> 00:21:34.313
startup from the next country can
actually get through.
335
00:21:35.773 --> 00:21:39.413
EU Inc., the planned
336
00:21:39.445 --> 00:21:43.197
single company form for startups across
Europe, is the most
337
00:21:43.213 --> 00:21:47.030
direct fix for the company structure
problem in this episode. Its supporters
338
00:21:47.528 --> 00:21:51.093
want the first company founded under it in
2027. My
339
00:21:51.093 --> 00:21:54.835
on the record, and I first made it in part
340
00:21:54.835 --> 00:21:58.497
6: no startup will be able to register as
an EU
341
00:21:58.850 --> 00:22:02.399
Inc. in any EU country before January 1st,
342
00:22:02.640 --> 00:22:06.446
2028. If a registry accepts one before
343
00:22:06.479 --> 00:22:10.188
that, I was wrong. My reason: a new
company form has to
344
00:22:10.188 --> 00:22:13.770
get through EU governments and the
European Parliament
345
00:22:13.964 --> 00:22:17.791
And then every single national registry,
tax office,
346
00:22:18.147 --> 00:22:21.877
and notary has to be ready to handle it.
Each step
347
00:22:22.103 --> 00:22:24.182
takes time. Confidence: 75%.
348
00:22:26.212 --> 00:22:29.950
And one more. To be precise, this
forecast's
349
00:22:30.015 --> 00:22:33.746
wording in a report, not the state of the
market.
350
00:22:34.517 --> 00:22:38.133
The next edition of the EU Startup and
Scale-Up Index will
351
00:22:38.133 --> 00:22:41.142
again name late-stage funding As a
constraint.
352
00:22:41.880 --> 00:22:45.570
Confidence: 85%. If
353
00:22:45.586 --> 00:22:49.308
no new edition appears by the end of 2027,
that one
354
00:22:49.357 --> 00:22:53.143
stays open. Europe has
355
00:22:53.143 --> 00:22:56.945
spent 2 decades building strong startup
and innovation regions.
356
00:22:57.443 --> 00:23:01.133
In many places, it worked. The next
question is a
357
00:23:01.133 --> 00:23:04.759
different one. When a company needs
something that exists
358
00:23:04.983 --> 00:23:08.709
elsewhere in Europe, How hard is it to
use? The data says
359
00:23:08.918 --> 00:23:12.596
the link between Europe's hubs is close to
American levels in
360
00:23:12.789 --> 00:23:16.291
research and much weaker in commercial
invention.
361
00:23:17.046 --> 00:23:20.451
That does not explain the whole scale-up
gap, but it
362
00:23:20.499 --> 00:23:24.001
points to a cost founders pay every day
and one
363
00:23:24.338 --> 00:23:27.872
that policy can reduce. Europe has its
364
00:23:28.032 --> 00:23:31.822
islands of excellence. The question is,
how cheap it
365
00:23:31.822 --> 00:23:35.590
becomes to travel between them. There's
more to this story. For
366
00:23:35.590 --> 00:23:39.261
Entrepreneurs Vault members, I go one
level deeper. I run the
367
00:23:39.261 --> 00:23:42.948
network test on the Munich, Vienna, and
Zurich corridor area
368
00:23:43.044 --> 00:23:46.475
by area as a desk assessment. Members get
a
369
00:23:46.475 --> 00:23:50.226
city-by-city table of where unicorn alumni
founded their next company,
370
00:23:50.739 --> 00:23:54.394
what it can and cannot tell you, and a
stage-by-stage
371
00:23:54.491 --> 00:23:58.242
checklist for founders deciding what to
reach for outside
372
00:23:58.274 --> 00:24:01.771
their home hub. The full Entrepreneurs
Vault analysis
373
00:24:02.059 --> 00:24:05.844
is available to members. Join either on
YouTube or
374
00:24:05.892 --> 00:24:09.517
on Substack. Both links are in the show
notes. Next week on Startup
375
00:24:09.517 --> 00:24:13.125
Radio, more interviews and news from
Germany,
376
00:24:13.173 --> 00:24:16.685
Austria, and Switzerland, and the European
scale-up question continues.
377
00:24:17.118 --> 00:24:20.695
This is Startuprad.io. That's all, folks.
Find more news, streams,
378
00:24:20.871 --> 00:24:24.507
events, and interviews at www.
Startuprad.io. Remember,
379
00:24:24.652 --> 00:24:25.443
sharing is caring.
1
00:00:00.208 --> 00:00:03.854
Say your company needs something your own
city
2
00:00:04.207 --> 00:00:08.046
does not have. A growth investor who knows
your sector,
3
00:00:08.544 --> 00:00:12.254
a head of sales who has done this, a big
customer
4
00:00:12.559 --> 00:00:15.996
in another country. It exists somewhere in
Europe.
