December 9, 2024 • 32 mins
Join host Adam Larson as he chats with Court Watson, Partner in Deloitte Advisory’s Controllership services practice. Discover the future of accounting through Court's insights on data science, generative AI, and modernizing financial practices. From tackling manual journal entry inefficiencies to exploring AI-driven dynamic account closures, this episode offers plenty of real-world takeaways.
Gain practical tips on integrating innovative tech, fostering continuous learning, and upskilling for the evolving landscape. Whether you're curious about AI's potential or seeking new ways to enhance your role, this engaging conversation is packed with inspiration and actionable advice.
Tune in to elevate your accounting game with invaluable insights from Court Watson. Don’t miss out!
We also invite you to listen and subscribe to Deloitte’s Resilient Controller podcast featuring controllers and financial executives who put the attributes of authenticity, determination, trust and grit into action.
Sponsor:
Today's episode is brought to you by U.S. Bank. U.S. Bank is a trusted financial partner for our clients, businesses and communities. We believe in doing the right thing and putting people first. It’s an honor to be recognized as one of the World’s Most Ethical Companies® by the Ethisphere Institute for the tenth consecutive year. From commercial credit cards and program management tools to innovative payment technologies and transportation offerings, U.S. Bank Corporate Payment Systems has the right solution to help your organization reduce payment costs, enhance control and streamline your entire payment processing function. We’ll partner with you to uncover your challenges and provide smart, clear and honest guidance to help you meet the financial goals for your business. Visit usbank.com/corporatepayments to learn more.
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Adam Larson (00:20):
Welcome back to Count Me In. I'm your host, Adam
Larson. And today, we're excitedto have Court Watson, a partner
in Deloitte's Advisory'scontrollership services
practice. In this episode, weexplore the evolving world of
accounting and how modernaccountants need to adapt to
technological advancements.We'll discuss the transition
from traditional reporting todata driven methods, the
(00:41):
importance of continuouslearning, and the impactful role
of AI.
Core provides insights onbridging the gap between
accountants and IT, leveragingAI and future proofing
accounting practices. Join usfor real world examples and
practical advice to help youstay ahead in the fast paced
accounting landscape. But beforewe get started, we wanted to
give a quick shout out toDeloitte's Resilient Controller
(01:03):
podcast. It's a fantasticresource for controllers and
financial executives. You canfind it in your favorite podcast
platform, but we'll also includea link in the show notes.
So, Court, just really excitedto have you on the podcast
today, and we're gonna becovering a lot of different
things around the controllershipand AI and how it's all it's all
(01:25):
being reshaped by everythingthat's happening within the
industry. And I figured we couldstart off by just talking a
little bit about AI and itseffect on the controllership
within organizations because,accounting and finance teams are
having to learn a lot of newskills now with all these new
tools coming up.
Court Watson (01:40):
Yeah. I mean, I think if we look at generative
AI, we've sort of changed thedefinite definition of AI to
include many things thathistorically weren't machine
learning. And we're sort of allbucketing them into that that
category. Generative AI is stillin its infancy. And I think,
really, the impact on accountingorganizations today is more of
(02:03):
an inward or internal pressureto adopt AI.
And controllership, I think, isa unique space, specifically if
you're a public company, right?Because there are requirements
around how you use technologythat don't necessarily exist in
the rest of the organization,and they are a an external
requirement or regulatoryrequirement. So I need to not
(02:25):
only know that I havecompleteness and accuracy around
everything that that a systemdid, but I also need to
understand at a pretty goodlevel how the system came to the
answer it did. And if you lookat a lot of what's out there in
AI right now, it's a little bitof a black box in understanding
(02:45):
how the output was created.Right?
We're sort of relying on it todo that, and I think there's
there's many cases where, youknow, you've asked some sort of
generative AI to do somethingand came out with an answer that
that didn't really align withwhat you thought. Yeah. But that
opaqueness needs to get resolvedbefore I think you're gonna see
wide adoption of of AI in theaccounting space. Now I still
(03:09):
think right now it's a reallygood sort of overeager assistant
who can give you a first pass,but the accountant still needs
to know on what what it did tohave the domain or subject
matter expertise to be able tosay, hey. That that makes sense,
or it doesn't.
(03:29):
Here's why. And so there's 2components. The accountant is
going to gradually have toupscale in that sort of the AI
domain and, I think, understandmore of an AI model uses linear
regression to produce a expectedoutcome on, let's say, an
accrual or a managementestimate. The accountant needs
to understand the metrics thatcome out of a linear regression
(03:54):
model and whether or not thatregression model is a usable
model or is not reliable. Andthat's a kind of a new skill set
for a lot of of accountants, sothere's going to be that
upskilling over time as it comesto AI before you I think you get
widespread.
Adam Larson (04:08):
Yeah. Well, and like you said, it's still in its
infancy. And when something isin infancy, everybody sees a
little infant baby. They'relike, oh, how cute. But then a
few years down the road, whenit's a crazy toddler making a
ruckus throughout therestaurant, nobody's saying how
cute.
They're like, oh, this is such anuisance. So it'll be
interesting to see whattrajectory this new generative
AI goes because and when it's inits emphasis, everybody's giving
(04:29):
each other high fives. And Ithink what people aren't really
talking about is some of thechallenges that it's bringing.
But, like, you mentioned the onething about controllers, you
know, having to know whysomething was, and you don't
always see that with the AI. Arethere other challenges that are
being brought up as their peopleare getting into it?
Court Watson (04:44):
Yeah. I mean, the skill gap is certainly there. In
order to run and maintain anysort of sort of technology, you
need to have the requisiteknowledge to make sure that it's
it's updating. Finance data isstill a big gap for a lot of
companies. Mhmm.
(05:04):
You're often gettingtransactions that are incorrect,
that are incomplete, that aren'tall sitting in one location,
that aren't all brought togetherfrom various source systems,
various inbound interfaces inone place that an AI model can
sort of access and tell anarrative or tell a story from.
Now as, you know, as AI getsmore robust, it's going to be
(05:28):
able to ping all the differentsystems and bring it all
together. And that's truly, Ithink, what eventually the
strength of these AI models willbe. But today, we don't we don't
have that. How do you quantifyrisk?
Risk and material misstatement,how do you put a number to that?
Like, yes, it might not savetime, but it might produce a
better quality outcome. How do Ijustify that in terms of ROI? So
that that's still being figuredout is what the ROI those use
(05:50):
cases are.
Adam Larson (05:52):
Well and you also need still need the human
interaction like you mentionedbecause we all know about the
hallucinations. We've all seenthe examples of them. We've seen
the examples of the lawyer whocreated this whole speech out of
it. Not that we're talking aboutlawyers, but there's lots of
popular examples that are outthere, and you still need that
human aspect because, like yousaid, it's a really eager
assistant giving a first draft.
Court Watson (06:14):
Correct. Yes. And I I will go every 3 to 5 months
into one of the language modelsthat I have access to, and I
will ask it the same accountingquestion each time. And I always
I always pick foreign currencybecause that's a topic that is
that is pretty complex but isvery rules based. Like, if you
know the rules, you can do thecalculations.
(06:37):
And the generative AI toolssound really, really smart, but
they're also really, reallyincorrect. And that's always my
fear is someone is relying onit, but doesn't understand the
subject matter, uses that, andthen produces something that
results in a incorrect orinaccurate filing related to
(06:59):
financial statements. And Ithink that I think, at least the
conversations I have, everycontrollership at every company
is skeptical enough that they'renot at that point. So I don't I
don't think it's a pervasiverisk, but I think it could
happen.
