Episode Transcript
Transcripts are displayed as originally observed. Some content, including advertisements may have changed.
Use Ctrl + F to search
0:01
The power of data is undeniable
0:03
and unharnessed, it's nothing but
0:05
chaos. The amount of data,
0:07
it was crazy. Can I trust it? You
0:10
will waste money, help the other way duct tape
0:12
to defy you. This season we're solving
0:14
problems in real time to reveal the
0:16
art of the possible. Making data your
0:18
ally, using it to lead with confidence
0:21
and clarity. Helping communities
0:23
and people thrive. This
0:25
is Data Driven Leadership, a show by
0:28
resultant. Hey
0:31
guys, welcome back. It's Jess Carter with Data
0:33
Driven Leadership and on today's episode, we're going
0:35
to be talking to Josh Richardson, who's the
0:37
Chief of Staff of the Department of Workforce
0:40
Development in Indiana. Side
0:42
note, if you ever listen to the
0:44
news and the beginning of each month, you might hear
0:46
them talk about the wage report, the wage data coming
0:48
out. That is federally
0:50
mandated as a collection of data from
0:52
the states to the feds. That
0:55
is kind of usually managed by the departments
0:57
of workforce development in your state. Josh
1:00
works in the department and he has spent
1:02
over a decade thinking deeply about the
1:05
data that they have and the unique data
1:07
that they hold and how we could leverage
1:09
that to better serve citizens. I
1:12
think this is an opportunity for anyone in leadership
1:15
to think about what data they
1:18
have in their organization that
1:20
is an asset that's unique to their own
1:22
organization and think about how
1:24
they can better leverage that creatively to
1:26
serve a purpose or to help meet
1:28
a need either in the organization or
1:31
with your constituents or clients, customers
1:33
even. As we get into this,
1:36
you'll kind of hear more about how Josh unpacks his
1:39
idea and you'll get to hear
1:41
it from idea and
1:43
vision to fulfill it.
1:46
I think the other thing that's neat
1:48
too that Josh highlights is there are a
1:50
lot of people around him that helped
1:53
bring this idea through fruition.
1:55
It's also looking at the
1:58
greater picture of what you're trying to accomplish. and recognizing
2:00
who's around you that you need to be for you
2:03
and for this vision and how do you pull them in
2:05
and help them get excited about it. I hope you enjoy
2:07
this episode. Let's get into it. Welcome
2:13
back to Data Driven Leadership. I'm your host, Jess
2:15
Carter. And today we have Josh Richardson, the Chief
2:17
of Staff at the Indiana Department of
2:20
Workforce Development. Let's get into it. Welcome, Josh.
2:22
I'm so happy to be here. Thank you, Jess. Let's
2:24
get into it. So Josh is maybe
2:26
the first person on the show that
2:29
I have had the pleasure of knowing
2:31
for nearly 10 years. Right.
2:35
It's been a minute. I remember you walking into
2:37
a room with the Commissioner, with Scott Sanders at
2:39
the time, to talk to the PMO about the
2:41
unemployment insurance system that was going live. And that
2:44
was the first time I got to meet you.
2:46
I remember the exact moment when I got to
2:48
meet you. I'm sure I was as memorable as you.
2:50
No, I do remember it. I would have had children
2:52
remembering the rest of it. But so here you are
2:54
promoted all the way up to podcast host. And I'm
2:56
still doing the same thing. It's
2:59
great to be with you, Jess. And yeah, lots of good stuff over
3:01
there. Sure, Chief of Staff.
3:03
You've also been promoted to something that's a little
3:05
more important. So OK,
3:08
this is also difficult. Let's go. So
3:11
when I first joined DWD,
3:13
or was helping as a consultant, I
3:15
had never heard of the Department of Workforce Development. So
3:18
when you pitch that to someone you meet or at
3:20
a kid's sports game or something, what do you explain
3:22
about, give me the elevator pitch for what is DWD?
3:25
So it's a good question. And the
3:27
thing is that that explanation almost switches
3:29
a little bit, depending on the economy,
3:31
because we have two real main things
3:33
that we do from an overarching perspective.
3:35
One is that we administer the state's
3:37
unemployment insurance system. So that's more than
3:40
half of the employees at the agency.
3:43
But so really, that's everything from collecting
3:45
the employer taxes that go in to
3:47
collect the trust fund all the way
3:49
to paying out those benefits to ensuring
3:51
fraud. We hold appeals within the agency.
3:53
But the other side of the agency
3:55
is really workforce training. And I think
3:57
largely that is how you would do
3:59
that. connect individuals that were
4:01
unemployed to work, but it includes things
4:03
like upskilling, existing workforce, but a huge
4:05
part of it is data, right? We
4:08
collect so much information about the workforce
4:10
through both sides of that system, information
4:12
about what's available, what
4:14
skills are in demand. And so it's
4:17
kind of that nexus that joins the agents, the
4:19
two sides of the agency. Okay, and
4:21
to your point earlier about the economy, whether
4:23
it's really, really great or really, really not
4:26
great, one of those sides of your agency
4:28
is sort of ebbs and flows. It's
4:30
an interesting piece where, when we've
4:33
gone through periods like the Great Recession or certainly
4:35
during the
4:37
pandemic, really everything that we
4:39
do at the executive level, so much
4:41
of the focus is on unemployment insurance,
4:43
but now the pendulum is almost entirely
4:47
in the other direction, where so much of the focus
4:49
is on workforce, because really across our
4:51
economy, from frontline work all the way
4:53
up through the highest skilled jobs that
4:55
we have, everybody needs skilled workers in
4:57
order to run their business. And so
4:59
lots of focus there. We're still cleaning
5:02
up some of those issues, but yeah,
5:04
it is really interesting how the
5:06
conversations change at that executive level.
