April 29, 2024 • 43 mins
Leverage the full potential of your Multifamily Real Estate investments with insights from Elizabeth Braman, CEO of RevolutionRE.
As our special guest, Elizabeth unravels the complexities of AI and data standardization in apartment investing, offering strategies for multifamily businesses looking to enhance their efficiency. We delve into the ways companies can align their data strategies with their overarching goals, transforming the way we think about property investment and management.
Navigating the often-treacherous terrain of data integration, our conversation with Elizabeth Braman focuses on creating a seamless and scalable system that can stand the test of time.
She highlights the importance of crafting KPIs that truly reflect a company’s performance and the critical nature of involving the entire team in the data journey.
We also tackle the real-world challenges that crop up when trying to wrangle various data sources into a single, coherent entity, particularly in the Multifamily Real Estate industry where standardization can feel like a distant dream.
Rounding out our deep-dive into the data-centric world of Multifamily Real Estate, Elizabeth sheds light on the game-changing impact of her innovative approaches to data standardization and the role AI plays in revolutionizing the multifamily industry.
We explore how historical data reveals more than just past performances.
Connect with Elizabeth Braman
Thank you for tuning in to today’s episode. If you found value in our conversation, please subscribe to our podcast on your favorite platform and leave us a rating and review.
Your feedback not only helps us improve but also helps others find us. And if today’s episode sparked a thought or provided a new insight, consider sharing it with a friend who might also benefit.
Together, we can grow our community and continue to learn and innovate. Thanks for listening, and until next time.
About the Multifamily Innovation® Council:
The Multifamily Innovation® Council is the executive level membership organization that makes a difference in your bottom line, drives a better experience for your employees, and allows you an experience that keeps demand strong for your company. The council is uniquely positioned to focus on the intersection of Leadership, Technology, AI, and Innovation.
The Multifamily Innovation® Council is for Multifamily Business leaders who want to unlock value inside their organization so they can create better experiences and drive profitability inside their company.
To learn more or to join, visit https://multifamilyinnovation.com.
For more information and to engage with leaders shaping the future of multifamily innovation, visit https://multifamilyinnovation.com/.
Connect:
Multifamily Innovation® Council: https://multifamilyinnovation.com/
Patrick Antrim: https://www.linkedin.com/in/patrickantrim/
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Speaker 1 (00:00):
Welcome back.
Today's guest is ElizabethBrayman.
She is the founder and CEO ofRevolution RE.
That's a SaaS company providingdata solutions to apartment
owners.
They do this for managers andreally any industry-related
company as more and more data iscoming together.
Prior to forming Revolution RE,elizabeth served as the chief
(00:23):
production officer at RealtyMogul.
Forming Revolution RE, elizabethserved as the chief production
officer at Realty Mogul.
This was an online platformraising capital for real estate
companies through privateplacements and A-plus REIT
offerings.
Prior to that, she was a chiefproduction officer at ReadyCap
Commercial, a small balancesheet commercial real estate
loan originator funding loansfrom private REIT offerings that
(00:44):
has since gone public.
Now Elizabeth is a certifiedcommercial investment member
this is the CCIM designation anda licensed attorney in
Washington DC, in theCommonwealth of Virginia.
She holds a Bachelor of Artsdegree from the American
University and a Master ofBusiness Administration and JD
from the George WashingtonUniversity.
(01:05):
Today, what I'm doing is I'mbringing in Elizabeth to share
her knowledge on how AI and datastandardization are key to
unlocking new efficiencies andopportunities in apartment
investing.
Elizabeth, welcome to theMultifamily AI Podcast.
Speaker 2 (01:20):
Awesome.
Thank you, patrick, it's greatto be here.
Speaker 1 (01:23):
Yeah, so much background and expertise in what
you do from a legal aspect.
I love that when we talk data,you're not here to give public
and legal advice, and nor am I,but it's interesting.
That type of background,serving at the leadership role
of a company solving data, iscompelling, obviously One of the
questions I'm thinking aboutfor leaders, business leaders.
(01:46):
So on this podcast I'm tryingto help owners and operators
that may be from sort ofnon-technical roles understand
the value that can be unlockedleveraging data in their
business, like they leveragedebt as a business tool, and
would love to know what youthink those questions should be
as people think about their datastrategy.
Speaker 2 (02:09):
Great question and I think it really obviously
depends on the organization.
I think the first thing Irecommend is really taking a
look in the mirror and sayingall right, what is a data
strategy?
What does that mean to ourorganization?
Because it can mean a lot ofdifferent things.
And what really drives the datastrategy should be your
(02:34):
business goals and objectives.
What are you looking toaccomplish with data?
What do you think you canaccomplish with data?
And then look at things likehow does it impact governance
and data security and all ofthose things.
But really starting with yourgoals, your objectives defining.
(02:56):
I'm a big fan of smart framework.
Is it specific, measurable,achievable, relevant and
time-bound?
And I go through this exercisewith our clients as well,
because when they're workingwith us on developing their data
strategy, it's really importantthat it's defined, that they
know who their key stakeholdersare, that they get buy-in from
(03:22):
their key stakeholders and thatthey're measuring what success
means to them.
Because data strategy is one ofthose things where, if it's too
amorphous, it's too vague,there's really no success metric
that an organization has.
That's when a lot of time,effort and energy is put into
(03:43):
things that eventually getabandoned because there's no end
point to it and people can'tlook and see why are we doing
this, which is important becauseyou need your team to really
participate in the process ofimplementing a strategy.
It can't be done by a vendor orin a silo.
(04:03):
One person can't take it on allby themselves if it's an
organizational data strategy.
Speaker 1 (04:10):
And getting into how to actually define this.
Many business leaders youmentioned we think about
dashboards, we think about thedata that we've always had to
make P&L decisions, knowingwhat's going to move the
investment needle in thebusiness.
So a lot of the businessleaders are more aware of their
data than they probably want toadmit, because it's really it's
(04:34):
decision making right.
But when we introduce AI anddifferent, we're talking about
standardization challenges andthings like that.
When it comes to solving thoseproblems, oftentimes a leader
will pass it off Okay, that's ITor that's somebody else's job
(04:55):
internally.
I'm curious, with theacceleration of technology and
talent, are these thingscompanies should be doing inside
their organization or relyingon outside parties to guide them
through right?
And then what I'm curious aboutis what's the leader's role in
(05:17):
understanding that data strategy?
Is it something they just handoff to IT or their technology
team?
Maybe they don't have one?
Speaker 2 (05:27):
Yeah, I was going to say goodness, no, I would hope
not.
You don't want to give the poorIT a data strategy rollout
without a lot of input from thebusiness stakeholders.
Those two things can't be donein a vacuum.
I think if an organization doeshave an IT group and a lot of
companies don't big real estatecompanies that have I shouldn't
(05:50):
say big the larger real estatecompanies obviously have
internal teams, but there arecompanies that have billions in
AUM that don't have a largeenough team to take this on
independently.
So it depends on the size ofthe organization.
But it's either done assomething in tandem with the IT
(06:12):
team or with external vendors.
But picking the right externalvendors can be great or it can
be not great.
Getting good references, butalso who you connect with, how
do you meld in terms of yourwork?
Because it's going to be aprocess and there is no quick
(06:34):
fix.
You can't bring a vendor in andjust wave a magic wand and say,
okay, data, that's not a thing.
Standardization obviously canhelp the process, but there's
still a requirement of everyreal estate company to
participate in the process ofand I decide whether it's a
(06:55):
long-term relationship that youwant to build from there and
look at this kind of as along-term strategy.
You can't spin up a completeintegration with every single
(07:18):
data source and data plan andhave it done in a week.
That's having reasonableexpectations is critical.
Getting your vendor or your ITteam to buy into those goals,
expectations, timelines, all ofthat is really critical to
having a successful datastrategy rollout.
(07:40):
But I think it's really up tothe company to work with
external or internal.
There's obviously benefits tohaving someone with completely
fresh eyes look at yourorganization and help guide that
process, and it's alsoimportant for internal folks who
really know all the nittygritty of how data has been
(08:02):
captured and stored.
So a little bit of both, I'dsay, or a lot of both, depending
on how big your data strategyis and your goals are.
