October 24, 2025 • 27 mins
After years of cautious experimentation, CRE’s biggest players are finally scaling artificial intelligence.
JLL has been at the forefront of it, launching an in-house AI platform called Falcon that is cutting deal timelines from weeks to hours and automating the drudgery that once bogged down teams. The result is higher revenue per head, higher success rates when pitching and faster closings.
On this week’s show, CEO Christian Ulbrich gave a peek under the hood of how JLL is getting the most out of AI and how he sees it reshaping the industry.
It’s difficult to get an edge from AI, he said. Tools that drive productivity become table stakes within six months. The real differentiation only comes from new-to-market strategy, agentic AI that reinvents parts of CRE deal-making.
Still, the industry must act.
“Don’t wait too long,” he said. “The train has left the station and it is going at Japanese speed levels of train — very very fast.”
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Mark Bonner (00:08):
Welcome. I'm Mark Bonner, business owner's editor
in chief, coming to you livefrom New York. It's Friday,
October 24. Well, JLL has beenon quite the tear this year. In
Q1, the firm pulled in about$5,700,000,000 in revenue, up
roughly 13% from last year andprofits climbed nearly 30%.
Then in q two, revenues pushedpast 6,000,000,000, up another
(00:31):
10% with earnings per sharejumping close to 30% again. And
a market still grinding underhigh rates and slow deal flow.
Those are rare numbers thesedays. And all eyes are on JLL's
next earnings release coming ina little more than a week on
November 5. But the story behindthose numbers isn't just another
strong quarter.
(00:52):
It's about a company rewiringhow CRE is actually operating.
While most of the industry isstill teasing AI or testing AI,
I should say, JLL is alreadyscaling it. Its enterprise
platform Falcon sits on top ofdecades of proprietary real
estate data turning lease terms,valuations and forecasts into
(01:14):
real time insight. JLL GPT givesemployees generative tools that
cut work cycles from weeks tohours. Azera helps clients model
space and cost in minutes.
And the new property assistantlets landlords ask, where's
when's my NOI dropping? And thenthey get the answer immediately.
(01:35):
It's not a pilot. It's aplatform. And it's quietly
redefining what it means to be abrokerage and advisor and a data
company all at once.
So today we're asking how farcan JLL pull push this
transformation and what does itsay about where commercial real
estate is heading next? Tounpack it, we're joined by the
(01:56):
man himself who's driving thatshift coming to us live from
Frankfurt, Germany, JLL CEOChristian Ulbrich. Christian,
welcome to First Draft Live.
Christian Ulbrich (02:05):
Mark, thank you. What a what a brilliant
introduction. That was that wasvery kind of you.
Mark Bonner (02:12):
So let's talk about Falcon. That's become the center
of JLL's AI play built forcommercial real estate and
fueled by decades of deal andproperty data as we just
discussed. It powers tools likeJLL, JPT, GPT, Azera, and the
new property assistant. ButChristian, give us the quick
version here. What exactly isFalcon, and what is it doing
(02:33):
inside JLL today?
Christian Ulbrich (02:36):
Well, as it says, Falcon is a platform. And
so that's kind of the the backof the house for all our AI
tools. It pulls, the data in andthen it distributes the data to
the different applications wehave. And the number of
applications are are growing aswe speak, and therefore, it is
(02:56):
important that you you have theframework, the house around it,
and that's what LL's Falcon is.
Mark Bonner (03:02):
But can you walk me through the the system and how
it actually works? What does auser have to put into it? And
then how quickly can he getsomething out of it to make a
real difference in theirworkflow?
Christian Ulbrich (03:15):
Well, as you know, AI is is based on accurate
data. So the story started withus kind of creating a really
thorough data platform where allwhat we do is kind of, finding
its home. We work very much onthe accuracy of that. That is
(03:35):
the biggest enemy. And then wepull that data into that
platform and then we distributeit out to whatever application
we are running.
And and so the the people willonly see the application they
are using. They don't have to doanything else than using an
application like they use anyother software. But behind that,
(03:57):
that is all hosted under theJLL's Filecoin platform.
Mark Bonner (04:00):
So where are you seeing the first real wins?
Leasing, valuations, portfolioplanning?
Christian Ulbrich (04:07):
Well, you see wins pretty much across the
board. What is probably fair tosay that in most organizations
who are fully embracing AI, thefocus sits on productivity
gains. So we are making that thelife of our colleagues easier
because work which they had todone to do in the past, kind of
(04:30):
more handishly and pull the datatogether. This is now done by an
agent. And so we are giving themliterally time back and it's
moving from there now much moreinto client facing tool.
You mentioned some of them inyour intro that we now have
already tools which our clientsare using, which is making their
(04:51):
life easier. And that's kind ofthe usual journey which you will
see probably across many verylarge organizations.
Mark Bonner (04:59):
So let's get into the efficiencies that are
created by this in a second. Butyou mentioned accuracy, right?
That's critically important.Know, I mean, we've all read or
experienced ourselves that AIhallucinates. Maybe it's 90%
accurate, but that 10% canrepresent a huge pitfall if
you're not in control of it.
How do you guys navigate thatfrom a policy perspective
(05:20):
internally in terms of how totrust or to verify what the AI
gives back to your workforce?
Christian Ulbrich (05:28):
Yeah. I mean, that is that is a great point
because people sometimes say AIwill immediately take over
instead of the humans. It's it'sthe combination. People, It's
but they still have to kind ofcontinue what humans are good
at, and that is to be criticalto ask questions. And so when
(05:49):
you get a response from AI, youhave to look at that response
and see does that really makesense.