5
00:00:16.494 --> 00:00:20.172
Can you actually use it? Here is one data
point that
6
00:00:20.188 --> 00:00:23.127
should make you ask. When scientists in
7
00:00:23.320 --> 00:00:26.452
European hubs publish together, The
8
00:00:26.452 --> 00:00:29.978
European Commission scores Europe's
network
9
00:00:30.283 --> 00:00:33.585
at 0.76. America,
10
00:00:33.682 --> 00:00:36.824
0.81. Almost level.
11
00:00:37.513 --> 00:00:41.088
When inventors in different hubs file
patents together,
12
00:00:41.745 --> 00:00:44.919
Europe drops to 0.40.
13
00:00:45.608 --> 00:00:48.895
America, 0.68.
14
00:00:49.536 --> 00:00:52.937
That does not tell us everything about how
startups scale,
15
00:00:53.630 --> 00:00:57.465
but it tells us that the connection
between Europe's hubs
16
00:00:57.884 --> 00:01:01.657
gets weaker as research moves toward
17
00:01:02.175 --> 00:01:05.783
a product. Today, islands of excellence.
18
00:01:16.463 --> 00:01:20.162
Hello and welcome everybody to episode 781
19
00:01:20.438 --> 00:01:24.213
of the Startuprad.io franchise. My name is
Joe. I'm
20
00:01:24.213 --> 00:01:27.955
coming to you from Frankfurt am Main,
Germany, and this is part 7
21
00:01:28.132 --> 00:01:31.361
of our series on the European scale-up
question.
22
00:01:31.842 --> 00:01:35.296
Here is what we are looking at today. In
the last
23
00:01:35.441 --> 00:01:39.087
18 months, the EU has published 3
24
00:01:39.087 --> 00:01:42.894
sets of numbers that belong together: a
ranking of
25
00:01:42.990 --> 00:01:46.749
Europe's most innovative regions, its
first startup and
26
00:01:46.781 --> 00:01:50.331
scale-up index for every member country,
and a
27
00:01:50.331 --> 00:01:53.413
study on how well Europe's innovation hubs
28
00:01:53.766 --> 00:01:57.520
connect with each other. Put them side by
side and you
29
00:01:57.520 --> 00:02:01.209
get a question every founder in Germany,
Austria, Switzerland, and
30
00:02:01.209 --> 00:02:04.401
beyond runs into sooner or later: when the
31
00:02:04.418 --> 00:02:07.979
capability you need sits in another hub,
what
32
00:02:07.979 --> 00:02:11.588
stands between you and using it? That is
the question for this
33
00:02:11.588 --> 00:02:15.278
episode. I'll give you the evidence, what
it shows, what it
34
00:02:15.278 --> 00:02:18.935
does not show, and in each of the 5 areas
we cover
35
00:02:19.112 --> 00:02:22.558
the decision I would make as a
36
00:02:22.558 --> 00:02:26.297
founder. Thesis. Here's the thesis
37
00:02:26.523 --> 00:02:30.034
I want to state carefully as usual. Europe
has
38
00:02:30.292 --> 00:02:33.852
outstanding innovation regions. What a
scaling
39
00:02:33.852 --> 00:02:37.274
company needs is strong capability at home
and the
40
00:02:37.291 --> 00:02:40.665
ability to reach capital, customers, and
41
00:02:40.762 --> 00:02:44.466
expertise elsewhere. The evidence suggests
that second
42
00:02:44.499 --> 00:02:47.987
part reaching beyond your home hub is
weaker in Europe
43
00:02:48.357 --> 00:02:52.110
than in the USA. And it suggests
44
00:02:52.239 --> 00:02:55.984
the gap is widest in the commercial layer,
not
45
00:02:56.097 --> 00:02:59.689
in the lab. That is a hypothesis we will
test today,
46
00:03:00.336 --> 00:03:04.086
not a law of nature. Continuity beat.
57
00:03:41.419 --> 00:03:45.094
Where does this sit in our series? In
58
00:03:45.174 --> 00:03:48.866
parts 1 to 5, we looked at the bottlenecks
one at a
59
00:03:48.866 --> 00:03:52.541
time. Fragmentation as a hidden growth
tax, capital that
60
00:03:52.733 --> 00:03:56.489
breaks between funding rounds, public
money that funds supply but
61
00:03:56.489 --> 00:04:00.164
rarely becomes a customer, and experienced
operators
62
00:04:00.501 --> 00:04:04.128
who do not recycle into a new company. In
part 6,
63
00:04:04.273 --> 00:04:07.788
Capital Gravity, we looked at companies
moving their parent
64
00:04:07.820 --> 00:04:11.635
company, their CEO, and their capital
relationships to the United
65
00:04:11.651 --> 00:04:15.402
States. Today's idea is that several of
66
00:04:15.498 --> 00:04:18.993
these bottlenecks may also be problems of
access
67
00:04:19.041 --> 00:04:22.776
between hubs, not all of them, but
several. And
68
00:04:23.112 --> 00:04:26.622
one correction on the record. In part 5,
we cited
69
00:04:26.895 --> 00:04:30.309
the figure that 56% of startups founded by
unicorn
70
00:04:30.421 --> 00:04:33.755
alumni are started in the same city as the
unicorn.