Speaker 1 (08:12):
Yeah, and I love the idea of the expectations and a
lot of times when we'reevaluating even technology, I
guess you can group data into aalmost like a product right
Productization of round, I thinkyou've called it.
Is it's standardization as aservice right?
Productization of round, Ithink you've called it.
What is it?
Standardization as a serviceright?
But, more importantly, in termsof those expectations,
sometimes leaders are lookingokay.
(08:32):
So when I have a problem,what's the problem?
This technology or thisstrategy is solving and in many
cases there may, depending onthe leader's role and how they
are up to speed on where thingsare in the business.
We may not know what we don'tknow.
So how do you come into anorganization and manage those
expectations and demonstrate thevalue of getting this
(08:55):
standardization correct?
Because it seems like a lot ofpeople want to do it, but
there's some people run intosome challenges doing it.
Speaker 2 (09:02):
Yeah, for sure.
I think I always laugh whensomeone asks me is it done?
It's never done.
It's a process, having a reallywell-structured process, and
that's why I use the SMARTframework and kind of work with
teams when we're engaged in theonboarding process to really
(09:24):
define everything they want toaccomplish and then trying to
prioritize and work throughtimelines for having small wins
so that you're iterating andbuilding it out over time.
You have to see some measurableimpact when you're looking at
data and so, depending on howcomplicated your tech stack is,
(09:50):
how many data sources and datasets you're looking to bring in,
how different the datastructure is to start, and it's
why people look for scalablesolutions to this stuff so that
they can find a way to make ameaningful impact right out of
(10:10):
the gate.
Just even just getting all yourdata in one place is an
accomplishment.
Right formats, having dailyaccess to property data from
multiple systems that's a bigwin to be able to get that done.
Then you look at okay, how dowe refine our KPIs?
(10:36):
What kind of metrics do we wantto start tracking?
What different teams within theorganization want to be
involved?
What different teams within theorganization want to be
involved?
Is it all marketing or is therethe leasing team, the
management boots on the ground?
Who is this really going toimpact and how do we get them
from step one to step twothroughout this journey of data,
(11:00):
but there is no done.
Of data but there is no done.
It's always just a continuous,never-ending product process,
which is why you have to stay ontop of it and create some
business roles internally.
But it's a commitment.
There's definitely a return,which is also why you want to
have measurable.
See how you're getting a goodreturn on rolling out a strategy
(11:23):
.
Yeah, I like the small wins.
Speaker 1 (11:26):
I like what you said about the small wins and making
it measurable when data iscoming from so many different
places.
What are the challenges ofgetting this?
If it's an accomplishment toget your data in one place right
?
Speaker 2 (11:37):
Yeah.
Speaker 1 (11:37):
Are there examples that even other industries are
doing that we could learn from?
Speaker 2 (11:43):
Yeah, it's.
The big difference between realestate and other industries is
it's not the most open framework.
I think.
If you look at banking, you canadd your credentials into a lot
of systems and all of a suddenyou've got a direct connection
with your bank.
Or marketing, if you want toget all of your Google Analytics
in one place, open APIs.
That's not quite the case inmultifamily, and so we're
(12:09):
extracting data from coresystems.
Sometimes we have one, two,three different methods of doing
it.
It's challenging because you'vegot many stakeholders involved,
whether it's the managementcompany that's generating and
storing the data, ownershipgroups who own some of the data
to different property managementsystems that all have different
(12:31):
data structures.
So extracting the data is thestep.
One is getting it into a commonand consistent format.
That's the transform.
So ETL is the extract.
The transform and then loadingit into a place where you can
now use it for BI is just afront-end visualization of data
(12:56):
that's in a structure that youcan use.
But if it's coming frommultiple systems and it's in
multiple structures and it's inmultiple structures, having a BI
front end and trying to createmultiple reports from those
source systems is a lot of workand it's why a lot of
organizations are challengedwith their reporting if they are
(13:17):
trying to pull it all togetherand generate portfolio-wide
analysis.
Speaker 1 (13:24):
And obviously having a standardization of data plays
a huge role in that process,right?
Are people skipping steps inthis?
Do you think right now?
Speaker 2 (13:34):
Yeah, I think if you're doing a bespoke approach
and it's reasonable, it makessense.
If you have one system of recordand you set it up to feed data
into a front-end visualizationand then you add a second one,
now you're doing the sameprocess twice side by side.
(13:54):
But it doesn't really meshunless you take a step back and
get all that data in one format.
Because these systems are notconsistent, there isn't an
industry-wide this is ourstandard for how data is
structured and because thereisn't one way of doing it, you
(14:16):
are limited to either having tostandardize it internally and
create a data model, and surenot a lot of real estate
companies want to create a datamodel internally.
It's expensive, time-consuming,requires specialized knowledge
and it's just messy.
So that's why we built what webuilt specifically for the
(14:38):
industry, because it didn't seemlike there was a solution that
was out there that would providethat same aggregation of data
for people to be able to thenuse it for AI applications, for
BI applications, for all thecool technological use cases,
(14:59):
because the data just wasn't ina format that made it easy to
use the data just wasn't in aformat that made it easy to use.
Speaker 1 (15:11):
And when it's in this format, when you mentioned the
ETL extract, transform and load,are there things that or
questions we should be askingaround?
Making sure that the quality ofthat process gets the like, how
do you ensure that process?
How do you make sure that ETLprocess gets the right data in
the right places?
Speaker 2 (15:27):
Yeah, that's a good question.
What do they say?
Garbage in, garbage out, so wecan clean up some of the data
that comes into our system.
You can't fill in empty holes.
If there's big blanks in data,there's no way to create it.
Ai is also helping to clean andcreate methods of cleaning data
(15:53):
, and a lot of these modelsdon't need perfect data.
They need a lot ofdirectionally correct data to
generate really interestingresults and to be impactful.
Speaker 1 (16:06):
This could just also be like just as simple as a
state record right, A fieldwhere it's maybe open text in
one database and the other it'sabbreviation.
Speaker 2 (16:18):
Absolutely.
Our resident name is last name,comma, first first name and
somewhere else it's a first namefield and a last name field.
It's all across the board,getting the data in just a
format that you can query, andeveryone who's used Excel with a
sort function knows that likeit needs to be a little bit
(16:39):
clean to be able to useeffectively.
Some of the AI solutions arenow starting to fill in the
blanks of helping to cleanthings up.
So we in our tool if it'sGoogle Analytics, google with
three O's, we can combine allthose and essentially for ad
sources, we're doing somecleanup on data like that, where
(17:02):
we're just making it moreconsistent when at the property
level it's being input slightlydifferently.
Our financial model, which takesmillions of rows of data,
different categories everycompany has what?
Four, five, 600 rows offinancial data and mapping it to
a more tight, cleaner set ofdata points so that you can do
(17:28):
benchmarking.
We call it our common model.
There's huge value in havingeven a little bit of cleanup of
the data so that you can do morecomparative analysis and market
analysis.
Speaker 1 (17:44):
What can go wrong in that process, or what should we
avoid doing?
Maybe someone's trying to dothis in-house, or oh, it's a
good question.
Speaker 2 (17:53):
There's tons of things to avoid, but I'd say you
want to.
There's a lot of monitoring andtracking that your data feeds
are actually active and theyaren't stopping for whatever
reason.
You might have a night where asystem just can't handle.
It times out.
It's too much data coming in.
Speaker 1 (18:15):
In those types of timeouts.
Does that stuff get queued oris it a restart?
Speaker 2 (18:21):
Yeah, that's a great question.
It depends.
So sometimes it just like stopsand you have to go and you get
notification that you have tostart over.
Then there's sometimes you canjust pick up where you left off.
And if it's a timeout, there'salso time zone issues, which are
really strange.
So if you're doing a nightly andyou have some data that's
(18:43):
coming in, but for the change intime zones things will happen
before 12 o'clock and thensometimes after, which makes it
a different day.
So you have to make sure thatyou time all of your data feeds
so that they're not impacted bydaylight savings or, if you're
(19:03):
in a different country,different things like that.
So there's definitely ways thatthe extract process can fail.
Then there's ways that the datacan be off.
A lot of people like to do.
Timing issues can happen.
So managers will come in onMonday after a long weekend and
(19:26):
they input their move outs fromover the weekend and then you've
got different timing issuesthat directionally, can make
data not match from one systemto the next.