And I guess we all have playedaround with different systems
and in our private lives. Andsometimes you are amazed about
the accuracy of the response.And sometimes you look at the
response and you know wellthat's not quite true. And so
(06:11):
that is why it's so importantthat the proprietary data which
we have within JLL in the systemis as accurate as possible
because then that helpsobviously our people to build
that trust into the systems.
Mark Bonner (06:26):
Got it. So, you know, inside JLL, and we'll get
to the now we're gonna get tothe the efficiency part of this.
I know it's cutting cycle timesto almost zero task that took
weeks or months now are takingdays, if not hours. Document
review is probably a place wherethis is really helping your
teams, client reporting. All ofit is happening in real time
(06:47):
faster than ever, probably inyour career, Christian, not to
put words in your mouth.
But you mentioned the time thatit's giving back to your
workforce. What do people dowith that extra time?
Christian Ulbrich (06:59):
Well, the most important thing they do is
that they have more clientcontact. They spend more time
interacting with our clients.And that is what we want for
various reasons here. First ofall, it's always great when they
speak to clients but they havemore to talk to the client about
because they are based with alot of facts, data driven facts
(07:24):
which provide insights and thatinsights makes our clients much
more successful. So our peoplehave to bring that information
to our clients.
Mark Bonner (07:35):
So have you seen that play out in the real world?
Like has that actually occurredwith your employees? Are they
doing it? And what are have theoutcomes been stronger because
they have more client contact?
Christian Ulbrich (07:46):
Oh, absolutely. Absolutely. I mean,
we we see we see that theproductivity they are gaining is
turning into higher revenue perhead. It's turning into higher
success rates when they arepitching. It's turning into
higher or faster closings.
(08:08):
We have now 56% of the bidderswhich are generated from our AI
tools are the ones who areclosing the deal on the capital
market side. So we havetremendous amounts of KPI which
are proving the value of thetools and so that's why
(08:29):
everything is now going at thatvery rapid speed level because
it's in your face the benefitsfor your clients when you
embrace AI.
Mark Bonner (08:39):
Look. I mean, this is obviously a deep competitive
advantage that you all have.Right? One of the giants of
commercial real estate globally.There's probably gonna be a lot
of people who listen to thisconversation who don't work at
one of those giants.
There may be a middle marketfirm or a small local firm
somewhere in The United Statesor in Europe or Asia and they're
thinking, how can I keep up? Howcan I realistically compete with
(09:00):
the JLL's of the world who arewho are able to pull this off
for their teams? I mean, whatwould your advice be, Christian,
to the majority of thecommercial real estate industry,
especially in The United States,which is not which is more
middle market and are findingtrying to figure out a way, what
can we learn from the big boys,if you will, and how can we
(09:20):
apply that for our small 50 to a100 person unit somewhere in the
Midwest Of The United States,for instance?
Christian Ulbrich (09:27):
Well, first of all, what type of companies
are we talking about in the realestate industry? If you talk
about small investors or thosetype of developers, those type
of companies, they can easilypartner with us or with others
who have the data in house andand we are very happy to help
them to be successful as as ourclients. If you talk about
(09:49):
smaller competitors to us, Idon't necessarily there's a
question around whether you arebig or small because the
advantages of being smallbecause you can move faster,
it's probably easier to convinceyour your teammates how
important it is to embrace AI ina smaller, store than in a
(10:09):
larger outfit. But the thedisadvantage is access to data.
I think that is for us thebiggest advantage.
The amount of proprietary datawhich we are obviously
complementing with publiclyavailable data. Now a smaller
(10:31):
outfit doesn't have that muchproprietary data, but there's
still a lot of data out in thepublic. So they can still have
the opportunity to use AIeffectively.
Mark Bonner (10:41):
But as part of the so okay. But part of the answer
is they should they should workwith JLL?
Christian Ulbrich (10:47):
Well, others will probably do the same what
we do. I don't think what we dois totally unique. Maybe we we
went on that journey a bitearly, but this is nothing which
will where you have forever acompetitive advantage because
everybody will embrace AI.Otherwise, they will be out of
business relatively soon.
Mark Bonner (11:08):
So the other big thing that's out there is
something that everyone'stalking about. Agentic AI, which
basically means systems thatdon't just answer questions, but
actually take actionindependently. You're building
that into Falcon. A lot of otherof your competitors are as well.
What does the general commercialreal estate public make of that?
(11:29):
How close are we to AI runningfull workflows inside CRE or
even at GLO?
Christian Ulbrich (11:37):
Well, I would say the whole development is in
comparison to previousinnovations we have seen at
lightning speed level. And andpeople shouldn't make the
mistake that they think AI isjust for the simple task. I
would say it's the opposite. AIis starting to prove, agentic AI
is starting to prove that themost complex tasks are actually
(12:00):
the one where you have thebiggest advantage because with
that operational complexity, AIcan really drive additional
value. And so it you can almostthink about the most complex
property types will will createthe biggest kind of value when
(12:22):
you apply AgenTic AI.
And and that's now in thedevelopment, but it goes really
outrageously fast.
Mark Bonner (12:31):
What I mean, within JLL, what property types are
showing the biggest operationalgains so far when it comes to
agentic AI?
Christian Ulbrich (12:38):
As I said, the more complex, the better. So
you see obviously a lot ofadvantage around when you deal
with multifamily with all thesetenant interactions and all the
necessary kind of work whichneeds to be done in those
properties around maintenance.Those are the things where I can
(13:00):
really play it out. But you canalso think about industrial
buildings where you have thesupply chain coordination.
Literally, the more complex, thebetter.