Adam Larson (07:13):
And should we be worried about, you know, the
next generation of accountantsthat are coming up? Because, you
know, my oldest is is is afreshman in college, and, you
know, I know she's using a lotof generative AI to kind of help
give first drafts or or getideas for things and stuff like
that. And a lot of people aredoing in their works. But if
people are utilizing those andnot necessarily understanding
the having a good foundation,will will we have a gap in a few
(07:37):
years when those this nextgeneration is coming up and
doing those low level jobs in inaccounting firms?
Court Watson (07:43):
So I think for future generations, I think
there is going to be a greaterneed for sort of logic,
philosophy, ethics. And thereason I say that is I think
you're going to see generativeAI replace a lot of the menial
tasks that are day to day, andthe individual is going to spend
(08:07):
less time preparing and moretime analyzing. And that
analysis requires sort of takinga step back, thinking about sort
of the macro picture and tellingthat story, the way the
regulatory environment is, it'snever going to be there's not
people, right? You always needyou can't have generative AI
sign off on a 3 zero twocertification for a financial
(08:29):
statement. There's always goingto need to be individuals.
And if you look like companiestoday, they don't close the
books in a day. They close it in5 or 6. And if you look at those
5 or 6, right now, there's a lotof manual work. I always say
manual journal entries are adefect of your technology.
Right?
(08:49):
It's just whether or not thecost benefit of resolving that
defect is there. If it'ssomething that takes 5 minutes
to prepare and book, are youreally going to industrialize
the solution? Because it couldtake 6 months and, like, a full
software development lifestyle.It's like, well, no, I'm not.
I'm going to keep doing themanual entry.
Adam Larson (09:05):
Yeah. And you
Court Watson (09:05):
sort of pick off your large ones and fix it. And
those are all processes in theclose. You're like, how do we
fix? If I look at, like,accounting technology, I
wouldn't be surprised, and Idon't think it'll happen soon,
but if someone sort of thinks ofa ground up accounting
technology solution around theERP, Like, could there be a
(09:25):
world where the tax provision,instead of booked on day 6 of
close after day 4 where you'vefully closed the books and
consolidated, is being booked onthe fly with every journal
entry? Because leveraging AI orleveraging technology, it's able
to sort of think about, okay,what goes into the provision?
How do these transactionstypically impact our results? Is
(09:48):
the technology able toconsolidate on the fly versus
now where you have to press abutton at the end of the period
and bring it all together? Likeall those things, I think it'd
be really cool. And I would loveto be part of something where
like blank slate, ground up,ERP, subledgers, closed solution
from scratch, embedding currenttechnology. I think you're
(10:08):
seeing the large manufacturersbring that in over time.
And and I think that's what alot of control ships waiting for
is is the proven industrialsolutions to have AI embedded in
their solutions because theyknow they're going to have
controls around that. They'regoing to have a SOC report. The
controlship organizations willbe able to say, hey, I can rely
on this because it's got it'sgot that right. And the
(10:30):
industrial solutions know whataccountants need in terms of
validating that that whathappened makes sense, and they
understand it. And so ratherthan building custom and bespoke
solutions like we're seeing inother domains in a company,
accounting is sort of like,let's see what the major players
do and how they bring it intotheir value proposition so that
we can
Adam Larson (10:51):
That makes a lot of sense. And what you describe
would be really cool, but I feellike because of how things move
in this industry, it would takea long time for adoption. You
know, a lot of times you see,you know, all these cool things
going up at rest of theorganization, but the accounting
team is still using the samesoftware they used 20 years ago.
Why do you think that is thatit's there's a long adoption
period, especially in theaccountings, accounting and
(11:13):
finance teams?
Court Watson (11:14):
So I will maintain that, like, you cannot have
agile software development as itrelates to accounting
technology. And if if you wantan agile solution, well, your
first release has to be all therequirements, which yeah. It has
to be. Right? Like Yeah.
If I build a revenue recognitionsolution and it meets half of
(11:37):
the accounting standard, well,then it's not usable. That
that's one piece of why it'sslow. The other thing is the
other piece is accounting is acost center. Right? And a, and a
business has to ration its ITbudget and where is it going to
spend, you know, where it thinksis most important and totally
(11:57):
makes sense that, you know,customer facing revenue
generating parts of the businessget that priority.
Totally get it. And then theother thing is, you know, there
is, and it's funny, I'll tellsort of an anecdotal story.
There's a gap between how anaccountant describes what they
need and how someone ininformation technology or a
(12:18):
developer builds a solution. Andone of my colleagues did a
presentation, a CPEpresentation, so you get, you
know, CPE credits for your CPAlicense. And he had accountants
write out the requirements formaking a peanut butter and jelly
sandwich.
He took one and he brought aloaf of bread, peanut butter,
(12:40):
jelly, a knife, and acted outthe requirements. And he ended
up smearing peanut butter on topof the bag of the bread because
there was never a requirement toremove the bread from the bag
and then just jamming the jar ofjelly on top because there was
never a requirement to open thejelly and use a knife to spread
it on the bread. So it's like inan accountant's head, it's like,
(13:03):
well, I need to, you know, Ineed to book a bad debt
estimate. And, okay, you mightin your you might verbalize 4
steps. Well, someone who has noidea how that comes about is now
in charge of developing it, andthey're not gonna come up with a
solution that meets theaccountant's needs.
And that just creates moremanual processes on top. And so
(13:27):
finding that balance, it'sreally hard to find a unicorn
who knows accounting, knowsaccounting standards, but also
knows how detailed requirementswork, how user stories work, how
testing works. Yeah. Andbringing those together is hard
and often leads to unsuccessfuloutcomes that actually, you
know, kind of make kind of makecontrollerships the
organizations go a bitbackwards. I've seen instances
(13:49):
where the close cycle haslengthened because of new
technology, and a lot of it isbecause companies did not
rethink the processes they haveand whether or not they could
standardize, and then the gapbetween what the requirements
are for the customer, theaccounting organization,
relationship organization, andthe team that implemented the
solution.
Adam Larson (14:07):
Well, and I think that really points how important
it is to have good internalcontrols, good to have good
processes. And if you're gonnaimplement any new technology,
you should look at all yourcurrent processes to see how
could you improve upon them whenyou're when you're looking at
those things.
Court Watson (14:20):
Yes. And and that ties back to control ship
organizations are alreadyreally, really lean. Yeah. How
do I spend meaningful time onthat when I barely have enough
headcount to do the day to day?So I'm trying to do the day to
day, but also trying to do thisextra incremental thing to spend
meaningful time revisitingprocesses, redesigning
(14:40):
processes, redesigning controls,and producing requirements.
Like, it's a massive challenge.I
have seen some companies go aroute that I thought was really
interesting where, you know,they baked into the budget of
the the initiative or thetransformation effort, co
sourcing. So they would bring ina third party to backfill their
accountants so their accountantscould focus on that. And that,
(15:01):
to me, makes a ton of sensebecause then you're leveraging
the people with the knowledgeand the expertise, and they're
focused on it. And sometimeswhat you see is your existing
processes that you're someoneelse is coming in to do, they'll
see some things and bring in adifferent perspective that
potentially makes it better aswell.
Adam Larson (15:18):
I sometimes hate, but I also love when somebody
new comes in and takes an SOPthat was written. And then you
you with by all based on thequestions they give you, you
realize how badly written yourSOP was. And I think it's a good
practice for all organizationsto, at some point, bake in, hey.
