Join David Knibbe, CEO of NN Group and Kevin McLoughlin, Partner & Co-Founder of MTech Capital as they dive into the role that technology has played in transforming NN into a major insurance player.
Transcript
0:01
[Music]
0:10
david welcome to this itc fireside chat
0:14
today we’re going to be discussing the
0:16
role of technology in the transformation
0:18
of nn group
0:21
first let me start by introducing us
0:23
briefly
0:24
david knieby is ceo of nn group
0:29
nn is a multi-line insurer based in the
0:31
netherlands
0:32
it has leading market positions in
0:35
corporate pensions
0:37
life and pnc
0:39
the group includes the number five bank
0:41
in the netherlands
0:43
uh leading businesses in life and
0:45
pensions across central and eastern
0:47
europe
0:48
and
0:49
japan
0:51
and my name is kevin mclaughlin
0:53
i’m partner and co-founder of mtech
0:55
capital we’re a vc firm focused
0:58
exclusively on insure tech including
1:01
asset management
1:03
we invest across north america
1:06
europe including israel
1:08
and we have offices in santa monica
1:11
california and london
1:15
david
1:16
last month you had a capital markets day
1:19
where you refreshed and reconfirmed
1:23
the strategic goals operational
1:25
objectives of an end group
1:28
so this interview is perfectly timed
1:31
um
1:32
if i can sum up that strategy and a
1:35
phrase
1:38
your goal is for nn to become a
1:40
customer-centric and data-driven company
1:44
uh what role will technology play in
1:46
this transformation
1:48
yeah well hello uh kevin and
1:51
and hello everybody very good to be here
1:53
first of all
1:54
um yeah indeed we just launched our
1:57
uh our new strategy uh which we launched
1:59
at our at our capital markets day it is
2:01
a strategy around creating value for for
2:04
all stakeholders
2:05
but the content of the strategy indeed i
2:07
think you summarize very well it’s very
2:09
much about being a customer-centric
2:11
organization and and very data-driven um
2:15
as well now clearly technology i mean
2:18
it’s obvious will play a crucial role uh
2:21
in that and maybe it’s helpful if i give
2:23
some examples i think part of the
2:25
strategy for example what we’re doing is
2:26
that we want to make sure that as an
2:28
insurance company we are really at the
2:30
at the forefront of our interactions
2:32
with customers i think at the end of the
2:34
day as an insurance company you need to
2:36
decide where you want to be in the value
2:37
chain and we want to be very much in the
2:39
forefront of our of all the interactions
2:41
with customers but if you want to do
2:43
that you have to understand your
2:44
customers you need to know them well and
2:46
then of course data and technology plays
2:48
a crucial role
2:50
in order to be able to know what’s going
2:52
on with our customers and how we can uh
2:53
can service them best
2:55
um so that’s an example where technology
2:57
is crucial
2:59
another example in our strategies of
3:00
course is platforms and and i’m sure
3:02
people talk a lot about it
3:04
we are very active uh particularly in
3:07
workforce management and carefree
3:08
retirement and again technology plays a
3:11
plays a crucial role in all of these all
3:13
of these new developments so i believe
3:15
in three years we’ll have a
3:16
fundamentally different company but a
3:19
lot more data and technology driven than
3:21
than today
3:23
and when you think about uh the
3:26
near-term priorities for transforming nn
3:29
with technology what is what is your
3:32
emphasis
3:33
is it you know life pensions pnc or in
3:36
terms of the value chain is it
3:38
distribution
3:39
you know uh policy administration
3:41
underwriting claims
3:44
yeah i know i think there’s there’s
3:45
indeed many topics that uh as a as a
3:48
broad company as a multi-line company
3:49
that uh that we are that we look at but
3:52
one important element for example is
3:54
scale
3:55
uh we have recently required delta lloyd
3:57
uh in the netherlands and uh
3:59
an even more recent fifa which made his
4:01
market leader in in both life and
4:03
non-life now to be able to to really
4:05
let’s say cash in on that scale
4:08
technology plays a crucial role if we’re