So that goes back toexpectations.
Speaker 1 (19:44):
Yeah, exactly, and I'm also like from all of this
work, it's the businessdecisions, almost like the
construction process is very itcan be messy, yeah, and it is.
Speaker 2 (19:56):
You're building a house a lot of dust, a lot of
unknown things, which is whyhaving a good blueprint ahead of
time keeps you from havingstructural problems.
If you set it up wrong or builda house wrong and then have to
go back and try and figure outlike oops, we didn't put the
plumbing in, or the electricityis in the wrong place, Like
(20:19):
those are the things that youwish you had done.
They say derf to do it rightthe first time.
Sometimes time makes it hard,because if you've been working
on this for a very long time,things change and you have to go
back and revisit some of thosestructural things.
But that's technology.
It's always an iterativeprocess.
Speaker 1 (20:41):
When you're in these conversations with executives
today and you're working throughthis, what are they leaning
into and what are they pushing?
Not only say pushing back, butmore cautious on?
Or if we don't understandsomething, it's hard to make a
decision about it right, and sothere's going to be a lot more
education, I think, around thisand unlocking the value.
Because you mentioned thedifferent stakeholders,
(21:03):
collaboration between teams andI always think about the
dashboard is like a lot of thetechnology that we're using is
making the assumption that we'regoing to be in an office
looking at a computer, using it.
So if you have this centralizeddata standardized and it's gone
(21:25):
through the ETL process, thenthat can be unlocked in mobile
like all over right.
You can take that wherever itneeds to be.
Speaker 2 (21:35):
Yeah.
Speaker 1 (21:36):
Yeah.
Speaker 2 (21:37):
That's the load portion.
You can load it into a numberof different applications or
uses.
But I think a lot of people arerightfully so leaning in on
corporate data governance,wanting more visibility into
tracking their data feeds andmaking sure that they're
reporting on the rightinformation, especially if
(21:59):
they're a public company or haveREITs with reporting
requirements.
All of those things arecritical to have a handle on how
the data is coming in.
Speaker 1 (22:11):
Tell me more about that data governance.
Speaker 2 (22:12):
Explain that, what you mean by that in, so explain
that what you mean by that.
A lot of companies, at leastlarge corporations, are looking
at how they secure and storetheir data, how they're
maintaining a handle on theaccuracy and also the security
issues related to data.
(22:34):
The multifamily industry we'redealing with residents and
residents' housing, and housingis obviously a highly regulated
industry.
The payments and activities ofresidents, their data is PII, so
the storage of their data canbe highly regulated.
(22:56):
There's a lot of different.
So when we look at corporategovernance, it's really how does
a company maintain the highestintegrity and policies and
documentation and legalstructure around how they are
storing, using, maintaining data?
(23:17):
So it's some effort to be doneinternally for big companies to
make sure that they're putting aplan in place and documenting
it appropriately.
Speaker 1 (23:28):
And do you offer?
Is your program a data store oris it wherever they want?
Is it going to other places, orhow does that work?
Speaker 2 (23:37):
It can be yeah, so some companies look to have
their data fed into a datawarehouse.
Some companies want us tomaintain and store their data
for them.
We have SOC 2 compliance, soall of our data is encrypted in
motion at rest.
(23:58):
Very few people have access tothe data and all of our policies
and procedures are available inour SOC 2 reports, which you
can get on our website.
But you have to sign an NDAbecause our auditors require
that.
But the whole process is not aninsignificant amount of effort,
but it's really important andsomething that I always
(24:19):
encourage people to talk totheir vendors about, because it
is incumbent upon the realestate companies to make sure
that they're the ones that couldbe liable for any data breaches
if they're not maintaining goodpractices, if they're being not
(24:39):
careful with how they store andmaintain their data.
Speaker 1 (24:44):
And leaning back into AI, how is AI impacted?
It seems like a lot of thesetransformations and
standardizations are gettingeasier.
There still needs to be thegovernance side of things.
It's one thing to turnsomething over to AI, but I'm
under the belief that peoplebegin and end the process, so
maybe it's taking out some levelof review.
(25:07):
But how is AI impacting thestandardization process for your
perspective?
Speaker 2 (25:12):
Well, it's a great question and there's obviously
lots of different applicationsfor AI.
The AI is thriving.
The need for standardizationbecause AI needs like data in a
structure that it doesn't havein its knowledge base.
(25:35):
It kind of lies and gives youanswers that you're a little
confused with because it's likewhat is it talking about?
It makes stuff up and that'sbecause it doesn't have the
right data to answer thequestions.
It tries to fill in the blanks.
It's being helpful, but notreally the blanks.
(25:58):
It's being helpful, but notreally.
So there's certain things thatAI is driving in terms of the
need for standardization, andthen there's, equally, ai is
providing standardization withmore ways to make data clean,
usable, accessible, so they feedeach other in certain ways, and
how much faster and better thetechnology is at letting us make
(26:24):
informed models using availabledata.
So it's requiring less and lessas long as the models are
trained and maintained.
But yeah, I know it'sinteresting.
They feed each other in someways back and forth.
But there's so many AIapplications and real estate has
(26:46):
been pretty lagging in howthese applications are being
utilized.
I think there was a good studyMcKinsey put out and if anyone's
interested, feel free to emailme.
I'll share it, but it wastalking about how the data is
really the shape of data.
The lack of a foundational datamodel can really hinder
(27:08):
people's ability to use data ina lot of these AI applications.
Speaker 1 (27:14):
And these large language models are moving so
fast and it's so expensive to dothat.
It's interesting, like when youcompare it with your private
company data in an environmentwhere it's not training the
large language model.
But it goes back to what we'retalking about here, which is, if
you don't have thestandardization of the data or
in the data in the right place,you can't retrieve it and then
(27:37):
get better responses when you'reusing AI applications inside
workflows.
And so I always wonder, asyou're moving companies' data
into a single source of truthand also standardizing it, what
are you seeing the challengesthere?
Is it they have differentsystems?
(27:58):
Maybe they're a third partyoperator and they don't just
have one system?
They're trying to managemultiple systems and I imagine
training people and all of thatcomes into play.
But in terms of the technical,what are those challenges in
bringing all these thingstogether?
Speaker 2 (28:15):
Yeah, that's exactly right.
It's ownership groups who workwith multiple property
management companies.
It's property managementcompanies that have multiple
ownership groups that requirethem to use different core
systems and or are buyingproperty management companies.
So sometimes it's just acompany buys another company,
(28:37):
and it's just the nature of realestate.
It's transactional in nature.
So when you have a sale of oneproperty to another buyer, are
they going to use the samesystem, are they going to use a
new system, and so you've got alack of kind of historic data
(28:59):
and then you're onboarding itinto a new system.
So anytime you have peoplemanually adding data into
systems, that also add some roomfor error, we might say and
there's just, they'reinconsistent.
They're not.
These systems weren't made tobe exactly the same and have the
same definitions and have thesame data structure.
(29:22):
They just they weren't set upthat way.
And there's not that manysystems out there.
It's not.
The percentage of the realestate market that's
professionally managed is thatit's consolidated, but those
companies aren't using a ton ofdifferent core systems and those
(29:42):
core systems really arecapturing all the financial data
, all the operational data, andthen you have all this ecosystem
of prop tech companies thatkind of have opened in the last
however many years, really takenoff, and so there's lots of
stuff, but that's just addingmore data to the problem.
Where is it being stored?
(30:03):
Where the core systems aren'tstoring that data, and these new
PropTech solutions aren'tstoring all the underlying
property data.
So you're just creatingadditional data silos.
You're exacerbating thefoundational issues that exist
with new technology.
So it makes your head explode.
Speaker 1 (30:26):
Yeah, it can.
When you think about thesecompanies, companies', entities
stay around longer thanproperties in the investment
cycle, are shorter term, right,and so we're asking people to
think longer term about thingswhen they're in a short term
investment cycle, right.
So that's always been achallenge.
But if you think about theentities themselves and also the
(30:48):
data, you may have company dataand then you have property data
and in the way that all of thatis organized, I'm just curious
if and it's probably true, thiswas my belief in that when a
developer builds an apartmentbuilding, the utilities and the
city infrastructure is valuethat a new buyer doesn't have to
(31:11):
come in and do and do.