Mark Bonner (13:13):
So, you know, we're talking about accuracy a few
minutes ago, right, in in adifferent context where there's
deep human intervention, as asthe guardrail or the fact
checker to make sure that it'saccurate, especially as you're
making big complex decisions orat least presenting things to
clients. On this, it's a littlebit more independent. What kind
of guardrails are in placeinternally to make sure that it
(13:34):
is accurate? Is it data quality?Is it compliance?
Are there ethical concerns thatyou all have discussed
internally on this front as asas you begin to really roll this
out?
Christian Ulbrich (13:46):
Yeah. I mean, obviously, everybody is is still
learning as we move forward, butthere are a couple of things
which we are particularlyconcerned about, most notably
the whole topic around cybersecurity. That's why we created
JLLGPT a closed system that ourdata doesn't leave the company,
(14:07):
that is very important to us andthat client data doesn't leave
the company, but that is a hugetopic around cyber. The other
challenge which is now comingup, you will have heard about
the term citizen development.This is moving so fast and you
cannot stop your talentedworkforce to kind of create
(14:32):
their own prompts which thenbecome a small little agent.
And so there's a massive,massive move towards citizen
development. Again, this iswithin JLL happening almost
exclusively within the frameworkof JLL GPT. But, you know,
(14:53):
people may bring their owndevice and then use a different
system. And that's where we haveto prevent that any kind of
corporate data is getting onother devices. So there are a
lot of guardrails in place, butit is something where we will
learn over the coming months andyears what kind of loopholes are
(15:17):
being created which become achallenge for that whole trend
around AI and the riskaccompanying it.
Mark Bonner (15:28):
If you're just joining us, this is First Draft
Live. We're talking with JLL CEOChristian Albrecht calling us
from Frankfurt, Germany abouthow the firm's Falcon platform
and AI strategy are reshapingglobal commercial real estate.
Christian, let's go to aquestion from the audience here.
Is the Falcon technology offeredto outside customers, or are
there plans to do so?
Christian Ulbrich (15:50):
Well, we have obviously offered it to our
clients very much in our workdynamics business. So you
mentioned it in the intro. Wehave clients now who used to ask
questions about their ownportfolio to a large group of
(16:10):
business analysts and then theywere trying to bring all the
information together and reportback a couple of days later. And
now they can just write theprompt into the tool which we
have provided to them and so theCIE lead will have the answer
instantly, which is fantasticfor them. It creates a bit of a
(16:32):
challenge for our industrybecause we have to now create
new ways of being rewarded forwhat we do.
We have to be paid for theoutcome. And so that is
something which is a change inthe industry and we are still at
the beginning to see thatdriving change into our
(16:52):
industry. But that is just oneexample. Are plenty of other
examples where we provide accessto clients to their portfolio
and they can look what otheropportunities similar to their
portfolio is being out there.And so again, this is just early
days, but it's coming everycouple of weeks, new products to
(17:14):
the market.
Mark Bonner (17:15):
You know, across capital markets AI is tightening
the deal clock. It's spend it'sspeeding up underwriting. It's
cutting through diligence. It'seven transitioning deal docs in
seconds. From your perspective,Christian, I mean, what capital
sources are leaning hardest intoAI assisted underwriting right
now?
Christian Ulbrich (17:33):
It's probably the investment management firms
and maybe even more privateequity. We saw some private
equity companies really doublingdown on their data accuracy over
the last couple of years andthey are probably now the ones
who are most embracing theopportunity AI is delivering. I
mean, we shouldn't forget at theend of the day, the real estate
(17:54):
business is a data business. DoI buy this building or do I buy
that building? Do I lease thisbuilding or do I lease that
building?
Those are not gut fields. Thoseare things which are supported
by facts, by data. And and sothere is no surprise that some
of the most successful realestate investors are the ones
(18:16):
who are embracing data the most,and they are the ones who are
now embracing AI the most.
Mark Bonner (18:21):
You have fierce competition everywhere around
the globe. And all of your yourfiercest competitors, which are
household names withincommercial real estate, are also
spending bootyles of cash onvery similar products. They can
speed up their their teams tocome to results faster and
obviously things that are moreclient facing. I guess the
question for you, Christian, isas you peer over the horizon
(18:42):
here, when everyone's fast andwhen everyone has this sort of
AI playbook, if you will, whatbecomes the new edge in this
world?
Christian Ulbrich (18:55):
Well, I I don't think that's a big change
to what it was before. Anyinnovation is being caught up by
others very quickly because whatwe can do, others will look at
it and they will try to do thesame. And if they don't do it,
then those type of companieswill be out of business. You
know, go back to whatever manyother examples, the the mobile
(19:17):
phone, you know, when we movefrom normal mobile phones to
smartphones, you know, the oneswho are still not providing
smartphones are just out ofbusiness, and it's the similar
trend is here. So I don't thinkthat any kind of AI tool which
is driving just productivitywill be in any form a game
(19:37):
changer.
It will become table stakeswithin three, six months. Where
there is potentially anopportunity is when you find a
way for a completely new go tomarket strategy. So there may be
agents which will completelyreinvent parts of our services
with a completely new approachto go to market. That then could
(19:59):
become a real differentiator.But there isn't anything out
there.
I guess everybody is working onthose. We are working very hard
on those kind of potentialagents, but nothing is at the
moment out there, which I wouldsee as a big game changer.
Mark Bonner (20:14):
You know, the headlines that we've all seen
here on another front has to dowith talent. Right? AI has been
rewriting job descriptions. It'sbeen cutting entry level roles.
It's been boosting output andchanging what junior, quote,
unquote, even means incommercial real estate these
days.
How do you balance thatautomation with while also
developing human talent?