4:09
not aligning our platforms if we’re not
4:11
bringing our customers all to one
4:13
platform and if we don’t use all the
4:15
data that we’re getting to the best way
4:17
possible we’re not really cashing in on
4:19
on the scale so that is a
4:22
very important element another important
4:24
element is our core business if you will
4:26
especially in the international markets
4:28
is around life protection
4:30
um
4:31
now clearly underwriting and data and
4:34
and pricing and underwriting become more
4:36
and more important more and more is
4:38
possible risk selection i mean i don’t
4:40
think we’re at the level where we can
4:42
price each risk individually for each
4:44
customer but uh we’re clearly on our way
4:47
in that uh in that direction and again
4:49
data and technology uh are an important
4:52
part there uh and the more we can
4:54
separate ourselves in terms of our
4:56
underwriting capabilities
4:58
the better so so these are important
5:01
elements uh
5:02
that are part of our strategy where we
5:04
want to create long-term value
5:07
and these are elements that are crucial
5:09
for the coming years to be successful
5:13
and just
5:14
staying for a minute on this topic of
5:16
where with an nn group you see the scope
5:19
for greatest value creation and
5:21
transformation by technology at mtec
5:23
capital for example
5:25
um we see
5:27
50 startups we meet with 50 startups per
5:30
month on average
5:32
the vast majority of these are focused
5:34
on pnc
5:36
so do you see kind of your your greater
5:39
scope for the adoption of technology in
5:41
the pnc side of your business than life
5:43
or not
5:44
yes it’s a good question the um and
5:47
indeed there is a a difference in in the
5:50
application of in the life and in the
5:51
non-life i i would argue that it’s not
5:54
the case that there
5:56
is more value on the non-life setting on
5:58
the live side it’s just more visible if
6:00
you look at a lot of the startups a lot
6:02
of is around uh you know these these
6:04
startups around whether it’s around
6:05
claim handling it’s about underwriting
6:07
and there’s a lot of front-end
6:08
applications and they tend to be
6:10
on the non-life side on the pnc side i
6:13
think also often because it’s easier to
6:15
get customer interaction on these type
6:17
of
6:18
products than on the on the live site
6:20
and uh but there’s clearly a lot of
6:22
value in it if i look at the live site
6:24
and we are a very large player in in
6:26
life insurance and in pensions um there
6:29
is enormous value however it’s often a
6:31
bit less sexy which is migration of
6:33
legacy platforms so we migrate
6:36
everything to new platforms with the
6:37
technology that we now have i remember
6:39
when i was running the live company
6:41
quite a while ago
6:42
i mean we would try to do conversions
6:44
and every weekend we would maybe get 60
6:46
70 000 policies and and we’ve now had
6:49
even remotely in those in these corona
6:51
days where we’ve been doing policy
6:53
migration from home
6:54
and literally a million policies go from
6:56
one platform uh uh to the other so the
6:59
technology there is very important
7:02
another element on the live side where i
7:03
see a big difference is we used to have
7:05
kind of a screen scraping where you know
7:08
you could have certain
7:10
amendments on policies you could do via
7:12
screen scraping but every time something
7:14
in the process changed you know the
7:16
thing didn’t work anymore and now we
7:17
have robots that in a way essentially
7:20
kind of learn and think that like some
7:22
of our employees do and say okay maybe
7:24
this this cell moved a bit but i
7:26
understand what the purpose is so i’ll
7:28
i’ll adapt it myself
7:30
so it’s a lot of applications also on
7:32
the live side that create a lot of value
7:34
it’s just a bit less visible maybe it’s
7:36
potentially maybe a bit less less sexy
7:38
but in terms of value creation i
7:40
certainly wouldn’t underestimate it then
7:44
and kind of
7:46
we can focus on life if you wish but
7:48
just when you think about
7:50
where you’ve introduced technology
7:52
already where do you see kind of the
7:54