71
00:04:34.108 --> 00:04:37.767
That figure comes from Dealroom and
Accel's 2022 data,
72
00:04:38.040 --> 00:04:41.815
and it covers Europe and Israel together.
We had put it next to
73
00:04:41.815 --> 00:04:45.590
their 2026 numbers. The figure is real.
The year
74
00:04:45.590 --> 00:04:49.334
and the region matter. Europe already has
very strong
75
00:04:49.848 --> 00:04:53.591
innovation regions. The EU's Regional
Innovation
76
00:04:53.687 --> 00:04:56.723
Ranking for 2025 covers
77
00:04:56.981 --> 00:05:00.595
241 regions. 38 are rated
78
00:05:00.949 --> 00:05:04.192
Innovation Leaders and 69 Strong
Innovators
79
00:05:04.770 --> 00:05:08.572
Between 2018 and 2025, performance
80
00:05:08.700 --> 00:05:12.325
went up in 233
81
00:05:12.951 --> 00:05:16.063
of the altogether 239 regions
82
00:05:16.641 --> 00:05:20.395
that could be compared. Number 1 is
actually
83
00:05:20.443 --> 00:05:23.683
Stockholm, then Copenhagen, London,
84
00:05:24.116 --> 00:05:27.261
Zurich, and Upper Bavaria around Munich.
85
00:05:28.111 --> 00:05:31.623
One thing to be precise about: this ranks
regional innovation
86
00:05:31.655 --> 00:05:35.422
systems, research, skills, companies
investing in
87
00:05:35.422 --> 00:05:38.756
innovation. It is not a ranking of startup
cities,
88
00:05:39.189 --> 00:05:42.555
but 2 out of the 5 are in our home market,
89
00:05:42.732 --> 00:05:46.210
Zurich and Munich. Now look at the
national
90
00:05:46.242 --> 00:05:49.673
picture. On the EU's new Startup and
Scale-up
91
00:05:49.769 --> 00:05:53.263
Index, Germany ranks 14th of
92
00:05:53.520 --> 00:05:57.367
27, at roughly the EU average. That is not
a
93
00:05:57.367 --> 00:06:01.118
contradiction. Big federal countries
always have strong
94
00:06:01.294 --> 00:06:05.140
and weak regions. What it tells you is
simply that a region can
95
00:06:05.156 --> 00:06:08.586
be excellent while the national conditions
around it are
96
00:06:08.746 --> 00:06:12.143
average. They measure different things.
And Europe
97
00:06:12.271 --> 00:06:16.053
overall is moving. The EU Startup and
Scale-Up Index
98
00:06:16.117 --> 00:06:19.610
improved by 13.5 percentage points between
99
00:06:19.610 --> 00:06:23.039
2020 and 2025. That is an index
100
00:06:23.039 --> 00:06:26.500
score, not 13.5% more startups.
101
00:06:26.869 --> 00:06:29.974
The Commission also said the process is
102
00:06:30.314 --> 00:06:33.920
uneven between countries, and that
late-stage money is still
103
00:06:34.033 --> 00:06:37.009
constrained. So the strong regions exist.
104
00:06:37.840 --> 00:06:41.535
The question is what it costs a company in
one of them
105
00:06:41.844 --> 00:06:45.552
to use what the others have. The
106
00:06:45.552 --> 00:06:48.863
framework: local capability and outside
access.
107
00:06:51.420 --> 00:06:53.935
Here is a model for you today. 2 ideas.
108
00:06:54.696 --> 00:06:58.047
Ecosystem density is how much startup
capability
109
00:06:58.416 --> 00:07:01.654
sits inside one single hub. Think
founders,
110
00:07:01.991 --> 00:07:05.181
angels, universities, seed funds, early
customers.
