00:00:05
this paper brings together two topics
00:00:07
one very old and well established and
00:00:10
one that's new and emerging the old
00:00:11
topic is Corporate valuation everybody's
00:00:14
talking about how you look at a
00:00:15
corporation and value it the new topic
00:00:17
is customer valuation can we look at
00:00:19
individual customers or groups of
00:00:21
customers and say what they're going to
00:00:22
be worth in the future what this paper's
00:00:25
all about is bringing the two together
00:00:26
in a really rigorous and and practical
00:00:29
Real World Way can we do corporate
00:00:32
valuation from the bottom up by looking
00:00:34
at the value of current and future
00:00:36
customers adding all that up and saying
00:00:38
that's the value of the corporation that
00:00:40
basic idea has been around for a while
00:00:42
it's been done by a few marketers in the
00:00:44
past but it's never had the real rigor
00:00:47
to win over the the respect of let's say
00:00:50
Financial people and accountants uh
00:00:52
that's what we're trying to do the right
00:00:53
way so it's it's customer-based
00:00:55
corporate valuation but done with with
00:00:58
all the rigor with all just the really
00:01:00
high standards and careful use of data
00:01:02
that accountants and and financial
00:01:04
professionals would respect of there
00:01:06
really have been kind of two silos of of
00:01:08
work here you the one in the marketing
00:01:10
domain and the one in the financial
00:01:11
domain and the financial domain has
00:01:13
really kind of hammered home you know
00:01:15
how you value a business by you know
00:01:18
projecting forward free cash flows
00:01:20
discounting those back doing all those
00:01:21
kind of nitty-gritty little Financial
00:01:23
details all in a in a very precise and
00:01:25
theoretically correct way so we want to
00:01:27
make sure to really draw upon that you
00:01:29
know know take this problem that really
00:01:31
has been solved in finance and apply it
00:01:33
in this you know marketing domain you
00:01:35
know where you perhaps uh some of the
00:01:37
financial details have been a little bit
00:01:39
looser one of the really interesting
00:01:41
aspects of This research is that we did
00:01:43
this valuation exercise for Dish
00:01:44
Networks big publicly traded company
00:01:47
many people especially in the US would
00:01:48
have heard of it the most interesting
00:01:50
part about it is that we have had no
00:01:52
contact with anyone at Dish Network I
00:01:54
have never exchanged emails phone calls
00:01:56
gotten data from anyone at dish this is
00:01:59
using purely publicly available data so
00:02:02
the data that we use anybody could get
00:02:04
access to uh and in fact the methods
00:02:06
that we're using are fairly common and
00:02:08
transparent as well so there's really no
00:02:11
secret sauce black magic here it's just
00:02:14
taking publicly available data
00:02:16
reasonably wellestablished methodss but
00:02:18
just combining them in what one might
00:02:21
call a clever but at least thoughtful
00:02:22
way in order to come up with this kind
00:02:24
of valuation and because we've done it
00:02:26
in this case for a company again that
00:02:28
we've had no contact with there's no
00:02:29
reason why we couldn't repeat this
00:02:31
exercise for other companies that make
00:02:33
similar kinds of data available yeah the
00:02:35
other thing that we're going to do to
00:02:36
kind of capitalize on you know we think
00:02:38
is a very fundamental methodology here
00:02:40
is make all this available you know both
00:02:43
we're going to be collecting data not
00:02:44
only on Dish Network but also on many
00:02:47
other companies that also disclose uh
00:02:49
the sort of metrics that we need to
00:02:50
perform customer-based corporate
00:02:52
valuation so we're going to release the
00:02:54
data make that available to everyone so
00:02:56
they can perform the same exercises
00:02:58
themselves and also release the
00:03:00
methodology so that people can actually
00:03:02
Implement these models too you we really
00:03:03
want to democratize this and make this
00:03:06
kind of core concept of valuing the
00:03:08
company by valuing the customers of
00:03:15
widespread yeah so we decided to perform
00:03:17
work on Dish Network really kind of by
00:03:19
happen stance uh there was no cherry
00:03:21
picking involved we didn't say let's
00:03:23
find all the companies that this works