Christian Ulbrich (20:35):
Yeah. I mean, there's a lot of talk out there
that entry level jobs willcompletely disappear and and and
and overall the amazing amountof potential job cuts. Let's
see. At the moment, it createsjobs. That doesn't mean that it
is always the same person, butat the moment it creates job.
That's the reason why SanFrancisco is coming back as a
(20:57):
real estate market really nicelybecause of all the AI jobs. So
going back to that questionaround analyst, I think what it
does is it provides smart youngpeople an ability to leapfrog in
their career experience in theirin their traction on gaining
knowledge and becoming a reallyvaluable member of a team much
(21:22):
earlier. What took three yearswill probably now only take six
to twelve months in thatjourney. And so overall, I think
for people with ambitions andthe willingness to learn, the
opportunities are only rising.The flip side of that is those
who are not that interested toreally embrace those new tools
(21:42):
and and go the extra mile, theymay find it more difficult to
have a career or even find arole going forward.
Mark Bonner (21:50):
You know, Gen Z, as we've discussed a lot on this
show over the last year, youknow, this is the most educated,
most tech enabled generation toever hit the global economy.
They're having a lot of hardthey're having a they're having
a hard time finding work rightnow even though they've went to
some of the top four yearschools, especially United
States and also in Europe andAsia. When you think about those
(22:13):
people, those are your futureleaders, for JLO and obviously
future leaders for commercialreal estate. What skills are now
must haves at JLO at that level?What is the fastest way to get
your resume to the top of theand for you to get a job at JLO
given this dynamic that we'rediscussing with artificial
intelligence?
Christian Ulbrich (22:34):
Well, that's a big question. I would say the
skill sets needed, is that youcan really communicate well,
that you can be part of a team.I think and that has not
necessarily anything to do withAI, but the world has become
incredibly complex. So it's nota world for solo fighters. It's
(22:58):
a world for team players.
And so people who can really bepart of a team and do their very
best for the overall success ofthe team are in a in a really
good position. And then withinthe team, you will have people
who are slightly more expert ona specific topic. Being
technology savvy, being AI savvyfor young people will be
(23:21):
completely table stakes, And andtherefore that will not be the
differentiator. But you can alsoturn that whole thing around. I
heard just yesterday a storythat some of the best
universities, their students arecreating their own agent while
they are going throughuniversity.
And then when they come to aninterview with a company, they
(23:42):
ask the company, what tech stackdo you have? Can I bring my own
agents Because that is my lifenow, my agent? It's my partner.
Because that's what people do.They use now agent as a partner.
So you're you're hiring you'rehiring Peter, but Peter brings
his agent. And so you get from acompany perspective almost two
(24:02):
for one.
Mark Bonner (24:05):
So, look, the bottom line is this. You know,
we're we're still in the veryinfancy of AI adoption in the
global economy, and certainly,that's true for commercial real
estate. We're in the pilotrefine scale phase right now.
The messy middle, I think whereevery firm's experimenting, some
are winning, some are not.Nobody's got the playbook down
(24:25):
just right.
And I think that includes JLO,right? You may be further along
on this journey than a lot ofother companies, but this is
still obviously being figuredout. And there's obviously a
Christian Ulbrich (24:35):
lot of
Mark Bonner (24:35):
opinions about AI. Right? In your opinion, what do
you think is the most commonoverreaction you're seeing
around AI adoption?
Christian Ulbrich (24:45):
Well, I think a lot of people kind of who are
maybe skeptics to AI are usingit for the first time. They are
bringing their data in, and theyare writing prompts, and they
are completely disappointed withthe outcome and then they blame
the AI for the outcome, but theyshould blame their own data
accuracy. You know this wholestory and it's voted many, many
(25:10):
times garbage in garbage out.That is coming to a new level
when you talk about AI. Andtherefore that is probably a
very common reaction from peoplewho are skeptics.
You have to do your homework ifyou want to take the benefit
from AI. And so that is, by theway, one of the big challenges
which every organization has,and I would say larger
(25:33):
organizations even more thansmaller organizations, that you
have to overcome your internalskeptics who wanna prove to you
that this is all rubbish andthey don't need to do that. And
and so that's why we have oursuccesses, but we also have our
failures. But very often, thefailures are kind of man made
failures within yourorganization.
Mark Bonner (25:54):
For the rank and file commercial real estate
leader, right, what is your bestadvice to them right now about
where they should start ifthey're still on the sidelines,
regarding experimenting with AIand trying to implement it into
workflows in their offices?
Christian Ulbrich (26:14):
Well, you know, I think it is important
that you are identifying two,three trials, and then you
really double down on them andtry to make them work. Don't
spread kind of yourself toothinly with twenty, thirty
different ideas, but nothing isreally getting the right
(26:34):
attention. Try out with two,three thoughtful ideas and then
go deep on them. But don't waittoo long. The train has left the
station, and this train is goingat kind of, I can say, Japanese
speed levels of trains really,really fast.
Mark Bonner (26:55):
Okay. That's our show for today, Christian. I'm
getting the hook here from myproducers, but I wanna thank you
very much for taking the time tobe with us today. I hope you
have a wonderful weekend.
Christian Ulbrich (27:03):
Thank you so much for having me, Mark. Have a
nice weekend too.
Mark Bonner (27:07):
If you missed any part of this conversation or
wanna catch earlier episodes,you'll find every show on
biznow.com or in your favoritepodcast feed. Just search First
Draft Live. Our conversationwith WeWork CEO, John Santora,
live from the Cretek Conferencein New York this week is now up
on the site and also in yourpodcast stream. And next week,
we're back with Industrious CEOand cofounder Jamie Haddari for
(27:30):
a look at where the flex officemarket goes next. This is First
Draft Live.