greatest where have you seen the
7:56
greatest benefits already and what has
7:58
not worked
8:00
yeah so i think we’ve seen a lot of
8:02
benefit already on the customer side uh
8:04
let me take for example our home market
8:05
as an example we now have close to seven
8:08
million customers uh which is you know
8:11
very substantial in in a market of 17
8:13
million uh 17 million people um they’re
8:17
all in one database and one of the
8:18
things that we’ve done is build an
8:20
engine around next best actions so
8:22
basically based on everything that we
8:24
know enriched with external data
8:27
and peer comparisons uh we’re getting
8:29
better and better at predicting towards
8:31
our customers what is the most logical
8:34
next step that this customer takes and
8:36
people often immediately think ah that’s
8:38
that’s cross seller deep sell but a lot
8:39
of it is actually servicing it could be
8:42
your car is overinsured maybe
8:44
the car is still all risk insured and
8:46
given you know the age of the car you
8:48
know you could have a a more simple
8:50
coverage uh maybe people don’t realize
8:53
that their children are becoming either
8:54
18 or they’re going to college and and
8:56
some things are changing in their uh in
8:58
their policies so a lot is also around
9:00
servicing where we’ve gotten much better
9:02
at it if we get it right whether it’s
9:04
via chat via it’s where they’re
9:07
calls so all the only channel
9:09
servicing that we do where we’re getting
9:11
better at is getting the right offer to
9:13
the customer what we’ve seen certainly
9:15
in the beginning if we get it wrong i
9:17
mean customers are very harsh so if you
9:19
you come up with a suggestion that is
9:21
wrong
9:22
you see it immediately backfires in your
9:25
mps and you’re worse off versus doing
9:27
nothing and so the learning has been
9:29
that we better really know what we’re
9:31
doing because if we do it well we get a
9:33
plus in our net promoters scores
9:36
if we get it wrong you really get uh get
9:38
punished for it so there’s been some
9:40
real value in terms of customer loyalty
9:42
in terms of nps but it’s it’s a longer
9:44
road to
9:46
to get it right another example is uh
9:49
we’ve used a lot of technology and again
9:51
this is probably
9:52
uh yeah might sound less less sexy than
9:55
than some of the other applications is
9:57
around email classification so we
9:59
literally get emails uh every day and we
10:02
get a lot of emails in and in a large
10:04
company a multi-line you don’t always
10:07
exactly know where these emails will go
10:09
and and how to classify them the first
10:11
step we’ve been doing is like on a fully
10:13
automated basis you get immediately the
10:15
email to the right person in the company
10:18
um that was step one
10:20
took quite some time to get that right
10:22
now we’re actually getting even better
10:23
because now we can go into the email and
10:26
already based on on what is in the
10:28
emails the suggested answers uh
10:31
is now the next phase that that we’re
10:33
working on now the amount of efficiency
10:36
that you create by these type of simple
10:38
things are
10:40
in my view are enormous certainly if you
10:42
have a large inforce book like like we
10:44
do
10:46
excellent um
10:50
of course insurance as an industry has
10:52
always been
10:54
driven by data
10:56
data and data analytics
10:58
but of course with ai including machine
11:01
learning nlp ocr
11:04
um
11:05
we have
11:06
you know tools tools that we can work
11:09
with
11:10
right across the value chain and
11:11
insurance
11:12
what’s nn’s general approach for using
11:15
ai and data science
11:17
right right yeah so we i think we’ve
11:19
been at the trying to be at the
11:21
forefront of all the uses of ai and i
11:23
think we’ve probably made you know the
11:27
mistakes that everybody did uh um you
11:30
know you you’d hope you avoid them but i
11:32
think quite a few of the mistakes
11:35
that generally are made
11:37
uh we also made which was
11:39
we started also with a separate team
11:42
that was doing this with some very
11:44
bright people and were really good at it
11:46
and
11:47
in hindsight we spent a lot of time on