111
00:07:06.159 --> 00:07:09.670
Ecosystem connectivity is how easily a
company can
112
00:07:09.926 --> 00:07:13.389
use what sits outside its home hub— an
113
00:07:13.389 --> 00:07:16.595
investor in Stockholm, an industrial
customer in Munich,
114
00:07:17.461 --> 00:07:20.893
an engineer in Paris, a company Starting
out
115
00:07:21.150 --> 00:07:23.863
mostly needs the first. A company
116
00:07:23.912 --> 00:07:27.621
scaling needs both, because almost no
single
117
00:07:27.621 --> 00:07:31.442
city has every investor, executive, and
customer
118
00:07:31.763 --> 00:07:35.360
you will need between 50 and 500
employees. The
119
00:07:35.376 --> 00:07:39.213
balance depends on your sector and your
stage. That is the
120
00:07:39.213 --> 00:07:42.858
lens for the rest of this episode. The
evidence—
121
00:07:43.083 --> 00:07:46.663
what it shows and what it does not. The
best
122
00:07:46.775 --> 00:07:49.984
evidence we have comes from a study of the
123
00:07:49.984 --> 00:07:53.160
Commission's research department published
on June
124
00:07:53.192 --> 00:07:56.610
25th. The researchers mapped collaboration
125
00:07:57.444 --> 00:08:00.941
between 168 European and
126
00:08:01.054 --> 00:08:04.391
144 American urban areas
127
00:08:04.808 --> 00:08:08.289
from 2000 to 2023. Europe
128
00:08:08.498 --> 00:08:12.124
here includes the UK, Switzerland, and
Norway,
129
00:08:12.525 --> 00:08:16.136
so it's not just an EU story. For patents,
they
130
00:08:16.136 --> 00:08:19.892
looked at where the investors on
international patent applications are
131
00:08:19.908 --> 00:08:23.375
based. Then they built a benchmark— the
links
132
00:08:23.680 --> 00:08:27.339
you would expect between hubs given how
big each
133
00:08:27.435 --> 00:08:31.127
hub is. The index measures how closely the
134
00:08:31.127 --> 00:08:34.834
real links follow that benchmark. 1 is a
perfect fit,
135
00:08:35.011 --> 00:08:38.781
0 means no relationship at all, between
the real network
136
00:08:39.085 --> 00:08:42.372
and the expected one. It is not a
percentage of links
137
00:08:42.372 --> 00:08:46.059
achieved. Here is the most important
138
00:08:46.059 --> 00:08:49.650
comparison in this episode. When European
scientists in
139
00:08:49.650 --> 00:08:53.080
different hubs publish together, Europe's
network score is
140
00:08:53.305 --> 00:08:56.174
0.76 against America's
141
00:08:56.255 --> 00:08:59.541
0.81. Close. When European
142
00:08:59.541 --> 00:09:03.132
investors in different hubs patent
together, Europe
143
00:09:03.132 --> 00:09:06.212
falls to 0.4 against America's
144
00:09:06.742 --> 00:09:10.319
0.68. And Europe's hubs
145
00:09:10.752 --> 00:09:14.345
are physically closer to each other than
America's. So
146
00:09:14.586 --> 00:09:18.147
physical distance alone does not explain
it. What
147
00:09:18.147 --> 00:09:21.837
separates Munich from Zurich is not the
train ride.
148
00:09:21.997 --> 00:09:25.703
It is a border, a currency, a tax system,
an
149
00:09:25.831 --> 00:09:29.649
employment law, and separate patent and
company registries.
150
00:09:31.448 --> 00:09:34.998
That last point matters. Boston to San
Francisco is a
151
00:09:35.047 --> 00:09:38.645
flight across the continent, but it
happens inside one currency,
152
00:09:39.304 --> 00:09:42.903
one securities regulator, one patent
office, and one
153
00:09:42.935 --> 00:09:46.389
capital market. Physical distance is small
in Europe.
154
00:09:46.791 --> 00:09:50.405
Institutional distance is large. The study
155
00:09:50.470 --> 00:09:53.747
also finds that European collaboration
clusters
156
00:09:54.117 --> 00:09:57.923
inside national borders and that the gap
with America is
157
00:09:58.180 --> 00:10:01.440
largest in complex technologies— think AI,
158
00:10:01.713 --> 00:10:05.486
biotech, and quantum computing. Now, what
this idea
159
00:10:05.486 --> 00:10:09.195
does not show: it measures invention
networks,
160
00:10:09.420 --> 00:10:13.049
not the whole business. An investor can
fund a startup in another
161
00:10:13.049 --> 00:10:16.629
country without any shared patent. A
company
162
00:10:16.886 --> 00:10:20.322
can license a technology or hire someone
remotely
163
00:10:20.659 --> 00:10:24.320
without a shared patent. And a part of the
American picture
164
00:10:24.770 --> 00:10:28.431
may be that More US companies have teams
in several hubs
165
00:10:28.848 --> 00:10:32.445
under one parent company, so the investors
in different cities
166
00:10:32.782 --> 00:10:36.555
show up as links. The study names possible
167
00:10:36.571 --> 00:10:40.393
causes— national rules, language,
different legal and of
168
00:10:40.409 --> 00:10:44.246
course tax frameworks— but it does not
provide which one matters
169
00:10:44.343 --> 00:10:48.084
most. So I will not tell you this number
proves where
170
00:10:48.277 --> 00:10:51.776
every European bridge breaks. What I will
say
171
00:10:52.242 --> 00:10:55.885
is this: the link between European hubs
172
00:10:56.238 --> 00:11:00.025
gets weaker as you move from research
towards commercial
173
00:11:00.025 --> 00:11:03.299
invention that is worth testing in the
174
00:11:03.380 --> 00:11:06.734
parts of company the patent data cannot
see.