00:03:25
for and then Whittle it down to the ones
00:03:27
where our model fits no you we basically
00:03:29
just pick dish just because it was one
00:03:31
of the first companies that you know we
00:03:33
happened to find a decently long time
00:03:35
series for we actually didn't even know
00:03:37
how long the time series was when we
00:03:39
first began uh working on the company so
00:03:42
we definitely have a very kind of I
00:03:45
almost call it like a pipeline type
00:03:46
procedure for whittling down and
00:03:48
identifying companies that you know you
00:03:50
could perform this analysis on uh I
00:03:53
think that process in and of itself has
00:03:54
been quite interesting basically we
00:03:57
identify Universe of companies there's
00:03:59
all the compan in the stock market and
00:04:01
some percentage of them basically say
00:04:04
words in their filings specifically the
00:04:07
10K and the 10 q that indicate that they
00:04:10
may release customer data the sort of
00:04:12
customer metrics that that we would need
00:04:14
so we identify all those companies and
00:04:16
we then go down and perform this
00:04:18
validation step where we kind of Whittle
00:04:20
those companies down to the list that
00:04:22
actually disclosed the metrics and now
00:04:24
we have a team of uh great Wharton
00:04:26
students here in undergrad as well as a
00:04:29
a few colleagues in India who've been
00:04:31
working with us to actually turn that
00:04:33
into raw data sets for each of those
00:04:36
each of those companies so dish is just
00:04:38
one example but uh there's many others
00:04:40
we found at least 35 that um also
00:04:43
disclose the sort of data that we would
00:04:45
need to to fit these models for so the
00:04:47
real key here would be what kinds of
00:04:50
statements companies put out about the
00:04:52
the number or the nature of their
00:04:53
customers it could be how many customers
00:04:56
they've acquired or how many customers
00:04:57
they have at the end of the period some
00:04:59
companies will go further and say
00:05:01
something about their retention rate or
00:05:02
their churn rate and and some kind of
00:05:05
derived measure that that take some of
00:05:06
the raw count data and turns it into
00:05:08
some kind of more diagnostic metric we
00:05:11
would rather work with the raw data we
00:05:13
just want to know how many customers
00:05:15
came in how long they stayed around how
00:05:17
many were there at any given point in
00:05:19
time uh right now there are absolutely
00:05:22
no standards for disclosing any of this
00:05:25
data or how it would be disclosed one of
00:05:27
the things that we want to do is to
00:05:29
start toate create some of those
00:05:30
standards it's not the objective of this
00:05:32
particular paper but if we can shine
00:05:34
light on the value of these customer
00:05:36
metrics and how they can be useful to
00:05:37
marketers to finance people to people
00:05:39
throughout organizations then maybe
00:05:41
we'll start to have a conversation about
00:05:43
which metrics would be reported how they
00:05:45
would be defined how often they would be
00:05:47
reported uh and standards around the the
00:05:50
derived metrics that come off of them as
00:05:51
well so it's very very early and none of
00:05:53
that has really happened yet in any kind
00:05:55
of formal way but we're hoping that this
00:05:56
is the first step in that direction
00:06:02
uh I'd say the big one is the procedure
00:06:05
really does work you know you can
00:06:07
perform this you can project out the
00:06:08
cash flows but now we're augmenting that
00:06:11
you know kind of core financial model
00:06:13
that you would use if you were just
00:06:14
working at an investment Bank
00:06:15
incorporate this customer metric data
00:06:17
and use that to get an even more precise
00:06:20
estimate of how much these companies are
00:06:21
worth we arrived at a price that was
00:06:23
within something on the order of 5% of
00:06:25
what the company had been trading at at
00:06:27
the time they disclosed uh the filing
00:06:29
that we would have used to train our
00:06:31
model on so we were very happy with the
00:06:33
fact that you know without really any
00:06:35
sort of jimming around with the numbers
00:06:38
uh we were able to come up with a very
00:06:39
sensible valuation for the firm uh there
00:06:42
were a number of other interesting
00:06:43
findings you know for example you really