11:50
getting the data in order getting the
11:52
right technology
11:54
trying to work with the technology
11:56
spend quite some time on all the
11:58
algorithms and what works and what
12:00
doesn’t work and then usa use cases came
12:02
out and i think that’s a model that
12:05
to be honest really doesn’t work that
12:06
well uh because there’s a couple of
12:08
issues uh with it in hindsight and that
12:11
is what we have changed now
12:13
i think the technology you know it’s
12:15
there the algorithms i think most of it
12:17
was probably invented by by
12:18
mathematicians in the 60s so the real
12:21
challenge is not that the real challenge
12:23
is getting you know the business
12:24
application of it how do you get large
12:26
organizations to actually adapt to it
12:29
and incorporate it instead of either
12:32
fighting it or proving it they don’t
12:33
need it or proving that the existing
12:35
model is is better or saying well it’s
12:38
great that you have these 10 brilliant
12:40
people there but i’m today really too
12:42
busy because i’m already launching a
12:44
product or launching a new platform and
12:46
i really don’t have time to to spend on
12:48
these type of things and i think that’s
12:51
been the challenge that that we also had
12:53
we’ve really converted it and turned it
12:55
around now so uh the overall goal is to
12:58
spend at least 70 of our time on
13:00
actually on use cases and maybe only 20
13:03
of our time on the actual algorithms and
13:05
on the
13:06
let’s say on the technology
13:09
and then now we’re starting to see that
13:11
even though we started with a lot of
13:13
cross-sell and other activities and now
13:16
you see so the example of the emails or
13:18
the robots are typically things that the
13:20
operational people really really needed
13:22
they add a lot of value
13:25
and at the same time uh you know the the
13:27
people that are now working in the ai
13:29
and in the the data science you know
13:31
suddenly now get flooded with requests
13:33
because people starting to see the um
13:35
you know the added value but to be
13:36
honest it’s been a journey we started i
13:38
think more in the classical way but
13:40
because we started early i do think
13:41
we’re well on our way now to to really
13:43
benefit from it
13:45
and in just in maybe a word on
13:48
to the extent you have found the um
13:51
perhaps a a critical obstacle being
13:54
corporate culture and just receptivity
13:56
to technology what specifically did nn
13:59
do to address that
14:01
i think the uh i’m not going to say
14:04
we’re there because this will continue
14:05
to be a
14:06
a challenge and but i think a lot
14:10
uh it helps a lot is as a senior
14:13
management you spend time on so that you
14:15
spend time talking to these people but
14:17
also talking to let’s say the business
14:19
units and the people in operations and
14:21
making sure that you continue to talk
14:23
about it and make it important
14:25
uh i think that’s one topic the other
14:27
topic is it’s very tempting after the
14:29
first attempt that we had where use
14:31
cases were built and didn’t work to to
14:33
be overly critical on it and explain to
14:35
everybody why it didn’t work and uh
14:38
but you should really avoid in my view
14:40
uh the finger pointing and and try to
14:43
get uh over that point as quickly as
14:46
possible and then obviously build on
14:48
success you know the small examples that
14:50
we had in the
14:52
in the beginning um you know i tried to
14:55
use as much as possible in town halls
14:56
and i i try to take it away from you
14:59
know this department is doing better
15:00
than the other one but just to to in a
15:03
positive way reinforce everybody that
15:05
this is the
15:06
uh this is the way forward um i think
15:08
we’re making progress there but like any
15:11
anything new uh it’s also fair to say
15:14
that we have uh i still see a lot more
15:16
opportunities than than we do today i
15:18
mean we just started we’ve been
15:19
working with the model towards customers
15:22
and so lessen the operations more
15:24
towards customers
15:26
well talking to customers on some of the