In other words, because somebodytook the risk and developed the
plumbing, the electrical, theutility, all the things that are
required under the ability togo vertical, the new buyer
obviously gets that right, butnobody really values it on the
sale of an asset, right?
Yes, but when we talk about, ifyou just think about the
(31:35):
listeners, just think about allthe different entities you have
you may have your company entityand then a holding entity and
then each property has an entity.
Is that a way to think throughdata in terms of if we get
property data at a specificlocation that later, if there's
a disposition or a consolidationor merger, that there's
enterprise value that can berealized on the exit by making
(32:02):
investments in getting thisclean data for the new buyer.
Speaker 2 (32:06):
Yeah, I totally think so, but it's really up to
sellers to view that as an assetor buyers to ask for the
historic data of the seller.
They may or may not feelcomfortable doing that.
Speaker 1 (32:22):
You're lucky to get a PDF today, I know right.
Speaker 2 (32:24):
Are they going to?
I guess, if the market getstight enough, you could put that
into purchase contracts, thatyou want two years of historic
data or access to historic datafiles.
I think it'd be really hard toget a seller to agree to it, but
it would be great for a buyer.
Speaker 1 (32:42):
Let me just ask though that's the business kind
to me again what would that meanif you had that on a new
acquisition?
Speaker 2 (32:51):
If you had it on a new acquisition, you could
probably track the trends ofyour performance a lot better.
So if you're trying to and Ithink this is one of the best
use cases of data is to showyour investors that your
operational strategy bringsalpha, that you have somehow
managed to create opportunity inthe asset by doing X, y and Z,
(33:18):
that's hard to prove when you'restarting with ground zero and
there's nothing to compare it to.
If you had the historic data,you could show what's been doing
this for the past 12, 24, 36months and even adjusting for
market conditions.
And even adjusting for marketconditions.
This is how much value we'vebeen able to add.
(33:38):
This is how much that we'reputting into the property.
That's creating a bump in NOIand this is the multiple that we
, as a data-driven companypeople like to say we're a
data-driven company to prove it,to actually show that what
you've done has created thismajor impact.
(33:59):
I think that's super powerfuland it's hard to raise capital
right now.
That's a tremendous use case.
Speaker 1 (34:07):
We're talking about the apartment car facts
basically.
Speaker 2 (34:10):
Yeah, some people have been called it what's like
the for hotels, the report thatkind of shows the market value
for that.
Speaker 1 (34:18):
yeah, sure, but it'd be interesting.
You buy a property and this isthe data from the time that we
owned it, but imagine, as ittransferred ownership over three
decades, that it would be morevaluable in terms of
underwriting, repairs,maintenance and remaining life.
There's so many other.
Where are we in all?
Speaker 2 (34:39):
of this.
We can do that with financials.
We can do that with thefinancials.
So if you have a property thatyou're looking to acquire and
you have the trailing 12 monthsand you can get maybe the past
couple of years, we can showfrom a financial perspective how
you can compare from afinancial perspective.
(35:00):
How you can compare because wecan convert it into a common,
consistent format with yourcurrent performance, your
current tracking.
So being able to take your proforma track to that track, to
budget, but tracking to historicperformance.
It's a really interesting usecase and, for sure, something
that I encourage people to do,if not with our platform, but
with their own internally seehow a property was performing
(35:25):
and how your efforts havechanged the story at that asset.
That asset is really compellingwhen you're going out to market
and trying to get people toinvest in you.
They say past performance isnot an indicator of future
results.
But when you can showcomparative how this property
was doing compared to the marketand then how we're doing
(35:46):
compared to the market, it justtells a really interesting story
of performance.
Speaker 1 (35:54):
Yeah, that's interesting and I know that
there's so much more that we cancover.
On all of this, I go back towhere we started, which was
getting those small wins.
I've been calling it crawl,walk, run right.
So it's maybe somebodylistening, has access to data,
obviously, but maybe doesn'thave a defined strategy in a
market where that alpha beatingthe returns in other ways than
(36:18):
just passing on rent increasesif we can find ways to innovate,
the business having that datais huge in that decision-making
process.
Do you ever find people getoverwhelmed with this because
they are not data scientists andthere's not an abundance of
people to help with this andrelying on the tools you have
(36:40):
available to you?
And there's just gaps, that theydon't see, that they're missing
, and we can go down the wholeunstructured data conversation
and getting that into astructured Like that's a whole
nother, like phase two probably,of unlocking value.
But are you feeling or do yousee that people get overwhelmed
with this conversation and howdo they prioritize making the
(37:02):
next move and making sure it'sgoing to drive actual
effectiveness in the measuredresults?
Speaker 2 (37:08):
That's a really good question, I think.
I say don't try and boil theocean.
Pick a couple of discrete KPIsthat every company has.
The thing that they think isthis is the thing that we are
really good at.
Speaker 1 (37:23):
Where would?
Speaker 2 (37:23):
be a good place to start.
Look at reducing days on marketor days to if you are turning a
unit.
If you can turn a unit in twodays less than you have been
Huge impact on a property.
So what we do is we set up, youtrack your goal, you put in
(37:49):
alerts, you have a start dateand an end date and you look at
what is the overall impact andthat's something that's truly
measurable.
If you're dropping two days fromthe days that it takes to turn
a unit, that reduces your dayson market overall, adding two
more days of rent to each unitthat you're turning.
That is something that you canshow adds to your NOI and adds
(38:14):
to the value of your asset.
Same thing with looking atreducing an expense, your
marketing expense, if you'reable to drop it.
Or I look at customeracquisition, your lifetime value
, looking at that as a ratioacross your portfolio at each
property and setting a targetfor I'm going to reduce my cost
(38:37):
to acquire residents but I'mgoing to increase the lifetime
value of that resident.
Those are things that you canreally show to investors who are
being really thoughtful, notjust to transactionally turn
these tenants but actually toincrease the value of the
housing experience for them.
(38:58):
Renewal rates, increasing thosethat's a huge way to increase
your overall.
If you're looking at yourtrade-outs, these types of
things, sitting down and picking, just start with three, start
with five, but looking at, if Idid this and then was able to
(39:19):
accomplish it, what does that doto my property and portfolio
and how do I communicate that tomy key stakeholders, whether
it's a property managementcompany wanting to tell their
ownership groups we've added$213,000 to the property value
this quarter by doing X, y and Zand we help to automate the
(39:41):
process of communicating that,of showing that value.
That has tremendous ROI.
Speaker 1 (39:48):
when you look at things that way, yeah, and do
you find that people want to dosomething like that?
Look at those.
They have tools that aim todeliver some element of that,
but it's in that environment,right?
Especially if you're inmultiple systems.
That's where it becomes trickyto understand that stuff.
And do you find that when youget these small wins under their
(40:11):
ability to measure somethingthat truly makes a financial
difference in the business thatthey lean into and get more
motivated into investing moreand more into the data strategy?
Speaker 2 (40:22):
Oh, absolutely, it's almost.
You want to gamify it in a lotof ways Get your managers
involved or your analystsinvolved in creating these goals
and tracking to them, becauseit's not one property, it's the
entire portfolio.
So, each person doing theirpart and making these small
(40:49):
incremental changes, they canhave a huge overall impact to
the value of the portfolio justby being more thoughtful about
how you look at your data andhow you track to performance in
small ways.
Overall, that adds.
It adds up.
It all adds up.
Speaker 1 (41:08):
Yeah, it sure does I always think about.
When you mentioned the slipdays of time to turn, especially
in renting.
It's like empty airline seatsthat we just can never recapture
.
And that's that alpha, as youmentioned, that you can offer a
definitive or differentiatingvalue in running your business
through this certain managementcompany.
Listen, this has been great,Elizabeth.
(41:30):
I really appreciate you comingon.
I know that I would love tohave you back for more on this.
For those of you listening, ifyou want to know more about AI
and data for better apartmentinvesting, reach out to us and
let us know what are yourquestions that you have around
this, specifically to yourportfolio, to your use case, and
just go tomultifamilyaipodcastcom.
You can click there and you caneven send us a message, but
(41:52):
reach out to us, let us know andgo find that information.