00:06:45
do need
00:06:46
to allow for the fact that different
00:06:49
customers are are different from each
00:06:51
other you have some who are valuable
00:06:53
some who are are not so valuable you
00:06:55
some who want to leave immediately and
00:06:56
some who are going to stay around for
00:06:58
for a very very long period of time so
00:07:01
even though these companies don't
00:07:02
disclose this sort of data you know we
00:07:04
were able to infer that dish for example
00:07:06
they had a kind of a hardcore loyal
00:07:10
segment of about you know 15% of the
00:07:12
customers that have been around for over
00:07:14
10 years just very very long term so by
00:07:17
allowing for different customers to have
00:07:19
different values uh we arrived at a much
00:07:21
higher valuation than we would have had
00:07:23
we done kind of the standard thing
00:07:25
within marketing which is to kind of
00:07:27
assume that all customers are kind of
00:07:28
like the average customer customer that
00:07:30
would kind of leave you with a price
00:07:31
that's much much lower than you know
00:07:33
what the company had been selling for at
00:07:35
the time anyone who knows my work knows
00:07:37
that I celebrate heterogenity I love to
00:07:40
find and exploit and explain and talk
00:07:43
about the importance of differences
00:07:45
across customers this project is quite
00:07:47
different because we have no direct
00:07:50
indications of heterogeneity we have no
00:07:52
ability to track individual customers to
00:07:54
say oh this one left after a month this
00:07:56
one left after 10 years we have no data
00:07:58
like that we just have aggregated counts
00:08:01
of how many customers came in at a given
00:08:03
time and how many were present at at the
00:08:05
end of a given period we need to infer
00:08:07
the heterogeneity we need to fit models
00:08:09
to this very poor very sparse overly
00:08:13
aggregated data and infer our best guess
00:08:16
about how the customers differ from each
00:08:18
other not only is that essential for us
00:08:20
to come up with accurate valuations but
00:08:22
it leads to all kinds of very very
00:08:24
useful Diagnostics that managers whether
00:08:27
it's of Dish Network or another firm
00:08:29
could really make great decisions from
00:08:31
so for instance we think about the 8020
00:08:33
rule that a small fraction of our
00:08:35
customers are going to capture a
00:08:37
disproportionate value from the
00:08:39
corporation it's it's a really valid
00:08:40
concept I spend a lot of my time trying
00:08:42
to take a company's internal data to try
00:08:44
to uncover the 8020 rule is it really as
00:08:47
concentrated as 8020 or maybe it's it's
00:08:49
less concentrated than that in this case
00:08:51
we don't have the data to do that but we
00:08:54
do it anyway because we're building
00:08:56
really good models on the sparse data we
00:08:58
can still make inferences about how long
00:09:01
different kinds of customers have stayed
00:09:02
around we find out in this particular
00:09:04
case our best guess is that it's it's
00:09:06
far less concentrated than 8020 but
00:09:09
still there's a good deal of het gen8
00:09:11
and if we ignore it our valuations and
00:09:14
our ability to make kind of action
00:09:16
oriented statements about the company
00:09:17
would be way
00:09:21
off one of the most interesting
00:09:24
conclusions from our work actually
00:09:25
doesn't appear in the paper in a lot of
00:09:27
the work that I do we like to make hold
00:09:29
out forecast let's look at the models
00:09:31
that we've built and project what sales
00:09:33
or whatever is going to be in the future
00:09:34
and see how well we do and we tend to do
00:09:36
pretty well but in this case besides
00:09:38
just coming up with with forecasts on
00:09:40
our own there already exists a universe
00:09:42
of forecasts which is to say Wall Street
00:09:44
analysts so we have thousands of of Wall
00:09:47
Street analysts or at least dozens of
00:09:49
people who on a very regular basis are
00:09:51
making statements about what the
00:09:53
earnings of Dish Network or any publicly
00:09:55
traded company is going to be one
00:09:57
quarter two quarter sometimes two years
00:09:59
from from now well we can compare our
00:10:01
models against their own models or
00:10:04
judgment or forecast and keep in mind
00:10:06
these are people whose jobs depend on
00:10:09
coming up with reasonably accurate