5:09
So few people
5:11
in my career so far have been
5:13
as passionate and innovative in their space
5:15
as I would argue that I've observed
5:18
you. I realize that you're not gonna
5:20
suit your own horn, but I will for you, that there
5:22
is a passion that just comes out of
5:24
the work. Was that always
5:26
there? How did you become so passionate about
5:28
data beauty? So I don't know because we
5:32
probably won't get into it on this podcast, but
5:34
it was kind of an accident that I'm here.
5:36
I will say that I've always really been in
5:38
love with public policy, the idea of having
5:41
big puzzles to solve that
5:43
actually lead to real outcomes for real people.
5:46
So I think that's always been there, but
5:48
it has been interesting that the
5:50
Department of Workforce Development has been a really
5:52
good mix, because so many of the things
5:54
that are my interest from economics to data
5:57
to law to politics,
6:00
All of it has ended up being able to, you
6:02
know, there are things that I get to do every
6:04
day in many of those different areas and I think
6:06
it's helped drive the passion. So
6:08
yeah, I hope so. You're right. I don't think
6:10
maybe I won't talk too much more about that.
6:12
But that's right. It's been great for me and
6:14
it's easy to find the passion to do this
6:17
work. Okay. Well, so that's fine.
6:19
We will get into sort of the interesting thing to
6:21
talk about here that I'm really excited about is, and
6:23
I want to talk about how it kind
6:25
of came to be because I have this memory, another
6:27
memory that's a really particular image of
6:30
me sitting across a lunch
6:32
table from you in
6:34
2014, 2015 and on
6:36
a napkin, you drew out
6:38
the concept for this project
6:40
that, and I'm not even sure, what are we calling
6:42
it? Yeah. Well, so
6:44
I think the workforce recommendation engine is sort
6:47
of like all encompassing tool that covers all
6:49
of it. Yeah. And then
6:51
pivot is really the name of the tool that thus
6:53
far we've launched. Okay. So that's
6:55
the interface that people are interacting with. But I
6:57
do think just as this broad concept, the workforce
6:59
recommendation engine probably covers everything.
7:02
Okay. And so you drew
7:04
this out for me in a really high level.
7:06
Can you kind of, again, elevator a
7:08
speech of what that high level was? What was the
7:10
idea? Well, so I mean,
7:13
there are a couple of different things that
7:15
all come together at once. One is that,
7:17
at the Department of Workforce Development, we are
7:19
privy to a significant amount that we mentioned
7:21
this earlier, like a significant amount of data
7:23
about the workforce. One of the biggest pieces
7:26
of that are unemployment
7:28
insurance wage records. And so just a
7:30
quick version, every quarter, every
7:32
employer in Indiana is required to
7:34
report at that individual level, the
7:36
amount of wages that they paid
7:39
to an individual that they employ. And,
7:42
you know, okay. So over the years, researchers
7:45
often ask for that data. So they
7:47
come through the Bureau of
7:49
Labor Statistics and they come to us with
7:52
a request and do all sorts of really
7:54
interesting research topics using these wage records. And
7:57
so that's one. is
8:00
on the workforce side of the
8:02
house, there's often
8:05
so many acronyms, so many
8:07
programs, so many different baskets
8:09
of funding that it gets
8:11
really difficult, even for our staff,
8:13
let alone the people that we serve to figure out
8:15
how to connect those. And so
8:18
I think as I started
8:20
to learn a little bit more about the
8:23
data we had access to and also as
8:25
technology changes and more focus on data-driven type
8:27
decision making, I think we're really looking at
8:29
this saying, I wonder if we can use
8:32
wage records, if we can use this data that
8:34
we have access to that the private sector really
8:36
can't get. Again, a lot of this is confidential
8:38
by federal law and for good reason, but
8:41
what if we could use it to help figure
8:43
out how to match up the services we have
8:45
with the people that would be most likely to
8:47
benefit from them? And it really
8:50
kind of grew from there and technology has changed
8:52
a little bit too. I
8:54
love the idea that I grew this
8:56
out on the napkin, but I think, and
8:58
I think largely, yes, that's true, the vision
9:00
has stayed the same, but what's happened is
9:02
that I think as the technology has advanced,
9:05
it both made it sort of more tangible,
9:08
but it also has allowed us to do it in
9:11
really a way that I think probably exceeded even
9:14
what I was hoping for. And so
9:16
for right now, just to get really pragmatic and make sure I
9:18
understand, because I've been on a
9:20
lot of projects adjacent to this one,
9:23
but not on this project. So if
9:25
I understand at brass tacks here, if
9:28
I'm on unemployment in Indiana, I have
9:30
the opportunity to leverage this workforce engine
9:32
to say, hey, what are the,
9:36
I hope I don't get this wrong here, the
9:38
most effective things I could be doing, skill
9:41
sets, education to gain to skill up,
9:43
essentially get re-employed. Right. So
9:45
where we're at right now is really
9:47
focused on occupational change, right? So we're
9:50
going to add more and more elements
9:52
to this as we move forward. And
9:54
obviously this is still a work in
9:56
progress. So, you know, again, think about
9:58
it this way. So. About
10:00
3.2 million Hoosiers in Indiana's
10:03
workforce. And
10:05
every quarter, those individuals are
10:07
having those wage records reported.