Probably Elizabeth has it topof mind, but would love to visit
back with you, Elizabeth, as wecontinue down this journey of
extracting the knowledge ETLextracting the knowledge,
transforming it in a way thatnon-technical business leaders
understand it and then loadingit into their business.
That was a pun on the ETL.
Speaker 2 (42:13):
I love it.
No, that's so great.
Speaker 1 (42:15):
Yeah, look, you're doing great.
We love that you're making animpact in this industry and we'd
love to talk to you more aboutthis.
But until then, if you want tolearn more about Elizabeth and
some of the projects she'sworking on, go to Multifamily AI
Podcast.
In the show notes you can clicklinks and get to her LinkedIn
and all the other great placesto connect with her.
Until then, wishing you guysthe best and we'll see you on
(42:37):
the
Welcome back.
Today's guest is ElizabethBrayman.
She is the founder and CEO ofRevolution RE.
That's a SaaS company providingdata solutions to apartment
owners.
They do this for managers andreally any industry-related
company as more and more data iscoming together.
Prior to forming Revolution RE,elizabeth served as the chief
(00:23):
production officer at RealtyMogul.
Forming Revolution RE, elizabethserved as the chief production
officer at Realty Mogul.
This was an online platformraising capital for real estate
companies through privateplacements and A-plus REIT
offerings.
Prior to that, she was a chiefproduction officer at ReadyCap
Commercial, a small balancesheet commercial real estate
loan originator funding loansfrom private REIT offerings that
(00:44):
has since gone public.
Now Elizabeth is a certifiedcommercial investment member
this is the CCIM designation anda licensed attorney in
Washington DC, in theCommonwealth of Virginia.
She holds a Bachelor of Artsdegree from the American
University and a Master ofBusiness Administration and JD
from the George WashingtonUniversity.
(01:05):
Today, what I'm doing is I'mbringing in Elizabeth to share
her knowledge on how AI and datastandardization are key to
unlocking new efficiencies andopportunities in apartment
investing.
Elizabeth, welcome to theMultifamily AI Podcast.
Speaker 2 (01:20):
Awesome.
Thank you, patrick, it's greatto be here.
Speaker 1 (01:23):
Yeah, so much background and expertise in what
you do from a legal aspect.
I love that when we talk data,you're not here to give public
and legal advice, and nor am I,but it's interesting.
That type of background,serving at the leadership role
of a company solving data, iscompelling, obviously One of the
questions I'm thinking aboutfor leaders, business leaders.
(01:46):
So on this podcast I'm tryingto help owners and operators
that may be from sort ofnon-technical roles understand
the value that can be unlockedleveraging data in their
business, like they leveragedebt as a business tool, and
would love to know what youthink those questions should be
as people think about their datastrategy.
Speaker 2 (02:09):
Great question and I think it really obviously
depends on the organization.
I think the first thing Irecommend is really taking a
look in the mirror and sayingall right, what is a data
strategy?
What does that mean to ourorganization?
Because it can mean a lot ofdifferent things.
And what really drives the datastrategy should be your
(02:34):
business goals and objectives.
What are you looking toaccomplish with data?
What do you think you canaccomplish with data?
And then look at things likehow does it impact governance
and data security and all ofthose things.
But really starting with yourgoals, your objectives defining.
(02:56):
I'm a big fan of smart framework.
Is it specific, measurable,achievable, relevant and
time-bound?
And I go through this exercisewith our clients as well,
because when they're workingwith us on developing their data
strategy, it's really importantthat it's defined, that they
know who their key stakeholdersare, that they get buy-in from
(03:22):
their key stakeholders and thatthey're measuring what success
means to them.
Because data strategy is one ofthose things where, if it's too
amorphous, it's too vague,there's really no success metric
that an organization has.
That's when a lot of time,effort and energy is put into
(03:43):
things that eventually getabandoned because there's no end
point to it and people can'tlook and see why are we doing
this, which is important becauseyou need your team to really
participate in the process ofimplementing a strategy.
It can't be done by a vendor orin a silo.
(04:03):
One person can't take it on allby themselves if it's an
organizational data strategy.
Speaker 1 (04:10):
And getting into how to actually define this.
Many business leaders youmentioned we think about
dashboards, we think about thedata that we've always had to
make P&L decisions, knowingwhat's going to move the
investment needle in thebusiness.
So a lot of the businessleaders are more aware of their
data than they probably want toadmit, because it's really it's
(04:34):
decision making right.
But when we introduce AI anddifferent, we're talking about
standardization challenges andthings like that.
When it comes to solving thoseproblems, oftentimes a leader
will pass it off Okay, that's ITor that's somebody else's job
(04:55):
internally.
I'm curious, with theacceleration of technology and
talent, are these thingscompanies should be doing inside
their organization or relyingon outside parties to guide them
through right?
And then what I'm curious aboutis what's the leader's role in
(05:17):
understanding that data strategy?
Is it something they just handoff to IT or their technology
team?
Maybe they don't have one?
Speaker 2 (05:27):
Yeah, I was going to say goodness, no, I would hope
not.
You don't want to give the poorIT a data strategy rollout
without a lot of input from thebusiness stakeholders.
Those two things can't be donein a vacuum.
I think if an organization doeshave an IT group and a lot of
companies don't big real estatecompanies that have I shouldn't
(05:50):
say big the larger real estatecompanies obviously have
internal teams, but there arecompanies that have billions in
AUM that don't have a largeenough team to take this on
independently.
So it depends on the size ofthe organization.
But it's either done assomething in tandem with the IT
(06:12):
team or with external vendors.
But picking the right externalvendors can be great or it can
be not great.
Getting good references, butalso who you connect with, how
do you meld in terms of yourwork?
Because it's going to be aprocess and there is no quick
(06:34):
fix.
You can't bring a vendor in andjust wave a magic wand and say,
okay, data, that's not a thing.
Standardization obviously canhelp the process, but there's
still a requirement of everyreal estate company to
participate in the process ofand I decide whether it's a
(06:55):
long-term relationship that youwant to build from there and
look at this kind of as along-term strategy.
You can't spin up a completeintegration with every single
(07:18):
data source and data plan andhave it done in a week.
That's having reasonableexpectations is critical.
Getting your vendor or your ITteam to buy into those goals,
expectations, timelines, all ofthat is really critical to
having a successful datastrategy rollout.
(07:40):
But I think it's really up tothe company to work with
external or internal.
There's obviously benefits tohaving someone with completely
fresh eyes look at yourorganization and help guide that
process, and it's alsoimportant for internal folks who
really know all the nittygritty of how data has been
(08:02):
captured and stored.
So a little bit of both, I'dsay, or a lot of both, depending
on how big your data strategyis and your goals are.
Speaker 1 (08:12):
Yeah, and I love the idea of the expectations and a
lot of times when we'reevaluating even technology, I
guess you can group data into aalmost like a product right
Productization of round, I thinkyou've called it.
Is it's standardization as aservice right?
Productization of round, Ithink you've called it.
What is it?
Standardization as a serviceright?
But, more importantly, in termsof those expectations,
sometimes leaders are lookingokay.
(08:32):
So when I have a problem,what's the problem?
This technology or thisstrategy is solving and in many
cases there may, depending onthe leader's role and how they
are up to speed on where thingsare in the business.
We may not know what we don'tknow.
So how do you come into anorganization and manage those
expectations and demonstrate thevalue of getting this
(08:55):
standardization correct?
Because it seems like a lot ofpeople want to do it, but
there's some people run intosome challenges doing it.
Speaker 2 (09:02):
Yeah, for sure.
I think I always laugh whensomeone asks me is it done?
It's never done.
It's a process, having a reallywell-structured process, and
that's why I use the SMARTframework and kind of work with
teams when we're engaged in theonboarding process to really
(09:24):
define everything they want toaccomplish and then trying to
prioritize and work throughtimelines for having small wins
so that you're iterating andbuilding it out over time.
You have to see some measurableimpact when you're looking at
data and so, depending on howcomplicated your tech stack is,
(09:50):
how many data sources and datasets you're looking to bring in,
how different the datastructure is to start, and it's
why people look for scalablesolutions to this stuff so that
they can find a way to make ameaningful impact right out of
(10:10):
the gate.
Just even just getting all yourdata in one place is an
accomplishment.