175
00:11:08.114 --> 00:11:11.918
Now the other side in its strongest
176
00:11:12.030 --> 00:11:15.866
form: maybe Europe's real problem is
177
00:11:16.058 --> 00:11:19.557
too few hubs have enough growth capital,
enough large
178
00:11:19.733 --> 00:11:23.215
customers, and enough experienced leaders.
Connecting them more
179
00:11:23.519 --> 00:11:27.129
will not create resources that are missing
everywhere.
180
00:11:28.075 --> 00:11:31.620
And a founder should choose the best
partner for the
181
00:11:31.620 --> 00:11:35.422
company wherever it is—in your own city,
in
182
00:11:35.470 --> 00:11:39.095
another European hub, or in the US. I
think that
183
00:11:39.095 --> 00:11:42.576
objection is partly right. If the resource
does
184
00:11:42.768 --> 00:11:46.394
not exist anywhere in Europe, no bridge
will help you. And the
185
00:11:46.442 --> 00:11:50.053
evidence today does not settle whether
Europe needs a
186
00:11:50.053 --> 00:11:53.763
few much stronger hubs or just better
links
187
00:11:53.763 --> 00:11:57.537
between many. Probably both, depending a
little bit on the
188
00:11:57.537 --> 00:12:01.247
sector. But here is where I part ways with
it.
189
00:12:01.793 --> 00:12:05.455
When a resource does exist in Europe, the
extra costs of
190
00:12:05.567 --> 00:12:09.132
reaching it across a border is real and is
paid by the
191
00:12:09.165 --> 00:12:13.003
founder. The EU's own research on venture
192
00:12:13.115 --> 00:12:16.857
and growth capital published in October
2025 describes
193
00:12:17.210 --> 00:12:20.646
Europe's market as still fragmented and
lists
194
00:12:21.625 --> 00:12:25.367
legal, tax, and market barriers to
investing
195
00:12:25.415 --> 00:12:28.835
across borders and to raising money from
pension funds and
196
00:12:28.851 --> 00:12:32.640
insurers. More money behind the same
borders would make
197
00:12:32.704 --> 00:12:36.044
every hub richer. It would not
automatically
198
00:12:36.397 --> 00:12:40.154
make them easier to reach. So the use for
gold is not
199
00:12:40.170 --> 00:12:43.568
connected. activity for its own sake, but
it's building valuable
200
00:12:43.696 --> 00:12:47.413
capabilities and making them cheaper to
reach. We should
201
00:12:47.606 --> 00:12:51.243
judge the links by what they let a company
do.
203
00:12:55.537 --> 00:12:59.046
So let us get practical. Think of 5 areas
where your
204
00:12:59.062 --> 00:13:02.603
company may need something from outside
your home hub:
205
00:13:03.003 --> 00:13:06.624
capital, people, customers, know-how, and
your company
206
00:13:06.624 --> 00:13:10.391
structure itself. For each, here's what we
know, what
207
00:13:10.439 --> 00:13:14.126
we do not know, and the decision I would
take.
208
00:13:15.328 --> 00:13:18.631
Keep in mind, this is no legal or
financial recommendation. Yeah,
209
00:13:19.592 --> 00:13:23.087
take it at your own risk and do your
homework. Capital.
210
00:13:23.584 --> 00:13:27.127
What we know: the EU's own capital
research finds
211
00:13:27.207 --> 00:13:30.622
barriers to cross-border investments and
to fundraising from large
212
00:13:30.670 --> 00:13:34.453
institutions. What we do not know from
public data is how
213
00:13:34.662 --> 00:13:38.111
Often a good round fails because of a
border.
214
00:13:38.592 --> 00:13:42.426
If I were a founder, before your next
round, look beyond
215
00:13:42.426 --> 00:13:45.715
your home network for investors who
understand your sector and
216
00:13:45.892 --> 00:13:49.582
can back your next stage. Ask
217
00:13:49.582 --> 00:13:53.368
every fund 2 questions: does your mandate
let
218
00:13:53.368 --> 00:13:56.769
you invest in my country, and how much do
you keep in
219
00:13:56.865 --> 00:14:00.700
reserve for follow-on rounds? A fund that
loves you
220
00:14:00.732 --> 00:14:04.291
but cannot follow you is not your lead.
People—
221
00:14:06.112 --> 00:14:09.720
what we know: Dealroom reported for Europe
and Israel in
222
00:14:09.752 --> 00:14:13.198
2022 that 56% of startups founded by
223
00:14:13.278 --> 00:14:16.901
unicorn alumni were started in the same
city as the
224
00:14:16.901 --> 00:14:20.475
unicorn. What that does not tell us is how
225
00:14:20.539 --> 00:14:24.368
far those founders' experience reaches. A
founder can stay in
226
00:14:24.480 --> 00:14:28.278
Berlin and still hire in Stockholm. Raise
in London and sell
227
00:14:28.278 --> 00:14:32.051
in France. Local founding and wide reach
can go
228
00:14:32.051 --> 00:14:35.776
together. If I were a founder: when you
hire
229
00:14:35.969 --> 00:14:39.453
your first leaders for the 50 to 500
employee
230
00:14:39.453 --> 00:14:43.098
stage, start from the capability gap, not
the postcode.