We need to look at our practicesand make sure we're doing things
efficiently because otherwise,you'll continue on and maybe
(15:40):
never grow Correct. Within yourwithin your team.
Court Watson (15:43):
Yeah. And, you know, that's not only logically
makes sense, but human nature islike, this is mine. I'm gonna be
defensive if you say it's bad.Right?
Adam Larson (15:51):
Exactly.
Court Watson (15:52):
We can we can make fun of our own family, but if
you're on the outside, don'tmake fun of my family. That
that's totally different. Andand it's not much difference. If
you have any pride when you do,you're like, who is this person
telling me I'm doing it wrong orshould do it differently? But
also like from a controlsframework, that's good
governance.
Like periodically rotatingpeople into different roles
(16:12):
helps you detect things that maybe amiss. And that's one of the
recommendations of a solidcontrols framework. So it's sort
of dual hatting in its purpose.
Adam Larson (16:20):
It really is. Now, do you think that as some of
these bigger software companiesthat are kinda old hats that
have been around for a while asthey incorporate some of this AI
features into their softwaresand people start to get into it,
do you think we'll start to seean ROI within those systems, or
is it just, hey, this is just anew cool, tool, but I'm not
really gonna use it very often?
Court Watson (16:39):
I'm gonna give an accounting answer and say it
depends. Like like like, I thinkit depends on how well
integrated it is, how clearlyexplained it is to the customer,
and then how willing thecustomer is to change. And I
think with a lot of companies,they do a disservice by not
being transparent of what thelimitations are or what
(17:02):
companies need to have in placefor this to work. And I think
when you do, especially withaccountants, it resonates much
better. Like, hey, I understandit's not gonna do it all, but I
understand what I need to do toget to that point.
The other thing is AI is stillquite expensive. Mhmm. And
you've seen situations where alicense will double in cost for
(17:25):
an AI functionality that is kindof an administrative assistant.
It's like, well, do I reallywanna pay double for that,
versus just like an incrementalvalue? So that pricing, I think,
needs to get figured out.
But, you know, there are a lotof industrial solutions, both in
your core finance technologyplus what I call edge systems,
so your reporting layer, yourclosed like closed tasks, closed
(17:48):
tools. And they're starting tolook at it a lot, and they're
starting to try and incorporatethings. And I think it's going
to be in incremental stepsbecause I think they're learning
along the way too of what workswith the customer, where the
feedback is, where the frictionpoints are, where clients, where
customers, you know, incontrollership are like, Hey, we
can't use this because we don'tunderstand what's happening, and
we can't get past regulatoryquestions around that. So it's,
(18:14):
it'll happen think it's gonnahappen in graduated steps.
Adam Larson (18:18):
Yeah. In graduated steps. That makes sense. Do you
think there's steps that we canbe taking now to make sure that
the, we can get over thatcommunication gap? Because
you're mentioning that we'rewe're talking about how, like,
it's hard to have somebody whounderstands all the steps and
accounting and understands howto write the stories and the
processes for for things to workproperly and and to make sure we
get all those steps down,especially for technology work
(18:39):
properly.
Are there are there ideas of ofor or steps people can take to
kinda get over that gap? Becausethat's probably the hardest part
of getting over the thecommunication gap that's there.
Court Watson (18:52):
Yeah. So there's a very famous technology
evangelist, and he recentlyMhmm. Posted and said, hey. You
know, artificial intelligence isthe gas, and regulations are the
brake.
Adam Larson (19:05):
Mhmm.
Court Watson (19:06):
And my first thought was, well, the better
the brake, the faster your topspeed can be. Right? Okay. If
you're on a bicycle and you haveno brake, are you going to go 40
miles per hour? Are you going togo 5?
Because you can't stop. Right?Whereas if you could, you can
stop and control. You're gonna,you're gonna be able to go
faster. And I think definingwhat that break is, what the
(19:28):
governance around the model is,what you actually need to see
from a model to get comfortablewith it.
I don't think anybody spentmeaningful time other than
saying this model doesn't giveme enough that I can rely on it.
It should be like, what do youneed to rely on it? I need
completeness. I need accuracy. Ineed to understand the inputs.
(19:48):
I need to understand thetransformation of the inputs and
how it derived it because at theend of the day, I need to be
able to back test it. I need tobe able to go back and look and
say, did it work? Did it do whatit was intended to do? So
spending time building out thatframework and what that looks
like, and for right now, it'sgoing to be company by company
basis because they are all goingto have different measures of
risk. They're all going to havedifferent materiality levels,
(20:10):
and they're going to have towork on that.
The other thing is cleaning yourfinancial data, going back,
where do we have gaps? And a lotof companies are doing this as
they have to and are required tomodernize their finance
technology because we're kind ofin a cycle of, like, the updates
are occurring, but going throughthat, really thinking with a
future sort of focus of, like,yeah, I don't use this dimension
(20:31):
today consistently, but I havean opportunity here to clean
that populate because I knowit's gonna have value in the
future. So those are the 2things that companies can do now
for sure.
Adam Larson (20:40):
Definitely. Do you think this is gonna have a
greater impact on small tomedium sized businesses?
Because, obviously, the largeorganizations might have the
capital to put the time in. Butsmall organizations, you know,
you've already mentioned thecontrollership is very lean. A
lot of finance and accountingteams have had to cut down and
and make it more lean and do thesame amount of work.
How are we gonna get these extraskills and upskill and be ready
for this in the midst of thisfast changing market?
Court Watson (21:04):
It can go both ways on that question. Okay. So
it may be easier for small tomedium businesses because their
systems are going to berelatively cleaner and simpler.
Adam Larson (21:17):
Okay.
Court Watson (21:19):
At the same time, AI is expensive. And for it to
be fully built into software andproducts, I think, you know, the
companies that produce financetechnology with the capital and
resources to embed AI
typically don't target the smallto medium sized business. Right?
(21:39):
That they're gonna go up, youknow, they are the the the
vendors for the Fortune 500companies. So there's going to
be that cost component.
But I can see where thedeployment will be much easier,
you know, at a small businessthat uses one sort of web based
ledger to track expenses andpayroll and everything and
(22:03):
really doesn't have a complexstructure of a multinational
that, over several acquisitions,uses 4 different ERPs and 7
different customer billingsystem. That just doesn't happen
in a small medium business, sothe opportunity to get the most,
I think, will be an easier lift.It's just in affordability.
Adam Larson (22:21):
Yeah. That makes a lot of sense. Well and and you
can go either way to from thefrom the the software company's
perspective because why not testit out with a bunch of small
companies, and then you can rollit out the big ones, but then
you're not doing anythingcomplicated. So it's a a give
and take depending on whatyou're doing. Yeah.
Court Watson (22:36):
And, you know, a lot a lot of what you do is is
proof of concepts where you picka component of a large business
and test it. I think that wouldbe a great idea to find that mid
market client that's willing toto be involved in those pilots
and provide feedback. That Ithink that's the best way you're
going to build a solution thatworks for everyone. It's just a
totally different beast when youget a multinational that has
(22:58):
been around forever. They haveprocesses they've had forever
that don't always alignnationally.
Again, systems are different allover the globe. It's just is a
very large effort to standardizeand align across multiple
geographical jurisdictions.