Right formats, having dailyaccess to property data from
multiple systems that's a bigwin to be able to get that done.
Then you look at okay, how dowe refine our KPIs?
(10:36):
What kind of metrics do we wantto start tracking?
What different teams within theorganization want to be
involved?
What different teams within theorganization want to be
involved?
Is it all marketing or is therethe leasing team, the
management boots on the ground?
Who is this really going toimpact and how do we get them
from step one to step twothroughout this journey of data,
(11:00):
but there is no done.
Of data but there is no done.
It's always just a continuous,never-ending product process,
which is why you have to stay ontop of it and create some
business roles internally.
But it's a commitment.
There's definitely a return,which is also why you want to
have measurable.
See how you're getting a goodreturn on rolling out a strategy
(11:23):
.
Yeah, I like the small wins.
Speaker 1 (11:26):
I like what you said about the small wins and making
it measurable when data iscoming from so many different
places.
What are the challenges ofgetting this?
If it's an accomplishment toget your data in one place right
?
Speaker 2 (11:37):
Yeah.
Speaker 1 (11:37):
Are there examples that even other industries are
doing that we could learn from?
Speaker 2 (11:43):
Yeah, it's.
The big difference between realestate and other industries is
it's not the most open framework.
I think.
If you look at banking, you canadd your credentials into a lot
of systems and all of a suddenyou've got a direct connection
with your bank.
Or marketing, if you want toget all of your Google Analytics
in one place, open APIs.
That's not quite the case inmultifamily, and so we're
(12:09):
extracting data from coresystems.
Sometimes we have one, two,three different methods of doing
it.
It's challenging because you'vegot many stakeholders involved,
whether it's the managementcompany that's generating and
storing the data, ownershipgroups who own some of the data
to different property managementsystems that all have different
(12:31):
data structures.
So extracting the data is thestep.
One is getting it into a commonand consistent format.
That's the transform.
So ETL is the extract.
The transform and then loadingit into a place where you can
now use it for BI is just afront-end visualization of data
(12:56):
that's in a structure that youcan use.
But if it's coming frommultiple systems and it's in
multiple structures and it's inmultiple structures, having a BI
front end and trying to createmultiple reports from those
source systems is a lot of workand it's why a lot of
organizations are challengedwith their reporting if they are
(13:17):
trying to pull it all togetherand generate portfolio-wide
analysis.
Speaker 1 (13:24):
And obviously having a standardization of data plays
a huge role in that process,right?
Are people skipping steps inthis?
Do you think right now?
Speaker 2 (13:34):
Yeah, I think if you're doing a bespoke approach
and it's reasonable, it makessense.
If you have one system of recordand you set it up to feed data
into a front-end visualizationand then you add a second one,
now you're doing the sameprocess twice side by side.
(13:54):
But it doesn't really meshunless you take a step back and
get all that data in one format.
Because these systems are notconsistent, there isn't an
industry-wide this is ourstandard for how data is
structured and because thereisn't one way of doing it, you
(14:16):
are limited to either having tostandardize it internally and
create a data model, and surenot a lot of real estate
companies want to create a datamodel internally.
It's expensive, time-consuming,requires specialized knowledge
and it's just messy.
So that's why we built what webuilt specifically for the
(14:38):
industry, because it didn't seemlike there was a solution that
was out there that would providethat same aggregation of data
for people to be able to thenuse it for AI applications, for
BI applications, for all thecool technological use cases,
(14:59):
because the data just wasn't ina format that made it easy to
use the data just wasn't in aformat that made it easy to use.
Speaker 1 (15:11):
And when it's in this format, when you mentioned the
ETL extract, transform and load,are there things that or
questions we should be askingaround?
Making sure that the quality ofthat process gets the like, how
do you ensure that process?
How do you make sure that ETLprocess gets the right data in
the right places?
Speaker 2 (15:27):
Yeah, that's a good question.
What do they say?
Garbage in, garbage out, so wecan clean up some of the data
that comes into our system.
You can't fill in empty holes.
If there's big blanks in data,there's no way to create it.
Ai is also helping to clean andcreate methods of cleaning data
(15:53):
, and a lot of these modelsdon't need perfect data.
They need a lot ofdirectionally correct data to
generate really interestingresults and to be impactful.
Speaker 1 (16:06):
This could just also be like just as simple as a
state record right, A fieldwhere it's maybe open text in
one database and the other it'sabbreviation.
Speaker 2 (16:18):
Absolutely.
Our resident name is last name,comma, first first name and
somewhere else it's a first namefield and a last name field.
It's all across the board,getting the data in just a
format that you can query, andeveryone who's used Excel with a
sort function knows that likeit needs to be a little bit
(16:39):
clean to be able to useeffectively.
Some of the AI solutions arenow starting to fill in the
blanks of helping to cleanthings up.
So we in our tool if it'sGoogle Analytics, google with
three O's, we can combine allthose and essentially for ad
sources, we're doing somecleanup on data like that, where
(17:02):
we're just making it moreconsistent when at the property
level it's being input slightlydifferently.
Our financial model, which takesmillions of rows of data,
different categories everycompany has what?
Four, five, 600 rows offinancial data and mapping it to
a more tight, cleaner set ofdata points so that you can do
(17:28):
benchmarking.
We call it our common model.
There's huge value in havingeven a little bit of cleanup of
the data so that you can do morecomparative analysis and market
analysis.
Speaker 1 (17:44):
What can go wrong in that process, or what should we
avoid doing?
Maybe someone's trying to dothis in-house, or oh, it's a
good question.
Speaker 2 (17:53):
There's tons of things to avoid, but I'd say you
want to.
There's a lot of monitoring andtracking that your data feeds
are actually active and theyaren't stopping for whatever
reason.
You might have a night where asystem just can't handle.
It times out.
It's too much data coming in.
Speaker 1 (18:15):
In those types of timeouts.
Does that stuff get queued oris it a restart?
Speaker 2 (18:21):
Yeah, that's a great question.
It depends.
So sometimes it just like stopsand you have to go and you get
notification that you have tostart over.
Then there's sometimes you canjust pick up where you left off.
And if it's a timeout, there'salso time zone issues, which are
really strange.
So if you're doing a nightly andyou have some data that's
(18:43):
coming in, but for the change intime zones things will happen
before 12 o'clock and thensometimes after, which makes it
a different day.
So you have to make sure thatyou time all of your data feeds
so that they're not impacted bydaylight savings or, if you're
(19:03):
in a different country,different things like that.
So there's definitely ways thatthe extract process can fail.
Then there's ways that the datacan be off.
A lot of people like to do.
Timing issues can happen.
So managers will come in onMonday after a long weekend and
(19:26):
they input their move outs fromover the weekend and then you've
got different timing issuesthat directionally, can make
data not match from one systemto the next.
So that goes back toexpectations.
Speaker 1 (19:44):
Yeah, exactly, and I'm also like from all of this
work, it's the businessdecisions, almost like the
construction process is very itcan be messy, yeah, and it is.
Speaker 2 (19:56):
You're building a house a lot of dust, a lot of
unknown things, which is whyhaving a good blueprint ahead of
time keeps you from havingstructural problems.
If you set it up wrong or builda house wrong and then have to
go back and try and figure outlike oops, we didn't put the
plumbing in, or the electricityis in the wrong place, Like
(20:19):
those are the things that youwish you had done.
They say derf to do it rightthe first time.
Sometimes time makes it hard,because if you've been working
on this for a very long time,things change and you have to go
back and revisit some of thosestructural things.
But that's technology.
It's always an iterativeprocess.
Speaker 1 (20:41):
When you're in these conversations with executives
today and you're working throughthis, what are they leaning
into and what are they pushing?
Not only say pushing back, butmore cautious on?
Or if we don't understandsomething, it's hard to make a
decision about it right, and sothere's going to be a lot more
education, I think, around thisand unlocking the value.
Because you mentioned thedifferent stakeholders,
(21:03):
collaboration between teams andI always think about the
dashboard is like a lot of thetechnology that we're using is
making the assumption that we'regoing to be in an office
looking at a computer, using it.
So if you have this centralizeddata standardized and it's gone
(21:25):
through the ETL process, thenthat can be unlocked in mobile
like all over right.
You can take that wherever itneeds to be.