231
00:14:43.852 --> 00:14:47.449
Search beyond your home hub when the skill
is not there, and
232
00:14:47.529 --> 00:14:51.142
check early what it takes to employ
someone and give
233
00:14:51.239 --> 00:14:54.517
them equity in the countries you are
likely to hire from.
234
00:15:00.590 --> 00:15:03.738
Hey guys, thanks for sticking with us.
Let's keep going.
235
00:15:04.916 --> 00:15:08.517
A quick word on how this show works.
Startup
236
00:15:08.517 --> 00:15:12.352
Radio is funded by partners, and the
partners we like best are
237
00:15:12.352 --> 00:15:16.077
the ones who make things easier to reach
across borders. Funds
238
00:15:16.222 --> 00:15:19.763
that invest across Germany, Austria, and
Switzerland, corporates that buy
239
00:15:19.812 --> 00:15:22.968
from startups in other countries, regions
that want
240
00:15:22.968 --> 00:15:26.578
founders in Berlin, Munich, Vienna, and
Zurich to know
241
00:15:27.420 --> 00:15:30.703
what they offer. If that is you, your
listeners
242
00:15:31.204 --> 00:15:34.923
are the founders and investors we— you are
trying to
243
00:15:34.939 --> 00:15:38.151
reach. Find out how we work together at
startuprad.io/become-a-partner.
244
00:15:42.450 --> 00:15:45.796
Now back to the 5 areas. Customers.
245
00:15:46.535 --> 00:15:50.278
A German startup should not have to become
international
246
00:15:50.326 --> 00:15:54.133
just because its next big customer is
French. In practice,
247
00:15:54.502 --> 00:15:57.923
it often does. New contract habits, new
248
00:15:57.956 --> 00:16:01.698
procurement rules, and maybe new language
and a new sales
249
00:16:01.891 --> 00:16:05.682
team. How big is that friction? The IMF
estimates
250
00:16:05.730 --> 00:16:09.248
that barriers inside Europe work like a
tariff of around
251
00:16:09.376 --> 00:16:11.816
44% on goods and
252
00:16:12.153 --> 00:16:15.523
110% on services. The IMF
253
00:16:15.523 --> 00:16:19.279
itself calls that an upper bound, and
other
254
00:16:19.375 --> 00:16:22.890
economists dispute it. It is a model of
trade
255
00:16:22.986 --> 00:16:26.822
costs, not a bill— not a bill your company
pays, but
256
00:16:26.854 --> 00:16:30.241
it tells you the friction is not
imaginary.
257
00:16:31.011 --> 00:16:34.558
Here's a decision I would make in your
second market:
258
00:16:35.024 --> 00:16:38.781
choose it where you have the strongest
evidence of customer demand
259
00:16:39.054 --> 00:16:42.522
and a credible way to sell, not where the
train is shortest.
260
00:16:43.790 --> 00:16:47.596
A nearby market like Austria or
Switzerland can be a cheap place
261
00:16:47.708 --> 00:16:51.529
to learn, to win a pilot customer and test
your sales process.
262
00:16:52.059 --> 00:16:55.415
But do not mistake Germany and Austria and
Switzerland for scale.
263
00:16:55.817 --> 00:16:58.899
Together, they're still smaller than the
market an American
264
00:16:58.964 --> 00:17:02.496
competitor gets at home. You use the
corridor to
265
00:17:02.576 --> 00:17:05.194
learn, then go where the demand is.
266
00:17:06.591 --> 00:17:10.091
Know-how. This is where Europe could have
an advantage.
267
00:17:11.247 --> 00:17:14.426
No city has to be best at everything.
268
00:17:15.406 --> 00:17:19.243
Munich, Zurich, Eindhoven, Grenoble,
Stockholm, and Cambridge
269
00:17:19.757 --> 00:17:23.450
can each specialize, but specialization
only helps
270
00:17:23.546 --> 00:17:26.324
if companies can reach it, and the patent
data
271
00:17:26.517 --> 00:17:30.017
suggests I would suggest this is where
Europe is weakest. If I were a
272
00:17:30.129 --> 00:17:33.822
founder in deep tech, identify the
capability you
273
00:17:34.031 --> 00:17:37.307
are missing, then find the best partner to
supply it,
274
00:17:37.901 --> 00:17:41.321
whether that is your own university or a
team in another
275
00:17:41.353 --> 00:17:44.853
country. And when you do partner across a
border,
276
00:17:45.527 --> 00:17:49.220
sort out who owns the resulting IP before
the
277
00:17:49.220 --> 00:17:52.351
work starts, not after. Investors care
278
00:17:52.351 --> 00:17:55.958
about clean ownership. Of course, your
company
279
00:17:55.958 --> 00:17:59.758
structure. In Part 6, Capital Gravity, we
describe
280
00:17:59.823 --> 00:18:03.599
companies moving their parent company CEOs
and listing
281
00:18:03.777 --> 00:18:07.449
plans to America as they grow. Sometimes
282
00:18:07.676 --> 00:18:11.397
that follows customers, capital, or the
executives they need.