Adam Larson (23:17):
Yeah. So we've talked a little bit about
upskilling and this new skillsthat finance and accounting
teams are gonna need to have.Maybe we can dig a little bit
more into that because, youknow, obviously, analytics,
analyzing, being able to analyzethis data that comes out, what
are the other skills thatthey're gonna need to have and
and that, you know, accountingand finance leaders are gonna
(23:38):
need to say, hey, guys. We needto get these skills. We need to
make time for this.
Court Watson (23:44):
If I think about skills that that a typical
accountant doesn't have today,there aren't too many that I
think an accountant can't becapable of. There needs to be a
better understanding of datamodels and data structure and
data tables, sort of thatinformation science component of
(24:04):
it. Yeah. Because theaccountant, if you look at the
applications that are comingout, they're going to be more
self-service. But thatself-service requires some of
that requisite knowledge of,like, how do I want to structure
this to get the most out of it?
If I want to build something ina visualization layer, I'm not
in the underlying table going tocreate a data structure that's
(24:26):
already a table output like theface financials. Think of let's
just go back to a basic, like apivot table. I think everyone
knows what a pivot table is.That's already a final report
structure. That is not a datatable structure, and a lot of
accountants think in that finalreport format.
Well, you need to prepare thedata in the data table structure
so that you can then leverageyour other systems and solutions
(24:48):
and capabilities to get that.And that's a that is actually a
big gap in accountants becausethey're always looking at it
from a balance sheet incomestatement format, which is a
final report structure. So beingable to sort of think about the
data component is going to be abig piece. But accountants,
again, have historically dealtwith a ton of data and have
worked with it to preparereports and done it in a way
(25:11):
where, like, there's controlsaround it. They can validate the
completeness and accuracy.
So they have the skill set, andoftentimes, if you can if you
can find the right analogy, itwill click in an account's eyes.
They will always say, like, whenthe new revenue standard came
out, there was a big concernabout the material right. And I
was like, well, the materialright isn't much different than
(25:33):
derivatives in derivativeaccounting. Right?
It's an option in the future toprovide, like, some sort of
benefit to a customer. And thenthey will blank on, like, no. I
don't know derivatives. I'mscared of those. I'm like, hey.
Have you ever, like, had a buddybought, like, a ticket to a
sporting event or a concert andinvited you and said you don't
have to pay for the ticket tillthe day of? That's that's a
(25:55):
derivative. You basicallyentered into an agreement. You
paid nothing up front, andyou're gonna get to go in, you
know, you're gonna pay for it atthe end. I'm, like, it's it's
not much different.
It's just, you know, accountantsjust need that understanding,
and then they, like, there's alot of intellectual capital in
in accounting. Lot of smartpeople. I'm confident that the
people who remain curious andcontinually learn will be
(26:18):
successful in this space.
Adam Larson (26:19):
Yeah. I I agree a 100%. It it's it's gonna take us
being curious to really, get thegenerative AI and these tools,
these new tools, and get themout of their infancy into
something that we can maturelyuse within our teams. You know?
Because you think about thingslike data analytics, you think
about things like, processingautomation, and you think about
(26:42):
understanding those things in abetter way.
We have to be curious and kindof look into that. Do you think
that we need to all accountsshould be looking and dabbling
in a little bit of data scienceto understand those things?
Court Watson (26:52):
Yes. 100%. And try and use it in day to day life.
You know, spend time buildingyour own ledger for your
personal finances. So I'm gonnadownload my credit card
statement, download my bankingstatement, and think about,
okay, if I was trying to putthis into, you know, some of
there's a lot of tools that arefree from a from a personal use
(27:13):
perspective.
That'll teach you those lessons.You know, I an example I did to
help understand some of thevisualization capabilities of
some of the solutions, I usedthem to track my me and my
college teammates' golf tripwhen we went to Ireland. And I
set it up so in my app, I couldjust enter the scores on my
phone, and it would update andproduce a graph that tracked,
(27:37):
like, who was leading. And wehad a complex point system. We
had to incorporate handicaps,had to calculate handicaps on
the fly, but it really taught mesort of the end to end
everything I need in a way thatI can then take back.
Okay. If I'm going to directlyaccess my ledger and my
consolidation ledger, and I needto do some transformations on
it, where that would sit, justgoing out and searching the web
(28:01):
on how to do those programminglanguages that really are more
accessible than a lot ofaccountants think. They're not
that hard once you spend alittle bit of time with them
because accountants are logical,and those languages are logical.
Now, you're not going to bedoing really complex things or
be the most advanced, but you'regonna do enough to be able to
add, like, column a to column bor filter out only your balance
(28:23):
sheet items in a dataset. Like,you can learn it.
You just have to be curious. Youhave to spend time on your own
trying to figure it out. Andthose are the people who will be
the most successful. And back toyour point of, like, what is the
risk to, like, the nextgeneration, not having that
curiosity and teaching thoseskills on your own is what's
going to be the risk to eachindividual.
Adam Larson (28:45):
Yeah. And for those who've been in the industry for
a long time, yeah, it might be aa little bit of a bigger curve,
but it's worth your time. Andit's also worth it for leaders
to say, hey, guys. I want you tospend the time here.
Court Watson (28:58):
Yes. Yes. It totally. And I think, you know,
one of the things that thinkprobably needs to be factored
into performance evaluationsMhmm. Is formalized and
structured learning.
So are you spending the timeeach year to develop your
(29:19):
skills? And part of that is notonly saying, hey, you know, you
need to have 40 hours oflearning a year, with the
company providing the rightlearning lead resources. And
that's not just, like, coursesonline that are video based. It
is, hey. We're gonna have asandbox, and we're gonna have
(29:40):
case studies.
I think case studies are a greatway. Right? Because an
individual has to follow sort ofinstructions, use the
technologies themselves, andcome to the solution. And maybe
that's just how I learn, but Ilearn best by doing. Some of the
vendors have really goodexamples and really good
documentation.
(30:01):
And if you say, I wanna learnhow to bring 2 datasets
together, you can go to thereference documentation, hold
they have example files that youcan use, and learn to do it. And
then you apply it to your realscenario.
Adam Larson (30:15):
Yeah. I and yeah. Hands on learning is probably
one of the best ways to do itbecause you make live mistakes
and you learn as you're going,and it's the best way to kinda
get you going. Now, you know,everybody learns in different
ways, but I think it's a it is areally effective way. And, you
know, you know, I think and italso goes beyond because a lot
of people have in the accountingspace have certifications.
(30:36):
You know, IMA offerscertification. But those courses
are geared toward one thing.They may not get you to the
other places you wanna go, soyou have to make sure you look
beyond that as well.
Court Watson (30:46):
Yes. And a lot of it is more theory based. You
know, for a lot of licensing,you need to demonstrate a
prerequisite amount of workingtime. That totally makes sense.
But to have your hands in atechnology being able to do
something, And what's moreinteresting when you can do
something that's personal toyou, because then you're like,
I'm not only learning, but I'mit's something I'm interested
(31:08):
in.
That that's the best way. Butthat doesn't always happen. The
other challenge is companieswill deploy a solution at a
production level. And if theywant a sandbox, that's an
incremental cost. So they haveto evaluate, do we want
something like a sandbox thatgives our people the ability to
go in and learn and kind of liketheir scratch pad or or
(31:28):
somewhere to play with, knowingthat it's gonna have an
incremental cost.
And I totally get theirconstraints, and you think about
that. But I do think if it'ssomething that company believes
in, we wanna leverage and wewanna get the most out of, it's
worth that investment because,ultimately, you're gonna get
more value out of thatproduction solution.