Have a great weekend y'all.
Welcome. I'm Mark Bonner, business owner's editor
in chief, coming to you livefrom New York. It's Friday,
October 24. Well, JLL has beenon quite the tear this year. In
Q1, the firm pulled in about$5,700,000,000 in revenue, up
roughly 13% from last year andprofits climbed nearly 30%.
Then in q two, revenues pushedpast 6,000,000,000, up another
(00:31):
10% with earnings per sharejumping close to 30% again. And
a market still grinding underhigh rates and slow deal flow.
Those are rare numbers thesedays. And all eyes are on JLL's
next earnings release coming ina little more than a week on
November 5. But the story behindthose numbers isn't just another
strong quarter.
(00:52):
It's about a company rewiringhow CRE is actually operating.
While most of the industry isstill teasing AI or testing AI,
I should say, JLL is alreadyscaling it. Its enterprise
platform Falcon sits on top ofdecades of proprietary real
estate data turning lease terms,valuations and forecasts into
(01:14):
real time insight. JLL GPT givesemployees generative tools that
cut work cycles from weeks tohours. Azera helps clients model
space and cost in minutes.
And the new property assistantlets landlords ask, where's
when's my NOI dropping? And thenthey get the answer immediately.
(01:35):
It's not a pilot. It's aplatform. And it's quietly
redefining what it means to be abrokerage and advisor and a data
company all at once.
So today we're asking how farcan JLL pull push this
transformation and what does itsay about where commercial real
estate is heading next? Tounpack it, we're joined by the
(01:56):
man himself who's driving thatshift coming to us live from
Frankfurt, Germany, JLL CEOChristian Ulbrich. Christian,
welcome to First Draft Live.
Christian Ulbrich (02:05):
Mark, thank you. What a what a brilliant
introduction. That was that wasvery kind of you.
Mark Bonner (02:12):
So let's talk about Falcon. That's become the center
of JLL's AI play built forcommercial real estate and
fueled by decades of deal andproperty data as we just
discussed. It powers tools likeJLL, JPT, GPT, Azera, and the
new property assistant. ButChristian, give us the quick
version here. What exactly isFalcon, and what is it doing
(02:33):
inside JLL today?
Christian Ulbrich (02:36):
Well, as it says, Falcon is a platform. And
so that's kind of the the backof the house for all our AI
tools. It pulls, the data in andthen it distributes the data to
the different applications wehave. And the number of
applications are are growing aswe speak, and therefore, it is
(02:56):
important that you you have theframework, the house around it,
and that's what LL's Falcon is.
Mark Bonner (03:02):
But can you walk me through the the system and how
it actually works? What does auser have to put into it? And
then how quickly can he getsomething out of it to make a
real difference in theirworkflow?
Christian Ulbrich (03:15):
Well, as you know, AI is is based on accurate
data. So the story started withus kind of creating a really
thorough data platform where allwhat we do is kind of, finding
its home. We work very much onthe accuracy of that. That is
(03:35):
the biggest enemy. And then wepull that data into that
platform and then we distributeit out to whatever application
we are running.
And and so the the people willonly see the application they
are using. They don't have to doanything else than using an
application like they use anyother software. But behind that,
(03:57):
that is all hosted under theJLL's Filecoin platform.
Mark Bonner (04:00):
So where are you seeing the first real wins?
Leasing, valuations, portfolioplanning?
Christian Ulbrich (04:07):
Well, you see wins pretty much across the
board. What is probably fair tosay that in most organizations
who are fully embracing AI, thefocus sits on productivity
gains. So we are making that thelife of our colleagues easier
because work which they had todone to do in the past, kind of
(04:30):
more handishly and pull the datatogether. This is now done by an
agent. And so we are giving themliterally time back and it's
moving from there now much moreinto client facing tool.
You mentioned some of them inyour intro that we now have
already tools which our clientsare using, which is making their
(04:51):
life easier. And that's kind ofthe usual journey which you will
see probably across many verylarge organizations.
Mark Bonner (04:59):
So let's get into the efficiencies that are
created by this in a second. Butyou mentioned accuracy, right?
That's critically important.Know, I mean, we've all read or
experienced ourselves that AIhallucinates. Maybe it's 90%
accurate, but that 10% canrepresent a huge pitfall if
you're not in control of it.
How do you guys navigate thatfrom a policy perspective
(05:20):
internally in terms of how totrust or to verify what the AI
gives back to your workforce?
Christian Ulbrich (05:28):
Yeah. I mean, that is that is a great point
because people sometimes say AIwill immediately take over
instead of the humans. It's it'sthe combination. People, It's
but they still have to kind ofcontinue what humans are good
at, and that is to be criticalto ask questions. And so when
(05:49):
you get a response from AI, youhave to look at that response
and see does that really makesense.
And I guess we all have playedaround with different systems
and in our private lives. Andsometimes you are amazed about
the accuracy of the response.And sometimes you look at the
response and you know wellthat's not quite true. And so
(06:11):
that is why it's so importantthat the proprietary data which
we have within JLL in the systemis as accurate as possible
because then that helpsobviously our people to build
that trust into the systems.
Mark Bonner (06:26):
Got it. So, you know, inside JLL, and we'll get
to the now we're gonna get tothe the efficiency part of this.
I know it's cutting cycle timesto almost zero task that took
weeks or months now are takingdays, if not hours. Document
review is probably a place wherethis is really helping your
teams, client reporting. All ofit is happening in real time
(06:47):
faster than ever, probably inyour career, Christian, not to
put words in your mouth.