00:10:10
forecasts so one of the things that we
00:10:12
did and again because the paper is so
00:10:15
rich there's so much going on we
00:10:16
actually didn't include it in there but
00:10:17
we're always happy to talk about it is
00:10:19
we compared our own forecast to those of
00:10:22
the Wall Street analysts and on average
00:10:24
they were actually considerably more
00:10:26
accurate uh as showing that not to say
00:10:29
that the Wall Street analysts do a bad
00:10:31
job but some of these very simple
00:10:32
marketing models applied to relatively
00:10:34
simple sparse data can do at least a
00:10:37
good a job as folks who have access to
00:10:39
the CEO and and and and uh and data that
00:10:42
that we couldn't even imagine having
00:10:44
yeah the Crux of this work is really
00:10:46
coming up with customer lifetime values
00:10:49
so we're going to take our crystal ball
00:10:51
out and we're going to project forward
00:10:52
what every customer is going to do into
00:10:54
the infinite future and you know really
00:10:57
that's kind of what's driving this model
00:10:58
for you know what revenues are going to
00:11:00
be and thus what the stock price is
00:11:02
going to be and what's most surprising
00:11:04
is that even when we're making such
00:11:06
short-term forecast we're able to do so
00:11:08
well you know so essentially to come up
00:11:10
with a stock price that's more of a
00:11:12
long-term game but we're actually able
00:11:14
to do very well uh kind of doing
00:11:16
short-term predictions as well so that
00:11:18
was definitely very striking to
00:11:23
me I've been spending a lot of time
00:11:25
lately talking about customer centricity
00:11:27
trying to get companies to understand
00:11:29
there's gold in them their Hills when it
00:11:31
comes to the customer data and it can
00:11:32
not only improve their marketing
00:11:34
practices but pretty much every part of
00:11:36
the organization we can make better
00:11:38
decisions about how our sales people are
00:11:40
doing about what kind of products we
00:11:41
develop how well our production people
00:11:43
are doing and on the finance side as
00:11:45
well so for me this idea of
00:11:47
customer-based corporate valuation has
00:11:49
been kind of a a Holy Grail I've been
00:11:51
going after that if I can win over the
00:11:52
CFO and other people who are running the
00:11:55
finance accounting and and control parts
00:11:57
of the business then I can get these
00:11:59
strategic ideas of customer centricity
00:12:01
to spread much more broadly much more
00:12:03
impactfully throughout the organization
00:12:06
so for me it it really is reaching out
00:12:08
and saying hey Finance guys a lot of the
00:12:10
stuff that I've been talking about to
00:12:11
the marketing people it's relevant for
00:12:13
you too and it will help you make better
00:12:14
decisions about your company and maybe
00:12:16
your competitors and so there I've been
00:12:19
really happy to see the results but of
00:12:20
course it's much more than just that
00:12:23
yeah one of the other kind of practical
00:12:24
implications is you kind of just
00:12:26
speaking from a financial standpoint
00:12:28
I've been a hedge fund for a number of
00:12:30
years before coming back seeing the
00:12:32
light and going to to Academia so you
00:12:34
can take the warten kit out of Finance
00:12:35
but you can't take the finance out of
00:12:37
the Wharton kid uh but you could
00:12:39
definitely construct some sort of a
00:12:40
trading strategy to be able to exploit
00:12:42
this data systematically across all the
00:12:44
companies that disclose it so again
00:12:47
there's on the order of you know say 00
00:12:49
companies let's say uh in the entire
00:12:52
universe who disclose enough data to be
00:12:54
able to construct some sort of
00:12:55
customer-based valuation model uh the
00:12:58
idea would be you could find some
00:12:59
companies that seem overvalued and some
00:13:01
that seem undervalued uh just by the
00:13:04
results that the models entail for each
00:13:05
of those individual companies so buy the
00:13:08
ones that are cheap sell the ones that
00:13:10
are expensive you know you're not taking
00:13:12
on any any real risk U I think such a
00:13:15
strategy could be potentially very
00:13:16
interesting and in turn that makes it
00:13:18
very interesting for the executives of
00:13:20
those firms knowing that you know people
00:13:22
are buying and selling their stocks