283
00:18:11.819 --> 00:18:15.337
Sometimes it is a missing European option.
From the outside,
284
00:18:15.596 --> 00:18:18.823
you often cannot tell which. If I were a
founder
285
00:18:19.283 --> 00:18:22.345
Once a year, ask yourself the Capital
Gravity questions.
286
00:18:23.531 --> 00:18:26.994
Where is our parent company? Where does
your
287
00:18:26.994 --> 00:18:30.713
CEO work? Where are your senior commercial
hires?
288
00:18:31.627 --> 00:18:35.362
Where do we expect to list? If the answers
are moving
289
00:18:35.474 --> 00:18:38.568
west, write down why. If the reason
290
00:18:39.466 --> 00:18:43.298
is that something exists in Europe but
costs too much to reach,
291
00:18:43.955 --> 00:18:46.891
that is exactly the friction this episode
is about.
293
00:18:53.266 --> 00:18:56.429
There's a good idea for making Europe's
spread-out map work.
294
00:18:57.520 --> 00:19:01.105
Dealroom has argued for treating the hubs
within a 4-hour
295
00:19:01.105 --> 00:19:04.868
train ride of London as one supercluster,
296
00:19:05.368 --> 00:19:09.176
which it calls the New Palo Alto. Global
Venturing
297
00:19:09.291 --> 00:19:12.926
reported that Dealroom Also sketched a
second one around Munich
298
00:19:13.344 --> 00:19:16.798
linking Berlin, Milan, Vienna, and the
Swiss
299
00:19:16.798 --> 00:19:20.574
cantons. Those are proposals. Drawing a
circle on a map
300
00:19:20.896 --> 00:19:24.318
does not mean a company can use what is
inside
301
00:19:25.057 --> 00:19:28.560
it on the same terms. For our listeners,
the real
302
00:19:28.672 --> 00:19:32.400
test case sits across Munich, Vienna, and
Zurich,
303
00:19:32.930 --> 00:19:36.497
2 of Europe's top 5 innovation regions, 3
countries,
304
00:19:36.722 --> 00:19:40.365
And Switzerland outside the EU. The
network test.
305
00:19:41.091 --> 00:19:44.670
So instead of asking every region to build
its own
306
00:19:44.670 --> 00:19:48.150
Silicon Valley, here's a test you can run
on your own company.
307
00:19:48.975 --> 00:19:52.822
Let's call it the network test. For the
next
308
00:19:52.870 --> 00:19:56.375
capability your company needs, find the
best option at home
309
00:19:56.666 --> 00:20:00.381
and the best option elsewhere. Can you
actually use the outside
310
00:20:00.381 --> 00:20:04.012
option? What extra time costs or
311
00:20:04.076 --> 00:20:07.623
eligibility rules come with it, and what
would
312
00:20:07.687 --> 00:20:11.186
change for your company if that friction
disappeared?
313
00:20:11.539 --> 00:20:15.118
A few rules so the test means something:
only count
314
00:20:15.230 --> 00:20:18.761
things your company actually needed. A
specialist fund
315
00:20:18.761 --> 00:20:21.425
saying no to a company that does not fit
316
00:20:22.436 --> 00:20:26.111
is not a border problem. Moving the whole
company
317
00:20:26.464 --> 00:20:30.234
should not be the only way to get access.
And count
318
00:20:30.298 --> 00:20:33.391
the cases where it worked too. To be
clear,
319
00:20:34.160 --> 00:20:37.878
this is a test we are proposing, not
320
00:20:37.990 --> 00:20:41.483
one we have validated. So here is why I
321
00:20:41.611 --> 00:20:45.425
ask, if you have run into this around
322
00:20:45.762 --> 00:20:49.159
a hire, a customer, or partnership that
worked or
323
00:20:49.159 --> 00:20:52.733
failed because it sat in another hub, send
it to
324
00:20:52.829 --> 00:20:56.576
partnerships@startuprad.io. I'll use real
cases to
325
00:20:56.608 --> 00:20:59.239
test this framework in a later episode of
this series.
326
00:21:01.208 --> 00:21:04.953
Investors, briefly.