Adam Larson (31:44):
Definitely. Well, Court, I think this has been a
great conversation. We'vecovered a lot of things, and I
hope our audience enjoyed it asmuch as I have. And I just wanna
thank you again for coming on.
Court Watson (31:53):
Yeah. Thank you for having me. Have fun.
Announcer (31:56):
This has been Count Me In, IMA's podcast providing
you with the latest perspectivesof thought leaders from the
accounting and financeprofession. If you like what you
heard and you'd like to becounted in for more relevant
accounting and financeeducation, visit IMA's website
at www.ima net.org.
Welcome back to Count Me In. I'm your host, Adam
Larson. And today, we're excitedto have Court Watson, a partner
in Deloitte's Advisory'scontrollership services
practice. In this episode, weexplore the evolving world of
accounting and how modernaccountants need to adapt to
technological advancements.We'll discuss the transition
from traditional reporting todata driven methods, the
(00:41):
importance of continuouslearning, and the impactful role
of AI.
Core provides insights onbridging the gap between
accountants and IT, leveragingAI and future proofing
accounting practices. Join usfor real world examples and
practical advice to help youstay ahead in the fast paced
accounting landscape. But beforewe get started, we wanted to
give a quick shout out toDeloitte's Resilient Controller
(01:03):
podcast. It's a fantasticresource for controllers and
financial executives. You canfind it in your favorite podcast
platform, but we'll also includea link in the show notes.
So, Court, just really excitedto have you on the podcast
today, and we're gonna becovering a lot of different
things around the controllershipand AI and how it's all it's all
(01:25):
being reshaped by everythingthat's happening within the
industry. And I figured we couldstart off by just talking a
little bit about AI and itseffect on the controllership
within organizations because,accounting and finance teams are
having to learn a lot of newskills now with all these new
tools coming up.
Court Watson (01:40):
Yeah. I mean, I think if we look at generative
AI, we've sort of changed thedefinite definition of AI to
include many things thathistorically weren't machine
learning. And we're sort of allbucketing them into that that
category. Generative AI is stillin its infancy. And I think,
really, the impact on accountingorganizations today is more of
(02:03):
an inward or internal pressureto adopt AI.
And controllership, I think, isa unique space, specifically if
you're a public company, right?Because there are requirements
around how you use technologythat don't necessarily exist in
the rest of the organization,and they are a an external
requirement or regulatoryrequirement. So I need to not
(02:25):
only know that I havecompleteness and accuracy around
everything that that a systemdid, but I also need to
understand at a pretty goodlevel how the system came to the
answer it did. And if you lookat a lot of what's out there in
AI right now, it's a little bitof a black box in understanding
(02:45):
how the output was created.Right?
We're sort of relying on it todo that, and I think there's
there's many cases where, youknow, you've asked some sort of
generative AI to do somethingand came out with an answer that
that didn't really align withwhat you thought. Yeah. But that
opaqueness needs to get resolvedbefore I think you're gonna see
wide adoption of of AI in theaccounting space. Now I still
(03:09):
think right now it's a reallygood sort of overeager assistant
who can give you a first pass,but the accountant still needs
to know on what what it did tohave the domain or subject
matter expertise to be able tosay, hey. That that makes sense,
or it doesn't.
(03:29):
Here's why. And so there's 2components. The accountant is
going to gradually have toupscale in that sort of the AI
domain and, I think, understandmore of an AI model uses linear
regression to produce a expectedoutcome on, let's say, an
accrual or a managementestimate. The accountant needs
to understand the metrics thatcome out of a linear regression
(03:54):
model and whether or not thatregression model is a usable
model or is not reliable. Andthat's a kind of a new skill set
for a lot of of accountants, sothere's going to be that
upskilling over time as it comesto AI before you I think you get
widespread.
Adam Larson (04:08):
Yeah. Well, and like you said, it's still in its
infancy. And when something isin infancy, everybody sees a
little infant baby. They'relike, oh, how cute. But then a
few years down the road, whenit's a crazy toddler making a
ruckus throughout therestaurant, nobody's saying how
cute.
They're like, oh, this is such anuisance. So it'll be
interesting to see whattrajectory this new generative
AI goes because and when it's inits emphasis, everybody's giving
(04:29):
each other high fives. And Ithink what people aren't really
talking about is some of thechallenges that it's bringing.
But, like, you mentioned the onething about controllers, you
know, having to know whysomething was, and you don't
always see that with the AI. Arethere other challenges that are
being brought up as their peopleare getting into it?
Court Watson (04:44):
Yeah. I mean, the skill gap is certainly there. In
order to run and maintain anysort of sort of technology, you
need to have the requisiteknowledge to make sure that it's
it's updating. Finance data isstill a big gap for a lot of
companies. Mhmm.
(05:04):
You're often gettingtransactions that are incorrect,
that are incomplete, that aren'tall sitting in one location,
that aren't all brought togetherfrom various source systems,
various inbound interfaces inone place that an AI model can
sort of access and tell anarrative or tell a story from.
Now as, you know, as AI getsmore robust, it's going to be
(05:28):
able to ping all the differentsystems and bring it all
together. And that's truly, Ithink, what eventually the
strength of these AI models willbe. But today, we don't we don't
have that. How do you quantifyrisk?
Risk and material misstatement,how do you put a number to that?
Like, yes, it might not savetime, but it might produce a
better quality outcome. How do Ijustify that in terms of ROI? So
that that's still being figuredout is what the ROI those use
(05:50):
cases are.
Adam Larson (05:52):
Well and you also need still need the human
interaction like you mentionedbecause we all know about the
hallucinations. We've all seenthe examples of them. We've seen
the examples of the lawyer whocreated this whole speech out of
it. Not that we're talking aboutlawyers, but there's lots of
popular examples that are outthere, and you still need that
human aspect because, like yousaid, it's a really eager
assistant giving a first draft.
Court Watson (06:14):
Correct. Yes. And I I will go every 3 to 5 months
into one of the language modelsthat I have access to, and I
will ask it the same accountingquestion each time. And I always
I always pick foreign currencybecause that's a topic that is
that is pretty complex but isvery rules based. Like, if you
know the rules, you can do thecalculations.
(06:37):
And the generative AI toolssound really, really smart, but
they're also really, reallyincorrect. And that's always my
fear is someone is relying onit, but doesn't understand the
subject matter, uses that, andthen produces something that
results in a incorrect orinaccurate filing related to
(06:59):
financial statements. And Ithink that I think, at least the
conversations I have, everycontrollership at every company
is skeptical enough that they'renot at that point. So I don't I
don't think it's a pervasiverisk, but I think it could
happen.
Adam Larson (07:13):
And should we be worried about, you know, the
next generation of accountantsthat are coming up? Because, you
know, my oldest is is is afreshman in college, and, you
know, I know she's using a lotof generative AI to kind of help
give first drafts or or getideas for things and stuff like
that. And a lot of people aredoing in their works. But if
people are utilizing those andnot necessarily understanding
the having a good foundation,will will we have a gap in a few
(07:37):
years when those this nextgeneration is coming up and
doing those low level jobs in inaccounting firms?
Court Watson (07:43):
So I think for future generations, I think
there is going to be a greaterneed for sort of logic,
philosophy, ethics. And thereason I say that is I think
you're going to see generativeAI replace a lot of the menial
tasks that are day to day, andthe individual is going to spend
(08:07):
less time preparing and moretime analyzing. And that
analysis requires sort of takinga step back, thinking about sort
of the macro picture and tellingthat story, the way the
regulatory environment is, it'snever going to be there's not
people, right? You always needyou can't have generative AI
sign off on a 3 zero twocertification for a financial
(08:29):
statement. There's always goingto need to be individuals.