10:09
And so what we can see is sort
10:11
of career transitions. We can watch people as
10:14
they move through and as their jobs change
10:16
and as their industry changes, the place where
10:18
they work changes, their wages change. And
10:21
so by using that, when an individual files an
10:23
unemployment claim, so you should
10:25
also say this, there
10:27
are something like at a higher level, we
10:30
can kind of classify jobs into 830 different
10:32
occupations. So
10:34
obviously it gets more specific than that when
10:37
you drill down. But I always
10:39
think, for a lot of people,
10:41
if you gave them a pencil and paper and
10:43
said, write down as many occupations as you could
10:45
think of, I don't think they'd get that. No.
10:48
And I think we have, is you have a person that's
10:50
filing an unemployment claim that can be a really stressful, it
10:52
can be a scary time. And
10:54
their first question is, how do I successfully
10:57
make a transition back to work? And
11:00
I like the job that I had and I'm really disappointed. It's gone,
11:02
but I didn't like the job that I had and I'm ready for
11:04
something new. But trying to be in
11:06
that spot where they've got to figure out what
11:08
to do next is tough. Well, so the idea
11:10
is that what we can maybe do is leverage
11:12
these wage records, leverage this information that we have
11:15
to figure out where we've seen people
11:17
make a successful transition that were in
11:19
that same position before. So, I'm a
11:22
factory worker in Richmond, Indiana, who lost
11:24
a manufacturing job after 20 years with
11:26
a high school education. We
11:29
can look at many of those same factors
11:31
and see, okay, of the people that were
11:34
in that similar spot that made a career
11:36
change, which ones were most effective? Putting
11:38
that as an option in front of that person. Great.
11:41
So it gives, in that moment when there might be
11:43
some high emotions, because we are unemployed and we don't
11:45
want to be, it gives
11:47
somebody some sense
11:49
of a shortened, more data-based
11:51
menu of where other people
11:54
have gone in case they haven't thought of that
11:56
or consider those as options that are easily attainable
11:58
potentially, right? Is that fair? Right. Okay.
12:00
And this is, is this the only place
12:02
this is happening? I mean, I know that
12:05
workforce agencies are federally mandated. They're in each
12:07
state. Is this, are we doing this in
12:09
other states? I mean, as far as we know,
12:11
this is the only one like it. Now you'll see,
12:13
you know, as we're adding AI
12:15
to essentially everything that exists, there are tools
12:17
out there that try to use artificial intelligence
12:20
to match people with jobs. A lot of
12:22
times those are skills based and those are
12:24
exciting too. It'll be interesting to see how
12:26
they develop, where they'll do things like you
12:28
tell us about your prior jobs. Right. We'll
12:30
try to identify skills. Then we're going to
12:32
look at job postings. We're going to try
12:34
to identify the skills and match them. So
12:37
it's an interesting component. But
12:39
what's really different about ours is
12:41
that we don't actually have to ask
12:43
this person for any additional inputs. And
12:46
we're using these administrative wage records. And so
12:48
instead of someone having to sit down and
12:50
try to match these skill pieces and try
12:52
to figure out what's relevant, we can just
12:55
see what's worked for other people. And so
12:57
nobody that we're aware of has figured out
12:59
how to use these state data sources as
13:02
the driver of this kind of
13:04
information. That is so cool. This is a
13:06
tool, this is data as an asset to
13:08
the agency that has not been utilized in
13:10
an effective way like this before. And we
13:12
can start from here. That's right. And I,
13:14
like I said, I think that the other
13:16
tools are really exciting too. And with all
13:18
this AI, really excited to see where it
13:21
goes. But I think, you know, they require
13:23
a decent amount of effort from that individual.
13:25
You usually have to try to drive them
13:27
to a standalone website or something and encourage
13:29
them to participate. And sometimes it's hard for
13:31
them to sort. You know, there's so many
13:33
different tools out there. We talked about this
13:35
before. How do I know which one to
13:37
use? And so we really love the idea
13:39
that this is just
13:41
a part of the unemployment claims filing process.
13:44
I mean, essentially, they sort of can't avoid
13:46
it. They're going to see
13:48
these options. And so it's low effort from that standpoint.
13:50
But obviously, you know, they're going to
13:52
get much out of it. We want them to interact. And,
13:55
you know, big part of this is the autonomy of that
13:57
individual. But yeah,
13:59
so I I'm excited about the skills-based
14:01
stuff, but in our case, we almost skipped
14:03
that step in the line, right, instead of
14:05
having to say, like, look, we know that
14:07
there's some skills mapping. When we
14:10
see people move from one occupation, we see
14:12
these successful careers into another. There absolutely are
14:14
going to be some skills there. But with
14:16
this tool, we don't really have to define
14:18
them. We just know that we've seen them
14:20
be successful. And I think that's really cool
14:22
and I think it will be really helpful.
14:26
You mentioned this, or alluded to it earlier, the,
14:30
call it customer experience of UI,
14:32
can be tricky because it is
14:34
a bunch of different federal or state programs, depending on
14:36
where you are, and then they get enrolled in one
14:38
or the other. And it can be a little confusing
14:40
to navigate. And the fact that you're
14:43
giving them an additional service
14:46
through this unemployment experience, right?