00:13:23
based on this information so the whole
00:13:26
idea of standardizing these metrics
00:13:29
making sure that you're improving the
00:13:31
the health of your business kind of from
00:13:33
a a Bottoms Up way you know making sure
00:13:36
your customer metrics are strong I think
00:13:38
it just reinforces something that they
00:13:39
really probably know they should be
00:13:40
doing in any case as Dan mentioned
00:13:43
there's all kinds of interesting hedge
00:13:44
fund possibilities and a lot of folks
00:13:46
who are managing hedge funds or working
00:13:48
with corporate valuation other ways have
00:13:50
been looking very carefully at this work
00:13:52
I want to say right now that I'm a
00:13:53
marketing guy I have no intention of
00:13:55
developing a hedge fund if other folks
00:13:56
want to take this research and do it
00:13:58
that's great I'll find that really
00:14:00
rewarding even if I don't have a piece
00:14:01
of it for me it's it's more about
00:14:04
corporate strategy more generally it's
00:14:05
more about making better use of the data
00:14:07
assets that we have for me the big
00:14:09
payoff would be if if folks with hedge
00:14:11
funds and other investors go pounding on
00:14:13
the doors of other public corporation
00:14:15
saying You must reveal some of these
00:14:17
other uh customer oriented metrics that
00:14:19
we can't make the right decisions unless
00:14:21
you do so and maybe for accounting
00:14:23
standards boards to start to to Make
00:14:25
Some Noise about that as well so if we
00:14:27
can change the kind of data that's being
00:14:29
put out there and the kind of
00:14:30
conversations that are having Beyond
00:14:32
just the making money thing I want to
00:14:33
change business practices and and the
00:14:35
kinds of data that we look at and the
00:14:37
way that we use it say one other very
00:14:39
practical implication is there's a whole
00:14:41
stream budding stream of research about
00:14:43
basically if I'm going to spend uh
00:14:46
marketing dollars to acquire customers
00:14:48
should I take that hit on my income
00:14:50
statement immediately or should I
00:14:52
capitalize that as an asset because I'm
00:14:54
acquiring these customers who are worth
00:14:56
a lot of money uh you know part of our
00:14:58
research is showing just how valuable
00:14:59
these customers are and one of the
00:15:01
implications is they're going to stay
00:15:03
around for a very long period of time so
00:15:05
we come up with an estimate that the
00:15:07
average customer at dish is going to
00:15:09
remain a customer for about 5.5 years
00:15:12
that's that's very very long and you
00:15:14
have some customers who are going to be
00:15:15
around for over 10 years so if I'm going
00:15:18
to spend subscriber acquisition expense
00:15:20
dollars yeah I really I'm investing in
00:15:23
an asset you know these customers are
00:15:25
assets but they're not acknowledged as
00:15:26
such so I think over 80% of Dish's
00:15:30
subscriber acquisition costs are
00:15:32
immediately expensed um and as a result
00:15:35
you know they kind of take the hit and
00:15:37
of profits don't look as good in the
00:15:39
short term so you know there's been work
00:15:42
by people like Shuba shason at Boston
00:15:44
University uh who are kind of saying
00:15:47
that doesn't really jive with the
00:15:48
underlying economics of the business if
00:15:50
I'm going to spend money and I'm going
00:15:52
to basically acquire an asset uh I
00:15:55
should be able to recognize the asset
00:15:58
and kind of have the expense associated
00:16:00
with the asset bleed away as the as the
00:16:03
assets value bleeds away so you know we
00:16:05
think that this kind of dovetails very
00:16:06
nicely with that stream of
00:16:12
research we see these ginormous
00:16:14
Acquisitions taking place in many cases
00:16:16
where where one company is buying
00:16:18
another largely because of their
00:16:20
customer base so you can think about
00:16:21
when Facebook bought what's app or just
00:16:24
endless rumors about Yahoo and AOL we're
00:16:27
just buying the customer base but what
00:16:29
are those customers worth well that's
00:16:31
what this research is all about and it's
00:16:32
interesting to see sometimes in the
00:16:35
Press you'll see kind of back of the
00:16:37
envelope calculations about what One
00:16:39
customer base is worth and they're
00:16:40
appalling the the the analyses that