And if you look like companiestoday, they don't close the
books in a day. They close it in5 or 6. And if you look at those
5 or 6, right now, there's a lotof manual work. I always say
manual journal entries are adefect of your technology.
Right?
(08:49):
It's just whether or not thecost benefit of resolving that
defect is there. If it'ssomething that takes 5 minutes
to prepare and book, are youreally going to industrialize
the solution? Because it couldtake 6 months and, like, a full
software development lifestyle.It's like, well, no, I'm not.
I'm going to keep doing themanual entry.
Adam Larson (09:05):
Yeah. And you
Court Watson (09:05):
sort of pick off your large ones and fix it. And
those are all processes in theclose. You're like, how do we
fix? If I look at, like,accounting technology, I
wouldn't be surprised, and Idon't think it'll happen soon,
but if someone sort of thinks ofa ground up accounting
technology solution around theERP, Like, could there be a
(09:25):
world where the tax provision,instead of booked on day 6 of
close after day 4 where you'vefully closed the books and
consolidated, is being booked onthe fly with every journal
entry? Because leveraging AI orleveraging technology, it's able
to sort of think about, okay,what goes into the provision?
How do these transactionstypically impact our results? Is
(09:48):
the technology able toconsolidate on the fly versus
now where you have to press abutton at the end of the period
and bring it all together? Likeall those things, I think it'd
be really cool. And I would loveto be part of something where
like blank slate, ground up,ERP, subledgers, closed solution
from scratch, embedding currenttechnology. I think you're
(10:08):
seeing the large manufacturersbring that in over time.
And and I think that's what alot of control ships waiting for
is is the proven industrialsolutions to have AI embedded in
their solutions because theyknow they're going to have
controls around that. They'regoing to have a SOC report. The
controlship organizations willbe able to say, hey, I can rely
on this because it's got it'sgot that right. And the
(10:30):
industrial solutions know whataccountants need in terms of
validating that that whathappened makes sense, and they
understand it. And so ratherthan building custom and bespoke
solutions like we're seeing inother domains in a company,
accounting is sort of like,let's see what the major players
do and how they bring it intotheir value proposition so that
we can
Adam Larson (10:51):
That makes a lot of sense. And what you describe
would be really cool, but I feellike because of how things move
in this industry, it would takea long time for adoption. You
know, a lot of times you see,you know, all these cool things
going up at rest of theorganization, but the accounting
team is still using the samesoftware they used 20 years ago.
Why do you think that is thatit's there's a long adoption
period, especially in theaccountings, accounting and
(11:13):
finance teams?
Court Watson (11:14):
So I will maintain that, like, you cannot have
agile software development as itrelates to accounting
technology. And if if you wantan agile solution, well, your
first release has to be all therequirements, which yeah. It has
to be. Right? Like Yeah.
If I build a revenue recognitionsolution and it meets half of
(11:37):
the accounting standard, well,then it's not usable. That
that's one piece of why it'sslow. The other thing is the
other piece is accounting is acost center. Right? And a, and a
business has to ration its ITbudget and where is it going to
spend, you know, where it thinksis most important and totally
(11:57):
makes sense that, you know,customer facing revenue
generating parts of the businessget that priority.
Totally get it. And then theother thing is, you know, there
is, and it's funny, I'll tellsort of an anecdotal story.
There's a gap between how anaccountant describes what they
need and how someone ininformation technology or a
(12:18):
developer builds a solution. Andone of my colleagues did a
presentation, a CPEpresentation, so you get, you
know, CPE credits for your CPAlicense. And he had accountants
write out the requirements formaking a peanut butter and jelly
sandwich.
He took one and he brought aloaf of bread, peanut butter,
(12:40):
jelly, a knife, and acted outthe requirements. And he ended
up smearing peanut butter on topof the bag of the bread because
there was never a requirement toremove the bread from the bag
and then just jamming the jar ofjelly on top because there was
never a requirement to open thejelly and use a knife to spread
it on the bread. So it's like inan accountant's head, it's like,
(13:03):
well, I need to, you know, Ineed to book a bad debt
estimate. And, okay, you mightin your you might verbalize 4
steps. Well, someone who has noidea how that comes about is now
in charge of developing it, andthey're not gonna come up with a
solution that meets theaccountant's needs.
And that just creates moremanual processes on top. And so
(13:27):
finding that balance, it'sreally hard to find a unicorn
who knows accounting, knowsaccounting standards, but also
knows how detailed requirementswork, how user stories work, how
testing works. Yeah. Andbringing those together is hard
and often leads to unsuccessfuloutcomes that actually, you
know, kind of make kind of makecontrollerships the
organizations go a bitbackwards. I've seen instances
(13:49):
where the close cycle haslengthened because of new
technology, and a lot of it isbecause companies did not
rethink the processes they haveand whether or not they could
standardize, and then the gapbetween what the requirements
are for the customer, theaccounting organization,
relationship organization, andthe team that implemented the
solution.
Adam Larson (14:07):
Well, and I think that really points how important
it is to have good internalcontrols, good to have good
processes. And if you're gonnaimplement any new technology,
you should look at all yourcurrent processes to see how
could you improve upon them whenyou're when you're looking at
those things.
Court Watson (14:20):
Yes. And and that ties back to control ship
organizations are alreadyreally, really lean. Yeah. How
do I spend meaningful time onthat when I barely have enough
headcount to do the day to day?So I'm trying to do the day to
day, but also trying to do thisextra incremental thing to spend
meaningful time revisitingprocesses, redesigning
(14:40):
processes, redesigning controls,and producing requirements.
Like, it's a massive challenge.I
have seen some companies go aroute that I thought was really
interesting where, you know,they baked into the budget of
the the initiative or thetransformation effort, co
sourcing. So they would bring ina third party to backfill their
accountants so their accountantscould focus on that. And that,
(15:01):
to me, makes a ton of sensebecause then you're leveraging
the people with the knowledgeand the expertise, and they're
focused on it. And sometimeswhat you see is your existing
processes that you're someoneelse is coming in to do, they'll
see some things and bring in adifferent perspective that
potentially makes it better aswell.
Adam Larson (15:18):
I sometimes hate, but I also love when somebody
new comes in and takes an SOPthat was written. And then you
you with by all based on thequestions they give you, you
realize how badly written yourSOP was. And I think it's a good
practice for all organizationsto, at some point, bake in, hey.
We need to look at our practicesand make sure we're doing things
efficiently because otherwise,you'll continue on and maybe
(15:40):
never grow Correct. Within yourwithin your team.
Court Watson (15:43):
Yeah. And, you know, that's not only logically
makes sense, but human nature islike, this is mine. I'm gonna be
defensive if you say it's bad.Right?
Adam Larson (15:51):
Exactly.
Court Watson (15:52):
We can we can make fun of our own family, but if
you're on the outside, don'tmake fun of my family. That
that's totally different. Andand it's not much difference. If
you have any pride when you do,you're like, who is this person
telling me I'm doing it wrong orshould do it differently? But
also like from a controlsframework, that's good
governance.
Like periodically rotatingpeople into different roles
(16:12):
helps you detect things that maybe amiss. And that's one of the
recommendations of a solidcontrols framework. So it's sort
of dual hatting in its purpose.