14:48
Yeah, so the tool
14:50
itself is absolutely the additional service. I
14:53
think one of the interesting things about
14:55
these problems are, you know, almost
14:58
in every case, we can
15:00
help, but the work is going to be on
15:02
that individual at the end of the day. They're going to be the
15:04
ones who have to take this leap
15:06
towards a new occupation or, you know, in
15:08
many cases enroll in the kind of training
15:10
that will better their outcomes. So like, we
15:12
can't do that for them, right? This is
15:14
going to take their work. It always will.
15:16
But I think that the key point is
15:18
trying to reduce the level of effort that
15:20
it takes to figure out what that path
15:22
looks like and also increase their confidence because
15:24
this tool helps them better understand the outcomes
15:26
they can expect. Yeah. You know,
15:28
again, based on people like them
15:30
who have made that transition. I love
15:32
that concept of you're almost immediately helping them
15:34
envision where they might go. That
15:37
can be a real gift to people that are
15:39
in the middle of this process. It's not exactly
15:41
a confidence inspiring moment in general. And the fact
15:43
that they can immediately start to see what might
15:45
work is a real kind of gift.
15:48
Yeah. Well, and I mean, you slowed me
15:50
down a couple of times already to talk through this just because
15:52
of how many different things that are happening. I think it's really
15:54
easy to take for granted on our side. Like,
15:56
you know, I've now spent 15, 16 years in
15:59
this world. So kind of know where
16:01
these programs are at, but for someone who's just
16:03
been working, this is an unexpected job loss. They
16:05
haven't done any of this background work. And
16:07
so really being able to cut
16:09
down on the amount of effort that it
16:11
takes in this stressful time to see what
16:14
might work is, I
16:16
think, the best thing that we're doing with the
16:18
tool. Yeah. Okay.
16:21
So question for you in
16:24
this whole process from ideation to
16:26
it's live, right? Right. Yeah.
16:29
So what surprised you about the process?
16:34
You know, I think that there
16:36
are so many different surprises as we move
16:38
through this. I think, you
16:40
know, you asked the question earlier about
16:43
whether other states were doing this. I
16:45
think one of the things that's become clear is that Indiana
16:47
has really done some things to set
16:49
the stage to make this possible. And
16:51
I think that it's, you know, there have been some
16:53
times through this where I realized like, oh, wow, you
16:56
know, like we're actually in really good shape. The conditions
16:58
were right to do this here. And
17:00
I think that the truth is that I, again,
17:02
I'm not an expert at truly what every single
17:04
other state has, but between our management and performance
17:06
hub and between these data sources
17:09
and sort of the maturity of these data
17:11
sources that we have, the relationships with other state
17:13
agencies, you know, we've been able to
17:15
do some, when we look at other states,
17:17
they're just starting different components of this process.
17:19
And so that's been exciting for me to
17:22
see Indiana's real leadership in this data driven
17:24
world. And so
17:26
that's been interesting and exciting. Yeah,
17:29
I think maybe that's the biggest surprise. Okay.
17:31
Well, and when you say that, so
17:33
I'm in a non public sector translate, some of what you
17:35
just said, and you tell me if I'm right. So the,
17:37
some of the themes of what I just heard is we
17:40
did go through this unemployment insurance modernization. So the
17:42
system, we went through that in 2014. It
17:47
was a long period, 14, kind of the launch, but yes, many
17:49
years ago. So kind of
17:51
new system with new data access that we have.
17:54
And then we have the management performance hub.
17:56
And so when data sharing data, privacy protection
17:58
that as we pulled data from different
18:01
agencies, there's protection, but there's collaboration
18:03
that's possible with inter-agencies. And so
18:05
when you talk about workforce,
18:07
so the wage data,
18:10
like my data would be just
18:12
Carter resultant, blah, blah, blah, blah, every quarter.
18:15
But then you have the unemployment data, so
18:17
you know if I'm unemployed too. But you
18:19
don't have the education data. That comes from
18:21
IDOE, right? Yeah, or
18:24
the Commission for Higher Ed or the other state
18:26
partner. Yeah, so this is, it's
18:28
necessitating that kind of a collaboration across all of those
18:30
data sets, is that fair? That is correct. Okay,
18:33
and that's where MPH plays a critical
18:35
role. Absolutely. Okay, and so
18:37
it's sort of like you have like
18:39
these modern systems and this interoperability of
18:41
data that's flowing between these entities that's
18:43
protected, that is timely as much as
18:45
it can be knowing that you know
18:47
the quarterly data, etc. Some of that
18:49
seems really important to me. When I
18:51
say IDOE, I mean, Indiana Department of
18:54
Education, I have to remember to get acronyms. And
18:57
MPH Management Performance Hub, which we already talked about. Okay,
19:00
so to your point you're saying we didn't
19:02
just build this. It was
19:04
built on a foundation of the commitment
19:06
to data and technology advances we've already
19:08
been making. That's so true. So I mean, you know,
19:11
just going back to the earlier question, sketching this out
19:13
on a cocktail napkin, I think, you know,
19:15
we could say, hey, wouldn't it be fun if
19:18
we could take all of this data and use
19:20
it in a way to help positively inform people
19:22
about outcomes. But the reason
19:24
why that's possible is because of a
19:26
history that Indiana has with the
19:28
State Longitudinal Education Database, SLDS
19:31
is the acronym, and I
19:33
don't even know Jess exactly what the acronym is for.