15:28
analytics that you do given privacy
15:30
concerns data concerns is again a
15:33
separate expertise so now we’re looking
15:35
at how do you interact with customers
15:36
once you have all this this data and you
15:39
think you have added value for them so
15:40
it’s also for us it’s still a very much
15:42
a learning journey
15:45
excellent um one other question um
15:49
which all insure techs want to know the
15:51
answer to um
15:54
uh
15:56
most of the insured techs that we
15:57
encounter are not out to disrupt uh to
16:01
displace incumbent insurers but rather
16:04
actually to service them through
16:06
products or or
16:08
technology different products and
16:10
services
16:11
um
16:12
having said that it can be a challenge
16:14
for uh these companies to work with
16:16
large complex insurance groups uh how do
16:19
you approach
16:20
at nn working with uh startup companies
16:24
why is that by the way kevin why are
16:25
they not trying to disrupt us that’s uh
16:27
sounds like a lot more fun than trying
16:29
to service these large insurance
16:31
companies uh
16:33
the vast majority there are certainly
16:34
some out there looking to disrupt and i
16:36
would say um
16:39
let’s say i would say to enhance and
16:42
improve
16:43
aspects of your business
16:45
and then provide that to insurance
16:48
incumbents is what we see more of than
16:51
the very large business models you know
16:53
the full stack insurers or
16:56
who are uh are out to displace you
16:59
so i think it’s uh
17:00
it’s we see both
17:02
but the vast majority i think are in the
17:04
business of probably servicing insurance
17:07
incumbents
17:08
right all right all right yeah so i
17:10
think going back to your question yeah i
17:12
mean clearly i mean there’s many times
17:13
that we are jealous of these intro texts
17:15
but i think the amount of innovation and
17:18
agility and and creative thinking uh
17:21
original thinking is something that uh
17:24
you know i would like to see in an end
17:25
all the time and and uh you know
17:28
we don’t always have that at the level
17:30
that some of the insured techs have and
17:31
i think that is tremendously valuable at
17:34
the same time we you know in a large
17:36
company we have a strong brand we have
17:38
capital we have a lot of customers we
17:40
have a lot of interaction and that’s
17:41
something that that obviously is very
17:43
valuable for any company and also for um
17:46
for insurer tax
17:48
i think the experience so far has been
17:50
that we have been working a bit more
17:52
with scale-ups than with really startups
17:55
and the reason being is not so much that
17:57
the ideas
17:59
yeah that we see an insure tax or in
18:01
really the startups are not as good as
18:03
the scale ups but a lot has to do also
18:05
with the scale i mean we’re a fairly
18:07
large organization and we’ve seen that a
18:09
lot of the you know the applications and
18:12
the ideas that people have um
18:15
become more complex
18:17
you know they’ve tested it with a
18:18
hundred with a thousand but if you
18:20
suddenly test it with millions of
18:22
customers of lots of operational streams
18:24
uh things become a lot more complex and
18:28
there you see that on the one hand we
18:30
always want to build size as a company
18:31
and then we’ve been you know active in
18:33
that field too but at the same time this
18:35
size also creates a lot of complexity
18:37
and from that point of view our
18:38
experience with scale-ups is a bit
18:39
better not because their concepts are
18:41
always better but usually they’re a bit
18:42
better in dealing with the complexity
18:45
that large uh
18:47
large organizations have but we do have
18:49
uh we have quite a bit of partnerships
18:51
all over the place and it’s uh some of
18:53
it is uh
18:54
as an investment and some of it is uh um
18:58
you know because we already see some uh
19:00
possibility to apply it within our
19:02
organization um so we are actively
19:05
engaging and um
19:07
uh trying to figure out who we’re going
19:09
to try and work with and
19:11
or invest in
19:15
david kneebe ceo of nn group thank you
19:19
thank you pleasure to be here
19:28
you