Speaker 2 (21:35):
Yeah.
Speaker 1 (21:36):
Yeah.
Speaker 2 (21:37):
That's the load portion.
You can load it into a numberof different applications or
uses.
But I think a lot of people arerightfully so leaning in on
corporate data governance,wanting more visibility into
tracking their data feeds andmaking sure that they're
reporting on the rightinformation, especially if
(21:59):
they're a public company or haveREITs with reporting
requirements.
All of those things arecritical to have a handle on how
the data is coming in.
Speaker 1 (22:11):
Tell me more about that data governance.
Speaker 2 (22:12):
Explain that, what you mean by that in, so explain
that what you mean by that.
A lot of companies, at leastlarge corporations, are looking
at how they secure and storetheir data, how they're
maintaining a handle on theaccuracy and also the security
issues related to data.
(22:34):
The multifamily industry we'redealing with residents and
residents' housing, and housingis obviously a highly regulated
industry.
The payments and activities ofresidents, their data is PII, so
the storage of their data canbe highly regulated.
(22:56):
There's a lot of different.
So when we look at corporategovernance, it's really how does
a company maintain the highestintegrity and policies and
documentation and legalstructure around how they are
storing, using, maintaining data?
(23:17):
So it's some effort to be doneinternally for big companies to
make sure that they're putting aplan in place and documenting
it appropriately.
Speaker 1 (23:28):
And do you offer?
Is your program a data store oris it wherever they want?
Is it going to other places, orhow does that work?
Speaker 2 (23:37):
It can be yeah, so some companies look to have
their data fed into a datawarehouse.
Some companies want us tomaintain and store their data
for them.
We have SOC 2 compliance, soall of our data is encrypted in
motion at rest.
(23:58):
Very few people have access tothe data and all of our policies
and procedures are available inour SOC 2 reports, which you
can get on our website.
But you have to sign an NDAbecause our auditors require
that.
But the whole process is not aninsignificant amount of effort,
but it's really important andsomething that I always
(24:19):
encourage people to talk totheir vendors about, because it
is incumbent upon the realestate companies to make sure
that they're the ones that couldbe liable for any data breaches
if they're not maintaining goodpractices, if they're being not
(24:39):
careful with how they store andmaintain their data.
Speaker 1 (24:44):
And leaning back into AI, how is AI impacted?
It seems like a lot of thesetransformations and
standardizations are gettingeasier.
There still needs to be thegovernance side of things.
It's one thing to turnsomething over to AI, but I'm
under the belief that peoplebegin and end the process, so
maybe it's taking out some levelof review.
(25:07):
But how is AI impacting thestandardization process for your
perspective?
Speaker 2 (25:12):
Well, it's a great question and there's obviously
lots of different applicationsfor AI.
The AI is thriving.
The need for standardizationbecause AI needs like data in a
structure that it doesn't havein its knowledge base.
(25:35):
It kind of lies and gives youanswers that you're a little
confused with because it's likewhat is it talking about?
It makes stuff up and that'sbecause it doesn't have the
right data to answer thequestions.
It tries to fill in the blanks.
It's being helpful, but notreally the blanks.
(25:58):
It's being helpful, but notreally.
So there's certain things thatAI is driving in terms of the
need for standardization, andthen there's, equally, ai is
providing standardization withmore ways to make data clean,
usable, accessible, so they feedeach other in certain ways, and
how much faster and better thetechnology is at letting us make
(26:24):
informed models using availabledata.
So it's requiring less and lessas long as the models are
trained and maintained.
But yeah, I know it'sinteresting.
They feed each other in someways back and forth.
But there's so many AIapplications and real estate has
(26:46):
been pretty lagging in howthese applications are being
utilized.
I think there was a good studyMcKinsey put out and if anyone's
interested, feel free to emailme.
I'll share it, but it wastalking about how the data is
really the shape of data.
The lack of a foundational datamodel can really hinder
(27:08):
people's ability to use data ina lot of these AI applications.
Speaker 1 (27:14):
And these large language models are moving so
fast and it's so expensive to dothat.
It's interesting, like when youcompare it with your private
company data in an environmentwhere it's not training the
large language model.
But it goes back to what we'retalking about here, which is, if
you don't have thestandardization of the data or
in the data in the right place,you can't retrieve it and then
(27:37):
get better responses when you'reusing AI applications inside
workflows.
And so I always wonder, asyou're moving companies' data
into a single source of truthand also standardizing it, what
are you seeing the challengesthere?
Is it they have differentsystems?
(27:58):
Maybe they're a third partyoperator and they don't just
have one system?
They're trying to managemultiple systems and I imagine
training people and all of thatcomes into play.
But in terms of the technical,what are those challenges in
bringing all these thingstogether?
Speaker 2 (28:15):
Yeah, that's exactly right.
It's ownership groups who workwith multiple property
management companies.
It's property managementcompanies that have multiple
ownership groups that requirethem to use different core
systems and or are buyingproperty management companies.
So sometimes it's just acompany buys another company,
(28:37):
and it's just the nature of realestate.
It's transactional in nature.
So when you have a sale of oneproperty to another buyer, are
they going to use the samesystem, are they going to use a
new system, and so you've got alack of kind of historic data
(28:59):
and then you're onboarding itinto a new system.
So anytime you have peoplemanually adding data into
systems, that also add some roomfor error, we might say and
there's just, they'reinconsistent.
They're not.
These systems weren't made tobe exactly the same and have the
same definitions and have thesame data structure.
(29:22):
They just they weren't set upthat way.
And there's not that manysystems out there.
It's not.
The percentage of the realestate market that's
professionally managed is thatit's consolidated, but those
companies aren't using a ton ofdifferent core systems and those
(29:42):
core systems really arecapturing all the financial data
, all the operational data, andthen you have all this ecosystem
of prop tech companies thatkind of have opened in the last
however many years, really takenoff, and so there's lots of
stuff, but that's just addingmore data to the problem.
Where is it being stored?
(30:03):
Where the core systems aren'tstoring that data, and these new
PropTech solutions aren'tstoring all the underlying
property data.
So you're just creatingadditional data silos.
You're exacerbating thefoundational issues that exist
with new technology.
So it makes your head explode.
Speaker 1 (30:26):
Yeah, it can.
When you think about thesecompanies, companies', entities
stay around longer thanproperties in the investment
cycle, are shorter term, right,and so we're asking people to
think longer term about thingswhen they're in a short term
investment cycle, right.
So that's always been achallenge.
But if you think about theentities themselves and also the
(30:48):
data, you may have company dataand then you have property data
and in the way that all of thatis organized, I'm just curious
if and it's probably true, thiswas my belief in that when a
developer builds an apartmentbuilding, the utilities and the
city infrastructure is valuethat a new buyer doesn't have to
(31:11):
come in and do and do.
In other words, because somebodytook the risk and developed the
plumbing, the electrical, theutility, all the things that are
required under the ability togo vertical, the new buyer
obviously gets that right, butnobody really values it on the
sale of an asset, right?
Yes, but when we talk about, ifyou just think about the
(31:35):
listeners, just think about allthe different entities you have
you may have your company entityand then a holding entity and
then each property has an entity.
Is that a way to think throughdata in terms of if we get
property data at a specificlocation that later, if there's
a disposition or a consolidationor merger, that there's
enterprise value that can berealized on the exit by making
(32:02):
investments in getting thisclean data for the new buyer.
Speaker 2 (32:06):
Yeah, I totally think so, but it's really up to
sellers to view that as an assetor buyers to ask for the
historic data of the seller.
They may or may not feelcomfortable doing that.
Speaker 1 (32:22):
You're lucky to get a PDF today, I know right.
Speaker 2 (32:24):
Are they going to?
I guess, if the market getstight enough, you could put that
into purchase contracts, thatyou want two years of historic
data or access to historic datafiles.
I think it'd be really hard toget a seller to agree to it, but
it would be great for a buyer.
Speaker 1 (32:42):
Let me just ask though that's the business kind
to me again what would that meanif you had that on a new
acquisition?
Speaker 2 (32:51):
If you had it on a new acquisition, you could
probably track the trends ofyour performance a lot better.