Adam Larson (16:20):
It really is. Now, do you think that as some of
these bigger software companiesthat are kinda old hats that
have been around for a while asthey incorporate some of this AI
features into their softwaresand people start to get into it,
do you think we'll start to seean ROI within those systems, or
is it just, hey, this is just anew cool, tool, but I'm not
really gonna use it very often?
Court Watson (16:39):
I'm gonna give an accounting answer and say it
depends. Like like like, I thinkit depends on how well
integrated it is, how clearlyexplained it is to the customer,
and then how willing thecustomer is to change. And I
think with a lot of companies,they do a disservice by not
being transparent of what thelimitations are or what
(17:02):
companies need to have in placefor this to work. And I think
when you do, especially withaccountants, it resonates much
better. Like, hey, I understandit's not gonna do it all, but I
understand what I need to do toget to that point.
The other thing is AI is stillquite expensive. Mhmm. And
you've seen situations where alicense will double in cost for
(17:25):
an AI functionality that is kindof an administrative assistant.
It's like, well, do I reallywanna pay double for that,
versus just like an incrementalvalue? So that pricing, I think,
needs to get figured out.
But, you know, there are a lotof industrial solutions, both in
your core finance technologyplus what I call edge systems,
so your reporting layer, yourclosed like closed tasks, closed
(17:48):
tools. And they're starting tolook at it a lot, and they're
starting to try and incorporatethings. And I think it's going
to be in incremental stepsbecause I think they're learning
along the way too of what workswith the customer, where the
feedback is, where the frictionpoints are, where clients, where
customers, you know, incontrollership are like, Hey, we
can't use this because we don'tunderstand what's happening, and
we can't get past regulatoryquestions around that. So it's,
(18:14):
it'll happen think it's gonnahappen in graduated steps.
Adam Larson (18:18):
Yeah. In graduated steps. That makes sense. Do you
think there's steps that we canbe taking now to make sure that
the, we can get over thatcommunication gap? Because
you're mentioning that we'rewe're talking about how, like,
it's hard to have somebody whounderstands all the steps and
accounting and understands howto write the stories and the
processes for for things to workproperly and and to make sure we
get all those steps down,especially for technology work
(18:39):
properly.
Are there are there ideas of ofor or steps people can take to
kinda get over that gap? Becausethat's probably the hardest part
of getting over the thecommunication gap that's there.
Court Watson (18:52):
Yeah. So there's a very famous technology
evangelist, and he recentlyMhmm. Posted and said, hey. You
know, artificial intelligence isthe gas, and regulations are the
brake.
Adam Larson (19:05):
Mhmm.
Court Watson (19:06):
And my first thought was, well, the better
the brake, the faster your topspeed can be. Right? Okay. If
you're on a bicycle and you haveno brake, are you going to go 40
miles per hour? Are you going togo 5?
Because you can't stop. Right?Whereas if you could, you can
stop and control. You're gonna,you're gonna be able to go
faster. And I think definingwhat that break is, what the
(19:28):
governance around the model is,what you actually need to see
from a model to get comfortablewith it.
I don't think anybody spentmeaningful time other than
saying this model doesn't giveme enough that I can rely on it.
It should be like, what do youneed to rely on it? I need
completeness. I need accuracy. Ineed to understand the inputs.
(19:48):
I need to understand thetransformation of the inputs and
how it derived it because at theend of the day, I need to be
able to back test it. I need tobe able to go back and look and
say, did it work? Did it do whatit was intended to do? So
spending time building out thatframework and what that looks
like, and for right now, it'sgoing to be company by company
basis because they are all goingto have different measures of
risk. They're all going to havedifferent materiality levels,
(20:10):
and they're going to have towork on that.
The other thing is cleaning yourfinancial data, going back,
where do we have gaps? And a lotof companies are doing this as
they have to and are required tomodernize their finance
technology because we're kind ofin a cycle of, like, the updates
are occurring, but going throughthat, really thinking with a
future sort of focus of, like,yeah, I don't use this dimension
(20:31):
today consistently, but I havean opportunity here to clean
that populate because I knowit's gonna have value in the
future. So those are the 2things that companies can do now
for sure.
Adam Larson (20:40):
Definitely. Do you think this is gonna have a
greater impact on small tomedium sized businesses?
Because, obviously, the largeorganizations might have the
capital to put the time in. Butsmall organizations, you know,
you've already mentioned thecontrollership is very lean. A
lot of finance and accountingteams have had to cut down and
and make it more lean and do thesame amount of work.
How are we gonna get these extraskills and upskill and be ready
for this in the midst of thisfast changing market?
Court Watson (21:04):
It can go both ways on that question. Okay. So
it may be easier for small tomedium businesses because their
systems are going to berelatively cleaner and simpler.
Adam Larson (21:17):
Okay.
Court Watson (21:19):
At the same time, AI is expensive. And for it to
be fully built into software andproducts, I think, you know, the
companies that produce financetechnology with the capital and
resources to embed AI
typically don't target the smallto medium sized business. Right?
(21:39):
That they're gonna go up, youknow, they are the the the
vendors for the Fortune 500companies. So there's going to
be that cost component.
But I can see where thedeployment will be much easier,
you know, at a small businessthat uses one sort of web based
ledger to track expenses andpayroll and everything and
(22:03):
really doesn't have a complexstructure of a multinational
that, over several acquisitions,uses 4 different ERPs and 7
different customer billingsystem. That just doesn't happen
in a small medium business, sothe opportunity to get the most,
I think, will be an easier lift.It's just in affordability.
Adam Larson (22:21):
Yeah. That makes a lot of sense. Well and and you
can go either way to from thefrom the the software company's
perspective because why not testit out with a bunch of small
companies, and then you can rollit out the big ones, but then
you're not doing anythingcomplicated. So it's a a give
and take depending on whatyou're doing. Yeah.
Court Watson (22:36):
And, you know, a lot a lot of what you do is is
proof of concepts where you picka component of a large business
and test it. I think that wouldbe a great idea to find that mid
market client that's willing toto be involved in those pilots
and provide feedback. That Ithink that's the best way you're
going to build a solution thatworks for everyone. It's just a
totally different beast when youget a multinational that has
(22:58):
been around forever. They haveprocesses they've had forever
that don't always alignnationally.
Again, systems are different allover the globe. It's just is a
very large effort to standardizeand align across multiple
geographical jurisdictions.
Adam Larson (23:17):
Yeah. So we've talked a little bit about
upskilling and this new skillsthat finance and accounting
teams are gonna need to have.Maybe we can dig a little bit
more into that because, youknow, obviously, analytics,
analyzing, being able to analyzethis data that comes out, what
are the other skills thatthey're gonna need to have and
and that, you know, accountingand finance leaders are gonna
(23:38):
need to say, hey, guys. We needto get these skills. We need to
make time for this.
Court Watson (23:44):
If I think about skills that that a typical
accountant doesn't have today,there aren't too many that I
think an accountant can't becapable of. There needs to be a
better understanding of datamodels and data structure and
data tables, sort of thatinformation science component of
(24:04):
it. Yeah. Because theaccountant, if you look at the
applications that are comingout, they're going to be more
self-service. But thatself-service requires some of
that requisite knowledge of,like, how do I want to structure
this to get the most out of it?
If I want to build something ina visualization layer, I'm not
in the underlying table going tocreate a data structure that's
(24:26):
already a table output like theface financials. Think of let's
just go back to a basic, like apivot table. I think everyone
knows what a pivot table is.That's already a final report
structure. That is not a datatable structure, and a lot of
accountants think in that finalreport format.