But you mentioned the time thatit's giving back to your
workforce. What do people dowith that extra time?
Christian Ulbrich (06:59):
Well, the most important thing they do is
that they have more clientcontact. They spend more time
interacting with our clients.And that is what we want for
various reasons here. First ofall, it's always great when they
speak to clients but they havemore to talk to the client about
because they are based with alot of facts, data driven facts
(07:24):
which provide insights and thatinsights makes our clients much
more successful. So our peoplehave to bring that information
to our clients.
Mark Bonner (07:35):
So have you seen that play out in the real world?
Like has that actually occurredwith your employees? Are they
doing it? And what are have theoutcomes been stronger because
they have more client contact?
Christian Ulbrich (07:46):
Oh, absolutely. Absolutely. I mean,
we we see we see that theproductivity they are gaining is
turning into higher revenue perhead. It's turning into higher
success rates when they arepitching. It's turning into
higher or faster closings.
(08:08):
We have now 56% of the bidderswhich are generated from our AI
tools are the ones who areclosing the deal on the capital
market side. So we havetremendous amounts of KPI which
are proving the value of thetools and so that's why
(08:29):
everything is now going at thatvery rapid speed level because
it's in your face the benefitsfor your clients when you
embrace AI.
Mark Bonner (08:39):
Look. I mean, this is obviously a deep competitive
advantage that you all have.Right? One of the giants of
commercial real estate globally.There's probably gonna be a lot
of people who listen to thisconversation who don't work at
one of those giants.
There may be a middle marketfirm or a small local firm
somewhere in The United Statesor in Europe or Asia and they're
thinking, how can I keep up? Howcan I realistically compete with
(09:00):
the JLL's of the world who arewho are able to pull this off
for their teams? I mean, whatwould your advice be, Christian,
to the majority of thecommercial real estate industry,
especially in The United States,which is not which is more
middle market and are findingtrying to figure out a way, what
can we learn from the big boys,if you will, and how can we
(09:20):
apply that for our small 50 to a100 person unit somewhere in the
Midwest Of The United States,for instance?
Christian Ulbrich (09:27):
Well, first of all, what type of companies
are we talking about in the realestate industry? If you talk
about small investors or thosetype of developers, those type
of companies, they can easilypartner with us or with others
who have the data in house andand we are very happy to help
them to be successful as as ourclients. If you talk about
(09:49):
smaller competitors to us, Idon't necessarily there's a
question around whether you arebig or small because the
advantages of being smallbecause you can move faster,
it's probably easier to convinceyour your teammates how
important it is to embrace AI ina smaller, store than in a
(10:09):
larger outfit. But the thedisadvantage is access to data.
I think that is for us thebiggest advantage.
The amount of proprietary datawhich we are obviously
complementing with publiclyavailable data. Now a smaller
(10:31):
outfit doesn't have that muchproprietary data, but there's
still a lot of data out in thepublic. So they can still have
the opportunity to use AIeffectively.
Mark Bonner (10:41):
But as part of the so okay. But part of the answer
is they should they should workwith JLL?
Christian Ulbrich (10:47):
Well, others will probably do the same what
we do. I don't think what we dois totally unique. Maybe we we
went on that journey a bitearly, but this is nothing which
will where you have forever acompetitive advantage because
everybody will embrace AI.Otherwise, they will be out of
business relatively soon.
Mark Bonner (11:08):
So the other big thing that's out there is
something that everyone'stalking about. Agentic AI, which
basically means systems thatdon't just answer questions, but
actually take actionindependently. You're building
that into Falcon. A lot of otherof your competitors are as well.
What does the general commercialreal estate public make of that?
(11:29):
How close are we to AI runningfull workflows inside CRE or
even at GLO?
Christian Ulbrich (11:37):
Well, I would say the whole development is in
comparison to previousinnovations we have seen at
lightning speed level. And andpeople shouldn't make the
mistake that they think AI isjust for the simple task. I
would say it's the opposite. AIis starting to prove, agentic AI
is starting to prove that themost complex tasks are actually
(12:00):
the one where you have thebiggest advantage because with
that operational complexity, AIcan really drive additional
value. And so it you can almostthink about the most complex
property types will will createthe biggest kind of value when
(12:22):
you apply AgenTic AI.
And and that's now in thedevelopment, but it goes really
outrageously fast.
Mark Bonner (12:31):
What I mean, within JLL, what property types are
showing the biggest operationalgains so far when it comes to
agentic AI?
Christian Ulbrich (12:38):
As I said, the more complex, the better. So
you see obviously a lot ofadvantage around when you deal
with multifamily with all thesetenant interactions and all the
necessary kind of work whichneeds to be done in those
properties around maintenance.Those are the things where I can
(13:00):
really play it out. But you canalso think about industrial
buildings where you have thesupply chain coordination.
Literally, the more complex, thebetter.
Mark Bonner (13:13):
So, you know, we're talking about accuracy a few
minutes ago, right, in in adifferent context where there's
deep human intervention, as asthe guardrail or the fact
checker to make sure that it'saccurate, especially as you're
making big complex decisions orat least presenting things to
clients. On this, it's a littlebit more independent. What kind
of guardrails are in placeinternally to make sure that it
(13:34):
is accurate? Is it data quality?Is it compliance?
Are there ethical concerns thatyou all have discussed
internally on this front as asas you begin to really roll this
out?