19:36
But it's essentially where we're able to link
19:38
up a lot of these workforce and education
19:40
related records together. And then MPH allows us
19:42
to do these data matches where, you know,
19:45
you've talked a lot on this show and
19:47
your listeners here, you talk about data silos
19:49
and everyone is familiar with this. But essentially
19:51
to allow this agency to allow
19:53
the data that these different agencies possess a lot
19:56
of a very sensitive a lot of it is
19:59
confidential. it needs to be protected. But
20:02
MPH has really allowed a secure
20:04
place to do this record linkage
20:06
that allows us to make smarter decisions
20:08
based on this data in a way
20:11
that protects people's privacy and security. And
20:14
so, you know, the from,
20:16
you know, the back of a napkin to
20:19
launch in November couldn't happen without those things
20:21
being in place. Yeah. So you kind
20:23
of already explained the difference, the experience
20:25
for somebody who's on unemployment or going
20:27
through the process compared to before, that
20:30
they wouldn't have had these kinds of
20:32
recommendations or suggestions. I want to emphasize
20:34
or ask you to emphasize the importance
20:36
of agency here. We're not telling them,
20:39
these are your only options, or these are
20:41
your options. We're saying this is what the
20:43
data, we're just presenting information to them. I
20:46
think this is such a critical part about
20:48
all of this. And so I definitely want to
20:50
answer that. But I will say like before
20:52
we started around this way, our agency has
20:54
had a long history of looking
20:56
at things like what are the most in-demand
20:58
jobs in an area. So you'll look at
21:01
things like, you know, what occupations are growing,
21:03
what are the wages. And so if you
21:05
think about a predecessor to what we're doing
21:07
here, they were posters and they would
21:09
be a poster that we would hang in a local office, maybe
21:12
send to a school, they would talk about,
21:14
you know, the top 50 jobs in their
21:16
area. But though if you think about those,
21:19
they were tough in a couple of areas.
21:21
There's one there very one size fits all.
21:23
And so, you know, it wasn't surprising to
21:25
see a lot of those require things like
21:27
higher education or master's degrees. And again, those
21:31
are absolutely still going to remain hot jobs
21:33
for the future. But for a lot of
21:35
individuals, if I again, if I've just lost
21:38
this manufacturing job, the only job that I've
21:40
known for 20 years, seeing a
21:42
list that includes a lot of things that feel inaccessible
21:44
to me isn't that helpful. The other part about this
21:46
is that people of course don't want to be told,
21:49
here's what you're supposed to do, here's your only answer.
21:51
And so I think we're really trying to do with
21:53
this tool is essentially have the approach, here's where others
21:55
have been successful, we can see this. And so the
21:57
way I like to think about it is that the
21:59
algorithm that, you know, this artificial
22:02
intelligence algorithm that we've built off of
22:04
all of these wage records, we're
22:06
really confident that it can show people
22:09
good jobs, often better jobs
22:11
than where they came from, but really they're the
22:13
only one who can figure out what the best
22:15
job is. And I think trying to have that
22:17
humbleness here to
22:19
know that we're not going to nail it,
22:23
we hope to facilitate their being able to
22:25
nail it. Right. Can you expand
22:27
on, so it's, when did it go live? We
22:29
went live on November 2nd. Okay.
22:32
So we're still in kind of
22:34
quarter one, if you will, of three
22:37
months or so of collecting the data. Do
22:39
you, are you getting feedback from users of how
22:42
it's been? What are you hearing?
22:44
So, I mean, so there's a handful of
22:46
things that we're getting. So the tool itself
22:48
has a feedback mechanism. Okay. Right. So
22:51
we're asking people when we show them an occupation for them to go
22:53
ahead and let us know. No, you
22:55
know, this isn't, this doesn't work for me yet. Maybe
22:58
like you're close. There's more that I'd like to know or yes. And
23:01
we're seeing people use that tool. So they don't have
23:03
to write. That's an optional part of this process. But
23:06
we've seen thousands of people respond already and
23:09
give us yes answers. The
23:11
tool then allows them to go and explore that
23:13
job further where they can look for training providers
23:15
that provide it. So we're seeing that. But the
23:17
other thing that's really interesting is, you know, when
23:19
we've done a lot of these campaigns before, if
23:21
you do things like, you know, an email campaign,
23:23
it's really hard to get people to open those
23:25
emails. It's even harder to get them to click
23:27
on it. But because this is really built into
23:29
the process, we're seeing really high rates of people
23:32
interact with this tool, you know, at some point
23:34
during their claim. We're still building that out a
23:36
little bit more. We're going to add a lot more clarity to
23:38
what we're seeing. But
23:40
we're seeing a significant percentage on a weekly
23:43
basis of, you know, initial of
23:45
the initial claims numbers turn into
23:47
activity within this tool.
23:50
Okay. And then I have
23:52
technical questions or context I
23:54
will try to provide. So Uplink is the name
23:56
of our unemployment system in Indiana. So
23:59
is it important? embedded within Uplink?
24:01
Absolutely. So yeah, so Indiana, again, we
24:03
talk about Indiana almost, for some reason,
24:05
the condition's just being right here. But
24:08
Indiana is essentially 100%
24:10
online claims filing. And
24:12
so other states still will use
24:14
some paper mechanisms or maybe some phone
24:16
center processes. But we're essentially
24:19
all online. And so when
24:21
that individual files that unemployment claim, they
24:23
use the Uplink system. But
24:26
so what we do is as they
24:29
register for the system, as they file an
24:31
unemployment claim, we're taking the information that we
24:33
needed for their unemployment claim. And that's the
24:35
information that we're using to generate the recommendation.