So if you're trying to and Ithink this is one of the best
use cases of data is to showyour investors that your
operational strategy bringsalpha, that you have somehow
managed to create opportunity inthe asset by doing X, y and Z,
(33:18):
that's hard to prove when you'restarting with ground zero and
there's nothing to compare it to.
If you had the historic data,you could show what's been doing
this for the past 12, 24, 36months and even adjusting for
market conditions.
And even adjusting for marketconditions.
This is how much value we'vebeen able to add.
(33:38):
This is how much that we'reputting into the property.
That's creating a bump in NOIand this is the multiple that we
, as a data-driven companypeople like to say we're a
data-driven company to prove it,to actually show that what
you've done has created thismajor impact.
(33:59):
I think that's super powerfuland it's hard to raise capital
right now.
That's a tremendous use case.
Speaker 1 (34:07):
We're talking about the apartment car facts
basically.
Speaker 2 (34:10):
Yeah, some people have been called it what's like
the for hotels, the report thatkind of shows the market value
for that.
Speaker 1 (34:18):
yeah, sure, but it'd be interesting.
You buy a property and this isthe data from the time that we
owned it, but imagine, as ittransferred ownership over three
decades, that it would be morevaluable in terms of
underwriting, repairs,maintenance and remaining life.
There's so many other.
Where are we in all?
Speaker 2 (34:39):
of this.
We can do that with financials.
We can do that with thefinancials.
So if you have a property thatyou're looking to acquire and
you have the trailing 12 monthsand you can get maybe the past
couple of years, we can showfrom a financial perspective how
you can compare from afinancial perspective.
(35:00):
How you can compare because wecan convert it into a common,
consistent format with yourcurrent performance, your
current tracking.
So being able to take your proforma track to that track, to
budget, but tracking to historicperformance.
It's a really interesting usecase and, for sure, something
that I encourage people to do,if not with our platform, but
with their own internally seehow a property was performing
(35:25):
and how your efforts havechanged the story at that asset.
That asset is really compellingwhen you're going out to market
and trying to get people toinvest in you.
They say past performance isnot an indicator of future
results.
But when you can showcomparative how this property
was doing compared to the marketand then how we're doing
(35:46):
compared to the market, it justtells a really interesting story
of performance.
Speaker 1 (35:54):
Yeah, that's interesting and I know that
there's so much more that we cancover.
On all of this, I go back towhere we started, which was
getting those small wins.
I've been calling it crawl,walk, run right.
So it's maybe somebodylistening, has access to data,
obviously, but maybe doesn'thave a defined strategy in a
market where that alpha beatingthe returns in other ways than
(36:18):
just passing on rent increasesif we can find ways to innovate,
the business having that datais huge in that decision-making
process.
Do you ever find people getoverwhelmed with this because
they are not data scientists andthere's not an abundance of
people to help with this andrelying on the tools you have
(36:40):
available to you?
And there's just gaps, that theydon't see, that they're missing
, and we can go down the wholeunstructured data conversation
and getting that into astructured Like that's a whole
nother, like phase two probably,of unlocking value.
But are you feeling or do yousee that people get overwhelmed
with this conversation and howdo they prioritize making the
(37:02):
next move and making sure it'sgoing to drive actual
effectiveness in the measuredresults?
Speaker 2 (37:08):
That's a really good question, I think.
I say don't try and boil theocean.
Pick a couple of discrete KPIsthat every company has.
The thing that they think isthis is the thing that we are
really good at.
Speaker 1 (37:23):
Where would?
Speaker 2 (37:23):
be a good place to start.
Look at reducing days on marketor days to if you are turning a
unit.
If you can turn a unit in twodays less than you have been
Huge impact on a property.
So what we do is we set up, youtrack your goal, you put in
(37:49):
alerts, you have a start dateand an end date and you look at
what is the overall impact andthat's something that's truly
measurable.
If you're dropping two days fromthe days that it takes to turn
a unit, that reduces your dayson market overall, adding two
more days of rent to each unitthat you're turning.
That is something that you canshow adds to your NOI and adds
(38:14):
to the value of your asset.
Same thing with looking atreducing an expense, your
marketing expense, if you'reable to drop it.
Or I look at customeracquisition, your lifetime value
, looking at that as a ratioacross your portfolio at each
property and setting a targetfor I'm going to reduce my cost
(38:37):
to acquire residents but I'mgoing to increase the lifetime
value of that resident.
Those are things that you canreally show to investors who are
being really thoughtful, notjust to transactionally turn
these tenants but actually toincrease the value of the
housing experience for them.
(38:58):
Renewal rates, increasing thosethat's a huge way to increase
your overall.
If you're looking at yourtrade-outs, these types of
things, sitting down and picking, just start with three, start
with five, but looking at, if Idid this and then was able to
(39:19):
accomplish it, what does that doto my property and portfolio
and how do I communicate that tomy key stakeholders, whether
it's a property managementcompany wanting to tell their
ownership groups we've added$213,000 to the property value
this quarter by doing X, y and Zand we help to automate the
(39:41):
process of communicating that,of showing that value.
That has tremendous ROI.
Speaker 1 (39:48):
when you look at things that way, yeah, and do
you find that people want to dosomething like that?
Look at those.
They have tools that aim todeliver some element of that,
but it's in that environment,right?
Especially if you're inmultiple systems.
That's where it becomes trickyto understand that stuff.
And do you find that when youget these small wins under their
(40:11):
ability to measure somethingthat truly makes a financial
difference in the business thatthey lean into and get more
motivated into investing moreand more into the data strategy?
Speaker 2 (40:22):
Oh, absolutely, it's almost.
You want to gamify it in a lotof ways Get your managers
involved or your analystsinvolved in creating these goals
and tracking to them, becauseit's not one property, it's the
entire portfolio.
So, each person doing theirpart and making these small
(40:49):
incremental changes, they canhave a huge overall impact to
the value of the portfolio justby being more thoughtful about
how you look at your data andhow you track to performance in
small ways.
Overall, that adds.
It adds up.
It all adds up.
Speaker 1 (41:08):
Yeah, it sure does I always think about.
When you mentioned the slipdays of time to turn, especially
in renting.
It's like empty airline seatsthat we just can never recapture
.
And that's that alpha, as youmentioned, that you can offer a
definitive or differentiatingvalue in running your business
through this certain managementcompany.
Listen, this has been great,Elizabeth.
(41:30):
I really appreciate you comingon.
I know that I would love tohave you back for more on this.
For those of you listening, ifyou want to know more about AI
and data for better apartmentinvesting, reach out to us and
let us know what are yourquestions that you have around
this, specifically to yourportfolio, to your use case, and
just go tomultifamilyaipodcastcom.
You can click there and you caneven send us a message, but
(41:52):
reach out to us, let us know andgo find that information.
Probably Elizabeth has it topof mind, but would love to visit
back with you, Elizabeth, as wecontinue down this journey of
extracting the knowledge ETLextracting the knowledge,
transforming it in a way thatnon-technical business leaders
understand it and then loadingit into their business.
That was a pun on the ETL.
Speaker 2 (42:13):
I love it.
No, that's so great.
Speaker 1 (42:15):
Yeah, look, you're doing great.
We love that you're making animpact in this industry and we'd
love to talk to you more aboutthis.
But until then, if you want tolearn more about Elizabeth and
some of the projects she'sworking on, go to Multifamily AI
Podcast.
In the show notes you can clicklinks and get to her LinkedIn
and all the other great placesto connect with her.
Until then, wishing you guysthe best and we'll see you on
(42:37):
the
Popular Podcasts
True Crime Tonight
If you eat, sleep, and breathe true crime, TRUE CRIME TONIGHT is serving up your nightly fix. Five nights a week, KT STUDIOS & iHEART RADIO invite listeners to pull up a seat for an unfiltered look at the biggest cases making headlines, celebrity scandals, and the trials everyone is watching. With a mix of expert analysis, hot takes, and listener call-ins, TRUE CRIME TONIGHT goes beyond the headlines to uncover the twists, turns, and unanswered questions that keep us all obsessed—because, at TRUE CRIME TONIGHT, there’s a seat for everyone. Whether breaking down crime scene forensics, scrutinizing serial killers, or debating the most binge-worthy true crime docs, True Crime Tonight is the fresh, fast-paced, and slightly addictive home for true crime lovers.
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
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.