Christian Ulbrich (13:46):
Yeah. I mean, obviously, everybody is is still
learning as we move forward, butthere are a couple of things
which we are particularlyconcerned about, most notably
the whole topic around cybersecurity. That's why we created
JLLGPT a closed system that ourdata doesn't leave the company,
(14:07):
that is very important to us andthat client data doesn't leave
the company, but that is a hugetopic around cyber. The other
challenge which is now comingup, you will have heard about
the term citizen development.This is moving so fast and you
cannot stop your talentedworkforce to kind of create
(14:32):
their own prompts which thenbecome a small little agent.
And so there's a massive,massive move towards citizen
development. Again, this iswithin JLL happening almost
exclusively within the frameworkof JLL GPT. But, you know,
(14:53):
people may bring their owndevice and then use a different
system. And that's where we haveto prevent that any kind of
corporate data is getting onother devices. So there are a
lot of guardrails in place, butit is something where we will
learn over the coming months andyears what kind of loopholes are
(15:17):
being created which become achallenge for that whole trend
around AI and the riskaccompanying it.
Mark Bonner (15:28):
If you're just joining us, this is First Draft
Live. We're talking with JLL CEOChristian Albrecht calling us
from Frankfurt, Germany abouthow the firm's Falcon platform
and AI strategy are reshapingglobal commercial real estate.
Christian, let's go to aquestion from the audience here.
Is the Falcon technology offeredto outside customers, or are
there plans to do so?
Christian Ulbrich (15:50):
Well, we have obviously offered it to our
clients very much in our workdynamics business. So you
mentioned it in the intro. Wehave clients now who used to ask
questions about their ownportfolio to a large group of
(16:10):
business analysts and then theywere trying to bring all the
information together and reportback a couple of days later. And
now they can just write theprompt into the tool which we
have provided to them and so theCIE lead will have the answer
instantly, which is fantasticfor them. It creates a bit of a
(16:32):
challenge for our industrybecause we have to now create
new ways of being rewarded forwhat we do.
We have to be paid for theoutcome. And so that is
something which is a change inthe industry and we are still at
the beginning to see thatdriving change into our
(16:52):
industry. But that is just oneexample. Are plenty of other
examples where we provide accessto clients to their portfolio
and they can look what otheropportunities similar to their
portfolio is being out there.And so again, this is just early
days, but it's coming everycouple of weeks, new products to
(17:14):
the market.
Mark Bonner (17:15):
You know, across capital markets AI is tightening
the deal clock. It's spend it'sspeeding up underwriting. It's
cutting through diligence. It'seven transitioning deal docs in
seconds. From your perspective,Christian, I mean, what capital
sources are leaning hardest intoAI assisted underwriting right
now?
Christian Ulbrich (17:33):
It's probably the investment management firms
and maybe even more privateequity. We saw some private
equity companies really doublingdown on their data accuracy over
the last couple of years andthey are probably now the ones
who are most embracing theopportunity AI is delivering. I
mean, we shouldn't forget at theend of the day, the real estate
(17:54):
business is a data business. DoI buy this building or do I buy
that building? Do I lease thisbuilding or do I lease that
building?
Those are not gut fields. Thoseare things which are supported
by facts, by data. And and sothere is no surprise that some
of the most successful realestate investors are the ones
(18:16):
who are embracing data the most,and they are the ones who are
now embracing AI the most.
Mark Bonner (18:21):
You have fierce competition everywhere around
the globe. And all of your yourfiercest competitors, which are
household names withincommercial real estate, are also
spending bootyles of cash onvery similar products. They can
speed up their their teams tocome to results faster and
obviously things that are moreclient facing. I guess the
question for you, Christian, isas you peer over the horizon
(18:42):
here, when everyone's fast andwhen everyone has this sort of
AI playbook, if you will, whatbecomes the new edge in this
world?
Christian Ulbrich (18:55):
Well, I I don't think that's a big change
to what it was before. Anyinnovation is being caught up by
others very quickly because whatwe can do, others will look at
it and they will try to do thesame. And if they don't do it,
then those type of companieswill be out of business. You
know, go back to whatever manyother examples, the the mobile
(19:17):
phone, you know, when we movefrom normal mobile phones to
smartphones, you know, the oneswho are still not providing
smartphones are just out ofbusiness, and it's the similar
trend is here. So I don't thinkthat any kind of AI tool which
is driving just productivitywill be in any form a game
(19:37):
changer.
It will become table stakeswithin three, six months. Where
there is potentially anopportunity is when you find a
way for a completely new go tomarket strategy. So there may be
agents which will completelyreinvent parts of our services
with a completely new approachto go to market. That then could
(19:59):
become a real differentiator.But there isn't anything out
there.
I guess everybody is working onthose. We are working very hard
on those kind of potentialagents, but nothing is at the
moment out there, which I wouldsee as a big game changer.
Mark Bonner (20:14):
You know, the headlines that we've all seen
here on another front has to dowith talent. Right? AI has been
rewriting job descriptions. It'sbeen cutting entry level roles.
It's been boosting output andchanging what junior, quote,
unquote, even means incommercial real estate these
days.
How do you balance thatautomation with while also
developing human talent?
Christian Ulbrich (20:35):
Yeah. I mean, there's a lot of talk out there
that entry level jobs willcompletely disappear and and and
and overall the amazing amountof potential job cuts. Let's
see. At the moment, it createsjobs. That doesn't mean that it
is always the same person, butat the moment it creates job.
That's the reason why SanFrancisco is coming back as a
(20:57):
real estate market really nicelybecause of all the AI jobs. So
going back to that questionaround analyst, I think what it
does is it provides smart youngpeople an ability to leapfrog in
their career experience in theirin their traction on gaining
knowledge and becoming a reallyvaluable member of a team much
(21:22):
earlier. What took three yearswill probably now only take six
to twelve months in thatjourney. And so overall, I think
for people with ambitions andthe willingness to learn, the
opportunities are only rising.The flip side of that is those
who are not that interested toreally embrace those new tools
(21:42):
and and go the extra mile, theymay find it more difficult to
have a career or even find arole going forward.