24:37
So again, you could think about building this
24:39
in a different way, where we gave it
24:42
its own URL, its
24:44
own website. And we'd have to drive people
24:46
there and encourage them to register and ask
24:48
these questions. And again, a lot of these
24:50
job matching tools do that. But we really
24:52
like the idea here that we didn't need
24:54
anything further from them to generate the recommendations.
24:56
We need something further. Obviously, like we talked
24:59
about this earlier, this is actually
25:01
going to turn into a good outcome. The individual says, yes,
25:03
they enroll in the training, they get the job. The work
25:05
is on them. But we're just trying
25:07
again to make it a lot simpler to find
25:09
this. So yes. All right. So in Uplink, I
25:12
filed my claim. Every week
25:14
they're coming back. If
25:16
they stay unemployed, they come back to
25:18
make that next week's claim for unemployment
25:20
benefits. And they would continue
25:22
to see this recommendation have
25:25
the option to enter this pivot
25:27
tool and interact with these
25:29
recommendations on a weekly basis. OK. And
25:31
then I am going to share an opinion
25:34
when I ask this question. I don't assume you will
25:36
share your own personal opinion. There
25:39
is usually in some of these federal programs around
25:41
employment, there is a requirement for a work search
25:44
that can be pretty goofy, in
25:46
my opinion, where in the past, how
25:48
you make sure an unemployment is
25:51
looking for work has not been
25:56
super effective. I don't know the right word I'd
25:58
use. and
26:00
how you measure that is difficult, is if they
26:02
click through and they say, yeah, I was in
26:04
this kind of manufacturing, I wanna be in this
26:06
one, is there any automation to work search requirements?
26:09
Okay, so we'll
26:11
talk through this here for a second. First,
26:13
I agree with you, work search is tough.
26:16
I think it's a really important concept that
26:18
someone is on unemployment benefits, that the idea
26:20
is that it's temporary and the transition to
26:23
their next job. And by the way, overwhelmingly,
26:25
there's very few people exhaust 26-week unemployment, most
26:28
claim less than that. So
26:30
we have this work search requirement every week, and
26:33
this has been with the unemployment insurance system since
26:35
it started in the 1940s. But
26:38
to a large extent, we're not really, I mean,
26:40
we're saying, hey, you have to go look for work, we're
26:42
not really telling them how to do it or how to
26:44
be effective as we do this. And so
26:46
that requirement remains, we want to see
26:48
this tool be useful. And so we're
26:50
waiting for this feedback, but I absolutely
26:53
can see the world where these things
26:55
sort of merge, where we're helping this
26:57
person connect and say, look, the use
26:59
of this tool would substitute for a
27:02
work search activity in this tool would
27:04
satisfy these work search requirements. But I think what we
27:06
want to make sure, again, is
27:08
we want to be able to look at
27:11
the data and see that we're actually achieving
27:13
better outcomes when we do that than the
27:15
traditional method. Do you have a hypothesis
27:17
about if, not
27:20
just if the algorithm
27:22
is helpful or not as we sort of
27:24
watch it play out, but is it possible
27:26
that it helps reduce exhaustion of
27:28
claims? Do you have a hypothesis about whether that would
27:30
be the case as well? Again, I hope so.
27:32
You know, we talk about this a lot, so we're
27:34
in these early days, and so our measures of success
27:37
are, hey, are people using the tool? Are they leaving
27:39
us feedback? How much time are they spending it? And
27:41
are they entering it another time? Ultimate
27:44
only measure is, is does it
27:46
make people's lives better? Can
27:49
we drive action that leads to
27:51
a better spot at the end
27:53
of it? The answer to that
27:55
is I very much hope yes, but I
27:57
think that there are also a lot of reasons to believe that
27:59
it could. I mean, I
28:01
just still think that if someone sat down and said,
28:03
look, I know I'm ready for a change in occupation,
28:06
but I've already got bills that are piling
28:08
up. Unemployment insurance, I mean, it doesn't replace
28:11
anywhere near their working wages. And in fact,
28:13
it's really set to always replace less than
28:15
half of their working wages. And
28:17
so the pressure is really on. I mean, if your
28:20
family depends on that income. So even if you wanted
28:22
to sit down and say, look, I really need to
28:24
go through a process to figure
28:26
out what I want my next job to
28:28
be, it's just really exhaustive. And so here,
28:31
I think that by speeding up that process, by
28:33
connecting that person with that next role, my hope
28:35
is that it really does put
28:37
them on a better path. You know, we do
28:39
different things within the tool. So they
28:42
can filter their recommendations by the amount of
28:44
time or training in training that it would
28:46
take. So we show them that information. So
28:48
they could say, look, I'm interested in training,
28:50
but only training that I can accomplish in
28:52
six months or less. They can look at
28:55
just that training. But they can also say,
28:57
look, no, I don't want any training, but
28:59
I'm just interested in an occupational change. And
29:01
again, like I said earlier, I think if
29:03
you ask people to sit down and make
29:05
a list of all the different occupations that
29:07
they could do with no additional training,
29:09
people just don't get very far enough. I wouldn't
29:11
get very far on it either. And so I
29:14
think this really helps that. And hopefully by doing
29:16
so will reduce the amount of time that they're
29:18
on employment. Even more, I
29:20
hope it makes it less likely that they ever
29:22
need unemployment again, or at least in the near
29:24
future. I mean, to me, that might even be
29:26
the better metric. I've sort of said this before.