Well, you need to prepare thedata in the data table structure
so that you can then leverageyour other systems and solutions
(24:48):
and capabilities to get that.And that's a that is actually a
big gap in accountants becausethey're always looking at it
from a balance sheet incomestatement format, which is a
final report structure. So beingable to sort of think about the
data component is going to be abig piece. But accountants,
again, have historically dealtwith a ton of data and have
worked with it to preparereports and done it in a way
(25:11):
where, like, there's controlsaround it. They can validate the
completeness and accuracy.
So they have the skill set, andoftentimes, if you can if you
can find the right analogy, itwill click in an account's eyes.
They will always say, like, whenthe new revenue standard came
out, there was a big concernabout the material right. And I
was like, well, the materialright isn't much different than
(25:33):
derivatives in derivativeaccounting. Right?
It's an option in the future toprovide, like, some sort of
benefit to a customer. And thenthey will blank on, like, no. I
don't know derivatives. I'mscared of those. I'm like, hey.
Have you ever, like, had a buddybought, like, a ticket to a
sporting event or a concert andinvited you and said you don't
have to pay for the ticket tillthe day of? That's that's a
(25:55):
derivative. You basicallyentered into an agreement. You
paid nothing up front, andyou're gonna get to go in, you
know, you're gonna pay for it atthe end. I'm, like, it's it's
not much different.
It's just, you know, accountantsjust need that understanding,
and then they, like, there's alot of intellectual capital in
in accounting. Lot of smartpeople. I'm confident that the
people who remain curious andcontinually learn will be
(26:18):
successful in this space.
Adam Larson (26:19):
Yeah. I I agree a 100%. It it's it's gonna take us
being curious to really, get thegenerative AI and these tools,
these new tools, and get themout of their infancy into
something that we can maturelyuse within our teams. You know?
Because you think about thingslike data analytics, you think
about things like, processingautomation, and you think about
(26:42):
understanding those things in abetter way.
We have to be curious and kindof look into that. Do you think
that we need to all accountsshould be looking and dabbling
in a little bit of data scienceto understand those things?
Court Watson (26:52):
Yes. 100%. And try and use it in day to day life.
You know, spend time buildingyour own ledger for your
personal finances. So I'm gonnadownload my credit card
statement, download my bankingstatement, and think about,
okay, if I was trying to putthis into, you know, some of
there's a lot of tools that arefree from a from a personal use
(27:13):
perspective.
That'll teach you those lessons.You know, I an example I did to
help understand some of thevisualization capabilities of
some of the solutions, I usedthem to track my me and my
college teammates' golf tripwhen we went to Ireland. And I
set it up so in my app, I couldjust enter the scores on my
phone, and it would update andproduce a graph that tracked,
(27:37):
like, who was leading. And wehad a complex point system. We
had to incorporate handicaps,had to calculate handicaps on
the fly, but it really taught mesort of the end to end
everything I need in a way thatI can then take back.
Okay. If I'm going to directlyaccess my ledger and my
consolidation ledger, and I needto do some transformations on
it, where that would sit, justgoing out and searching the web
(28:01):
on how to do those programminglanguages that really are more
accessible than a lot ofaccountants think. They're not
that hard once you spend alittle bit of time with them
because accountants are logical,and those languages are logical.
Now, you're not going to bedoing really complex things or
be the most advanced, but you'regonna do enough to be able to
add, like, column a to column bor filter out only your balance
(28:23):
sheet items in a dataset. Like,you can learn it.
You just have to be curious. Youhave to spend time on your own
trying to figure it out. Andthose are the people who will be
the most successful. And back toyour point of, like, what is the
risk to, like, the nextgeneration, not having that
curiosity and teaching thoseskills on your own is what's
going to be the risk to eachindividual.
Adam Larson (28:45):
Yeah. And for those who've been in the industry for
a long time, yeah, it might be aa little bit of a bigger curve,
but it's worth your time. Andit's also worth it for leaders
to say, hey, guys. I want you tospend the time here.
Court Watson (28:58):
Yes. Yes. It totally. And I think, you know,
one of the things that thinkprobably needs to be factored
into performance evaluationsMhmm. Is formalized and
structured learning.
So are you spending the timeeach year to develop your
(29:19):
skills? And part of that is notonly saying, hey, you know, you
need to have 40 hours oflearning a year, with the
company providing the rightlearning lead resources. And
that's not just, like, coursesonline that are video based. It
is, hey. We're gonna have asandbox, and we're gonna have
(29:40):
case studies.
I think case studies are a greatway. Right? Because an
individual has to follow sort ofinstructions, use the
technologies themselves, andcome to the solution. And maybe
that's just how I learn, but Ilearn best by doing. Some of the
vendors have really goodexamples and really good
documentation.
(30:01):
And if you say, I wanna learnhow to bring 2 datasets
together, you can go to thereference documentation, hold
they have example files that youcan use, and learn to do it. And
then you apply it to your realscenario.
Adam Larson (30:15):
Yeah. I and yeah. Hands on learning is probably
one of the best ways to do itbecause you make live mistakes
and you learn as you're going,and it's the best way to kinda
get you going. Now, you know,everybody learns in different
ways, but I think it's a it is areally effective way. And, you
know, you know, I think and italso goes beyond because a lot
of people have in the accountingspace have certifications.
(30:36):
You know, IMA offerscertification. But those courses
are geared toward one thing.They may not get you to the
other places you wanna go, soyou have to make sure you look
beyond that as well.
Court Watson (30:46):
Yes. And a lot of it is more theory based. You
know, for a lot of licensing,you need to demonstrate a
prerequisite amount of workingtime. That totally makes sense.
But to have your hands in atechnology being able to do
something, And what's moreinteresting when you can do
something that's personal toyou, because then you're like,
I'm not only learning, but I'mit's something I'm interested
(31:08):
in.
That that's the best way. Butthat doesn't always happen. The
other challenge is companieswill deploy a solution at a
production level. And if theywant a sandbox, that's an
incremental cost. So they haveto evaluate, do we want
something like a sandbox thatgives our people the ability to
go in and learn and kind of liketheir scratch pad or or
(31:28):
somewhere to play with, knowingthat it's gonna have an
incremental cost.
And I totally get theirconstraints, and you think about
that. But I do think if it'ssomething that company believes
in, we wanna leverage and wewanna get the most out of, it's
worth that investment because,ultimately, you're gonna get
more value out of thatproduction solution.
Adam Larson (31:44):
Definitely. Well, Court, I think this has been a
great conversation. We'vecovered a lot of things, and I
hope our audience enjoyed it asmuch as I have. And I just wanna
thank you again for coming on.
Court Watson (31:53):
Yeah. Thank you for having me. Have fun.
Announcer (31:56):
This has been Count Me In, IMA's podcast providing
you with the latest perspectivesof thought leaders from the
accounting and financeprofession. If you like what you
heard and you'd like to becounted in for more relevant
accounting and financeeducation, visit IMA's website
at www.ima net.org.
Popular Podcasts
24/7 News: The Latest
The latest news in 4 minutes updated every hour, every day.
Therapy Gecko
An unlicensed lizard psychologist travels the universe talking to strangers about absolutely nothing. TO CALL THE GECKO: follow me on https://www.twitch.tv/lyleforever to get a notification for when I am taking calls. I am usually live Mondays, Wednesdays, and Fridays but lately a lot of other times too. I am a gecko.
The Joe Rogan Experience
The official podcast of comedian Joe Rogan.