Mark Bonner (21:50):
You know, Gen Z, as we've discussed a lot on this
show over the last year, youknow, this is the most educated,
most tech enabled generation toever hit the global economy.
They're having a lot of hardthey're having a they're having
a hard time finding work rightnow even though they've went to
some of the top four yearschools, especially United
States and also in Europe andAsia. When you think about those
(22:13):
people, those are your futureleaders, for JLO and obviously
future leaders for commercialreal estate. What skills are now
must haves at JLO at that level?What is the fastest way to get
your resume to the top of theand for you to get a job at JLO
given this dynamic that we'rediscussing with artificial
intelligence?
Christian Ulbrich (22:34):
Well, that's a big question. I would say the
skill sets needed, is that youcan really communicate well,
that you can be part of a team.I think and that has not
necessarily anything to do withAI, but the world has become
incredibly complex. So it's nota world for solo fighters. It's
(22:58):
a world for team players.
And so people who can really bepart of a team and do their very
best for the overall success ofthe team are in a in a really
good position. And then withinthe team, you will have people
who are slightly more expert ona specific topic. Being
technology savvy, being AI savvyfor young people will be
(23:21):
completely table stakes, And andtherefore that will not be the
differentiator. But you can alsoturn that whole thing around. I
heard just yesterday a storythat some of the best
universities, their students arecreating their own agent while
they are going throughuniversity.
And then when they come to aninterview with a company, they
(23:42):
ask the company, what tech stackdo you have? Can I bring my own
agents Because that is my lifenow, my agent? It's my partner.
Because that's what people do.They use now agent as a partner.
So you're you're hiring you'rehiring Peter, but Peter brings
his agent. And so you get from acompany perspective almost two
(24:02):
for one.
Mark Bonner (24:05):
So, look, the bottom line is this. You know,
we're we're still in the veryinfancy of AI adoption in the
global economy, and certainly,that's true for commercial real
estate. We're in the pilotrefine scale phase right now.
The messy middle, I think whereevery firm's experimenting, some
are winning, some are not.Nobody's got the playbook down
(24:25):
just right.
And I think that includes JLO,right? You may be further along
on this journey than a lot ofother companies, but this is
still obviously being figuredout. And there's obviously a
Christian Ulbrich (24:35):
lot of
Mark Bonner (24:35):
opinions about AI. Right? In your opinion, what do
you think is the most commonoverreaction you're seeing
around AI adoption?
Christian Ulbrich (24:45):
Well, I think a lot of people kind of who are
maybe skeptics to AI are usingit for the first time. They are
bringing their data in, and theyare writing prompts, and they
are completely disappointed withthe outcome and then they blame
the AI for the outcome, but theyshould blame their own data
accuracy. You know this wholestory and it's voted many, many
(25:10):
times garbage in garbage out.That is coming to a new level
when you talk about AI. Andtherefore that is probably a
very common reaction from peoplewho are skeptics.
You have to do your homework ifyou want to take the benefit
from AI. And so that is, by theway, one of the big challenges
which every organization has,and I would say larger
(25:33):
organizations even more thansmaller organizations, that you
have to overcome your internalskeptics who wanna prove to you
that this is all rubbish andthey don't need to do that. And
and so that's why we have oursuccesses, but we also have our
failures. But very often, thefailures are kind of man made
failures within yourorganization.
Mark Bonner (25:54):
For the rank and file commercial real estate
leader, right, what is your bestadvice to them right now about
where they should start ifthey're still on the sidelines,
regarding experimenting with AIand trying to implement it into
workflows in their offices?
Christian Ulbrich (26:14):
Well, you know, I think it is important
that you are identifying two,three trials, and then you
really double down on them andtry to make them work. Don't
spread kind of yourself toothinly with twenty, thirty
different ideas, but nothing isreally getting the right
(26:34):
attention. Try out with two,three thoughtful ideas and then
go deep on them. But don't waittoo long. The train has left the
station, and this train is goingat kind of, I can say, Japanese
speed levels of trains really,really fast.
Mark Bonner (26:55):
Okay. That's our show for today, Christian. I'm
getting the hook here from myproducers, but I wanna thank you
very much for taking the time tobe with us today. I hope you
have a wonderful weekend.
Christian Ulbrich (27:03):
Thank you so much for having me, Mark. Have a
nice weekend too.
Mark Bonner (27:07):
If you missed any part of this conversation or
wanna catch earlier episodes,you'll find every show on
biznow.com or in your favoritepodcast feed. Just search First
Draft Live. Our conversationwith WeWork CEO, John Santora,
live from the Cretek Conferencein New York this week is now up
on the site and also in yourpodcast stream. And next week,
we're back with Industrious CEOand cofounder Jamie Haddari for
(27:30):
a look at where the flex officemarket goes next. This is First
Draft Live.
Have a great weekend y'all.
Popular Podcasts
Stuff You Should Know
If you've ever wanted to know about champagne, satanism, the Stonewall Uprising, chaos theory, LSD, El Nino, true crime and Rosa Parks, then look no further. Josh and Chuck have you covered.
Dateline NBC
Current and classic episodes, featuring compelling true-crime mysteries, powerful documentaries and in-depth investigations. Follow now to get the latest episodes of Dateline NBC completely free, or subscribe to Dateline Premium for ad-free listening and exclusive bonus content: DatelinePremium.com
The Bobby Bones Show
Listen to 'The Bobby Bones Show' by downloading the daily full replay.