29:29
If they needed to spend an extra
29:31
couple of weeks on unemployment this round,
29:33
but it resulted in them obtaining the
29:36
kinds of skills they needed so that
29:38
they were less likely to be laid
29:40
off in the future, it'd be a
29:42
positive return to the unemployment fund itself.
29:45
Certainly better for them and better for an
29:47
employer out there who really desires that skill
29:49
set. Yeah, I love it. I
29:52
can imagine that there's a world
29:54
where not just citizens on unemployment
29:56
want this. Is there
29:58
a world where that's possible? I really hope so. So
30:01
we're in the process of doing this
30:03
and we're scoping this out. And the
30:05
only things, there will be some slight
30:07
differences, right? Because the question is, is
30:09
we're going to have to obtain that
30:11
information from them upfront. So
30:13
in the unemployment insurance system, we're able to do this as a matter of
30:16
course. But yeah, we
30:18
want to make this tool available to everyone.
30:20
I think that it would be really great
30:22
if we could use the data that we
30:24
sit on so that someone could ask this
30:26
question. Look, if I wanted to make $5
30:28
more an hour than I did today, what would the
30:31
path look like to get there? And
30:33
so not only to be able to answer that question,
30:35
but to be able to show them the paths that
30:37
people like them have made that successful transition to the
30:39
$5 more an hour. Again,
30:41
I think that's something that almost only
30:44
the state could do because others couldn't get access
30:46
to the data. So yeah, I mean, we're working
30:49
on that now, a different
30:51
interface, sort of a different front door to it
30:54
as we continue to try to improve the tool
30:56
that's already available. Awesome. Because then
30:58
I'm like, it's not just a solution for
31:00
unemployment insurance, it's a solution for wages
31:02
in Indiana, which is good for everyone,
31:04
right? Absolutely. I
31:07
think what you really want
31:09
is for people to be able to meet their
31:11
needs and their goals when it comes to their
31:13
career. And the more that we
31:15
can reduce the effort that it
31:17
takes to figure out which path that would be, the
31:19
more that I think it allows that. So absolutely excited
31:22
to do that. And then so many other potential use
31:24
cases of this, the way that we might be able
31:26
to help employers identify where
31:28
future talent pools come from, those sorts of
31:30
things that are really exciting uses of. Dude,
31:32
the same data set, it's just going to
31:34
be different, sort
31:36
of different approaches and probably
31:38
different interfaces. That
31:41
makes your earlier point land for me even
31:43
further. If
31:46
it can allow employers to select the
31:48
most right employees upfront, would
31:50
they stay longer? Would there be better retention?
31:52
Would there be less unemployment in the first place? Yeah, I
31:54
think just hopefully just information to help
31:57
them understand where. There
32:00
were week we might be able to look
32:02
at where their potential pool comes from and
32:04
also were those workers might be going instead
32:06
just providing employers at El Bulli better information.
32:08
The opposite attract the kind a talented. They
32:10
need to be successful often. Okay some
32:13
money as he this when you look
32:15
ahead the next so lucky the faq
32:17
and look ahead or start the he
32:19
had still have the next five years
32:22
and you can ahead of posts on
32:24
unemployment and underemployment for he said while
32:26
to fifteen. Years. The Oh yeah, bolivia.
32:28
Fifteen Plus Fifteen. Plus okay race.
32:32
To the I know. What?
32:34
what do you anticipate is gonna be next?
32:36
Big challenge. Oh. And
32:39
you know I've I wish that we were better
32:41
at anticipating as early. Nobody saw the challenge that
32:43
we were that we were facing when we did.
32:45
I, you know, look, I'm not one of these
32:47
dues. There's about a I, in fact I'm really
32:49
excited about. but I do think that there are
32:52
a lot of folks suggesting that what will see
32:54
his, you know, sort of just. More.
32:58
Need for individuals to
33:00
continuously upscale throughout their
33:02
career to stay. Here
33:05
to do to stay employed to gainfully employed
33:07
to job progress in their wages. and so
33:09
I think that's you know one thing that
33:11
will look at where that plays out I'm
33:13
not quite sure but what we know, his
33:15
dad skills and education are really in demand.
33:17
As we're moving to the stays were more
33:19
and more can be automated and so I
33:21
know that that ah be a job for
33:23
the system. but I'm not sure that I
33:25
can quite predict exactly what the effects will
33:27
be. But the guys
33:29
are you eating Will will definitely keep an
33:31
eye on that to see if the types
33:33
of you the types. of occupation start to
33:35
change that we see from through the
33:38
unemployment insurance system. Ok, I'm
33:40
looking back as the
33:42
leader. it has had the point
33:44
of half her name an app that i
33:46
got his race you seen this through a
33:49
whole bunch of see that say that i
33:51
think a lot of leaders come in for
33:53
a moment and and are critical to one
33:55
one caught one moment any projects life cycle
33:57
not they don't get to see it through
33:59
for essence yeah Do you need
34:01
advice for a leader that is embarking on
34:03
a journey or something that you thought it
34:05
would be important to look back on and
34:08
say, you know, this is how
34:10
I, the staying power I had, like some
34:12
days was that hard, I imagine. So
34:15
you know, absolutely. It's
34:17
hard not, I mean, so I like the question
34:19
at the same time, it's hard not to almost
34:21
fall into cliche type of, you know, answers, you
34:23
know, we really tried to stay the course, I
34:26
think, you know,
34:28
so much of it when you depend on others
34:30
to buy in, to gain the support for something,
34:32
which isn't, you know, which isn't unique to government,
34:34
I think there are maybe some components of that
34:36
to our, but in any organization, and I think,
34:39
you know, it's to continue to listen to the
34:41
concerns and the objections of others and to try
34:43
to make sure that you're reflecting on
34:46
that as you make changes going forward.