327
00:21:05.548 --> 00:21:09.291
if I were running a fund in Germany,
Austria, or Switzerland, I
328
00:21:09.323 --> 00:21:13.100
would build co-investment relationships
with funds in other hubs
329
00:21:13.325 --> 00:21:17.102
early so you have partners for later
rounds you cannot
330
00:21:17.118 --> 00:21:20.300
carry alone. If you are a limited partner
331
00:21:20.747 --> 00:21:24.451
ask your funds which follow-on rounds they
had to pass
332
00:21:24.531 --> 00:21:28.220
on because of a border. If you are a
corporate, the
333
00:21:28.220 --> 00:21:31.603
cheapest improvement you can make is a
buying process that a
334
00:21:31.603 --> 00:21:34.313
startup from the next country can
actually get through.
335
00:21:35.773 --> 00:21:39.413
EU Inc., the planned
336
00:21:39.445 --> 00:21:43.197
single company form for startups across
Europe, is the most
337
00:21:43.213 --> 00:21:47.030
direct fix for the company structure
problem in this episode. Its supporters
338
00:21:47.528 --> 00:21:51.093
want the first company founded under it in
2027. My
339
00:21:51.093 --> 00:21:54.835
on the record, and I first made it in part
340
00:21:54.835 --> 00:21:58.497
6: no startup will be able to register as
an EU
341
00:21:58.850 --> 00:22:02.399
Inc. in any EU country before January 1st,
342
00:22:02.640 --> 00:22:06.446
2028. If a registry accepts one before
343
00:22:06.479 --> 00:22:10.188
that, I was wrong. My reason: a new
company form has to
344
00:22:10.188 --> 00:22:13.770
get through EU governments and the
European Parliament
345
00:22:13.964 --> 00:22:17.791
And then every single national registry,
tax office,
346
00:22:18.147 --> 00:22:21.877
and notary has to be ready to handle it.
Each step
347
00:22:22.103 --> 00:22:24.182
takes time. Confidence: 75%.
348
00:22:26.212 --> 00:22:29.950
And one more. To be precise, this
forecast's
349
00:22:30.015 --> 00:22:33.746
wording in a report, not the state of the
market.
350
00:22:34.517 --> 00:22:38.133
The next edition of the EU Startup and
Scale-Up Index will
351
00:22:38.133 --> 00:22:41.142
again name late-stage funding As a
constraint.
352
00:22:41.880 --> 00:22:45.570
Confidence: 85%. If
353
00:22:45.586 --> 00:22:49.308
no new edition appears by the end of 2027,
that one
354
00:22:49.357 --> 00:22:53.143
stays open. Europe has
355
00:22:53.143 --> 00:22:56.945
spent 2 decades building strong startup
and innovation regions.
356
00:22:57.443 --> 00:23:01.133
In many places, it worked. The next
question is a
357
00:23:01.133 --> 00:23:04.759
different one. When a company needs
something that exists
358
00:23:04.983 --> 00:23:08.709
elsewhere in Europe, How hard is it to
use? The data says
359
00:23:08.918 --> 00:23:12.596
the link between Europe's hubs is close to
American levels in
360
00:23:12.789 --> 00:23:16.291
research and much weaker in commercial
invention.
361
00:23:17.046 --> 00:23:20.451
That does not explain the whole scale-up
gap, but it
362
00:23:20.499 --> 00:23:24.001
points to a cost founders pay every day
and one
363
00:23:24.338 --> 00:23:27.872
that policy can reduce. Europe has its
364
00:23:28.032 --> 00:23:31.822
islands of excellence. The question is,
how cheap it
365
00:23:31.822 --> 00:23:35.590
becomes to travel between them. There's
more to this story. For
366
00:23:35.590 --> 00:23:39.261
Entrepreneurs Vault members, I go one
level deeper. I run the
367
00:23:39.261 --> 00:23:42.948
network test on the Munich, Vienna, and
Zurich corridor area
368
00:23:43.044 --> 00:23:46.475
by area as a desk assessment. Members get
a
369
00:23:46.475 --> 00:23:50.226
city-by-city table of where unicorn alumni
founded their next company,
370
00:23:50.739 --> 00:23:54.394
what it can and cannot tell you, and a
stage-by-stage
371
00:23:54.491 --> 00:23:58.242
checklist for founders deciding what to
reach for outside
372
00:23:58.274 --> 00:24:01.771
their home hub. The full Entrepreneurs
Vault analysis
373
00:24:02.059 --> 00:24:05.844
is available to members. Join either on
YouTube or
374
00:24:05.892 --> 00:24:09.517
on Substack. Both links are in the show
notes. Next week on Startup
375
00:24:09.517 --> 00:24:13.125
Radio, more interviews and news from
Germany,
376
00:24:13.173 --> 00:24:16.685
Austria, and Switzerland, and the European
scale-up question continues.
377
00:24:17.118 --> 00:24:20.695
This is Startuprad.io. That's all, folks.
Find more news, streams,
378
00:24:20.871 --> 00:24:24.507
events, and interviews at www.
Startuprad.io. Remember,
379
00:24:24.652 --> 00:24:25.443
sharing is caring.