34:48
But you know, it really has just
34:50
been to continue to share the message
34:52
and try to generate excitement around the
34:54
potential of the tool. That's
34:58
really been, I think, what ultimately turned us from
35:00
an idea that just kind of sat for a
35:02
while into an actual tool. Well, you
35:05
know, I know that there are a ton
35:07
of people that were part of it. And you,
35:09
you constantly mentioned that it's not just Josh Richardson,
35:11
that you had a great IT team, you had
35:13
great team around you, you had buy in you
35:16
were, it sounds like there was some, to
35:18
your point, how to galvanize how to get that
35:20
buy in and help people see the vision. And
35:23
then we can go figure out how to execute it. But sometimes I think the
35:25
hardest thing is just to get that buy in. Well, absolutely.
35:27
And I mean, again, you, you
35:30
need skeptical
35:32
people throughout the process. You need people to
35:34
challenge these things. Certainly even the people who,
35:37
you know, may have not been sure that
35:39
this could work or that it could
35:41
launch, they've been really important themselves in making it better.
35:43
So yeah, I think I'm
35:45
really excited about where it's at. And I think
35:47
there are a lot of people who did really
35:50
sort of positive things to contribute to it. But
35:53
I think even a lot of those cases, people
35:55
with the biggest concerns, you know, it's just about
35:57
sort of adding another like, you know. Part
36:00
of the problem is that we have a lot
36:02
of different tools and resources and we're saying it's
36:04
hard for people to navigate them And so your
36:06
solution to that is another tool, right? That was
36:08
a really important point that helped clarify some things
36:10
about how What
36:13
we didn't want this to look like right? So that even
36:15
those sorts of things have been really helpful Okay,
36:19
wrapping up on my questions, but one of them I have to ask were
36:22
you ever nervous? Uh,
36:24
yeah, I mean Sort
36:28
of maybe every moment. Do you mean like maybe every
36:30
moment that it wouldn't launch? You
36:33
know, I so what I do feel is
36:35
like some of the big things I use
36:37
a data show and again We're dealing with
36:40
really confidential data one thing I didn't Necessarily
36:42
have to spend a lot of time sort
36:44
of you know I'm really nervous about was
36:46
sort of our privacy and again, it's because
36:48
of the structure in Indiana The
36:51
advice that we have here and those different pieces ain't
36:53
that made it really helpful You know
36:55
those nerves to see like well people use it or
36:57
not, right? We've been I've been telling people Yeah years
36:59
that hey, there's this idea that could really be helpful
37:02
So I think there may be some nerves there early
37:04
on Okay But I think and you know now they
37:06
remain because like I said like really fun to have
37:08
a tool and I think that a Lot of us
37:10
we can sit and we're really happy with sort of
37:12
the data science work and all these different components that
37:14
go into it And so now we're sort of at
37:17
this point where I guess we're kind of we can
37:19
pat ourselves on the back and say hey Look, we
37:21
really feel like we know what people ought to be
37:23
doing But it really won't
37:25
matter unless they ultimately do it again It's
37:27
one of these things about it is that
37:29
it's going to take significant effort on that
37:31
workers part to achieve
37:33
success and so our
37:35
hope is just that we can reduce that effort by
37:37
that littlest bit or Give
37:39
them greater confidence that it's worth persisting
37:42
through that training Because they
37:44
know the outcomes better. That's awesome. I I really
37:46
appreciate The way that
37:48
from that first day on a napkin You
37:51
saw a gap in something and were able to
37:53
try and come up with some ideas and you
37:55
had some solutions on how to sell it And
37:58
I think I also I really respect how
38:00
you hold it loosely. It's not like,
38:03
well, this has to work now because it's been my
38:05
idea, I'm gonna shove it through. It's like, well, do
38:07
people use it? Is there a meaningful outcome? If not,
38:09
okay. Like I would think, but
38:11
okay. And so I really respect the way that,
38:13
I feel like that's tied back to sort of
38:15
your missional alignment with the agency of like, it's
38:17
gotta make a difference. That's why we're here. I
38:20
think that that's neat. Absolutely. Yeah, I
38:22
mean, yeah, you can launch tools, but I
38:24
think the real critical thing here is gonna be, does
38:26
it, does it make it? Is this the only way
38:28
that it will stay around, right? Is obviously if it
38:31
makes a difference, but it's
38:33
obviously the only thing that will validate that it was
38:35
ultimately worth doing. Yep. What
38:37
have we not talked about that we should? No,
38:40
I think it's been pretty thorough. We covered
38:42
most of it. I
38:44
hope that it's clear. It's
38:47
been sort
38:49
of a really fun process, like I said, to
38:51
move through it. And so it was just really
38:53
good to talk to you about it. Awesome.
38:56
Thank you for listening. I'm your host Jess
38:58
Carter. And don't forget to follow the Data
39:00
Driven Leadership wherever you get your podcasts. Rate
39:02
and review letting us know how these data
39:04
topics are transforming your business. We can't wait for
39:06
you to join us on the next episode.
Podchaser is the ultimate destination for podcast data, search, and discovery. Learn More