00:16:42
they're doing are terrible the numbers
00:16:44
that they're coming up with are way way
00:16:45
off so we like the idea that people are
00:16:48
are having that kind of conversation but
00:16:50
we want to make it a little bit more
00:16:51
educated we want the answers to those
00:16:53
questions to be as rigorous as the kinds
00:16:56
of things that we would expect to see in
00:16:57
a conversation about Finance and
00:16:59
Accounting not kind of lightweight
00:17:01
approximations that might be Associated
00:17:04
more with marketing yeah so one example
00:17:06
that's been in the news lately is kind
00:17:08
of on the topic Dish Network uh they're
00:17:11
in the news about potentially being
00:17:13
acquired or merging with other company
00:17:16
so you know they've been in the news
00:17:17
themselves of you trying to find a
00:17:19
partner in some way to move forward so I
00:17:22
think this is a perfect example of how
00:17:24
customer-based Corporate valuation could
00:17:26
be useful actually for the company that
00:17:28
that you know we specifically are
00:17:30
looking at now our research is all about
00:17:32
valuing the customers so all these other
00:17:35
assets that the company may have that
00:17:37
you know really aren't coming from the
00:17:38
value of the customers we'll leave that
00:17:40
to the to the finance guys but
00:17:42
specifically for the value of the
00:17:43
customers we think there's real gold in
00:17:46
basically looking at them kind of the
00:17:48
way that that we are and all the better
00:17:50
if we had internal company data we could
00:17:52
just make those you know uh forecasts
00:17:55
even more precise but you know we think
00:17:58
even with this Network itself there's
00:18:00
definitely real potential to think about
00:18:01
it in this
00:18:05
way yeah I'd say perhaps one of the
00:18:08
biggest misperceptions that we've seen
00:18:09
is not acknowledging the Paramount
00:18:12
importance of the value of the customer
00:18:14
to the value of the firm really
00:18:16
establishing that link and then just
00:18:17
focusing on the value of the customers
00:18:19
I'd say perhaps one of the most striking
00:18:21
examples of this is just in the
00:18:23
accounting rules themselves so if a
00:18:25
company was to buy a chair and put it
00:18:28
right down over here like we're sitting
00:18:30
in now they would be able to capitalize
00:18:32
that as an asset on their balance sheet
00:18:34
and they don't recognize an expense
00:18:36
immediately but if I was to spend money
00:18:38
to acquire a customer that customer
00:18:41
doesn't show up anywhere they're called
00:18:42
an intangible asset and one of the
00:18:45
interesting byproducts of our research
00:18:47
is that for Dish example for example
00:18:50
these customers are expected to
00:18:52
basically stay with the firm for about
00:18:54
5.5 years so they have very very long
00:18:56
useful lives and in fact dish in
00:18:59
particular their example of the chair is
00:19:01
is the actual dish itself so the money
00:19:04
that they spend on these dishes that you
00:19:05
see uh they're able to capitalize and
00:19:08
they have a useful life of only four
00:19:10
years so actually you have this kind of
00:19:12
striking example It's Like A Tale of Two
00:19:13
Worlds that you have this kind of inert
00:19:17
object that has a short useful life I'm
00:19:19
able to capitalize that but you know
00:19:21
here are these valuable customers that
00:19:23
are generating cash flow over very long
00:19:25
periods of time and I need to expense
00:19:27
that immediately so we think that's
00:19:29
you know really something that should be
00:19:30
looked into more carefully another
00:19:33
important point that the paper addresses
00:19:34
would be the the gap or I'd rather say
00:19:36
the Synergy between marketing and
00:19:39
finance that that marketers will often
00:19:41
talk about this idea of customers as
00:19:43
assets we often say almost just in a
00:19:44
purely conceptual way if we add up the
00:19:47
value of the customers boom we get the
00:19:48
value of the corporation and and while
00:19:50
it's it's a nice idea we haven't seen
00:19:53
work so far that has really won over the
00:19:56
Finance and Accounting Community say
00:19:57
yeah you know what there there really is
00:19:59
something there now I'm happy to say
00:20:01
that there have been a few Mavericks in
00:20:02
Finance and Accounting who have been
00:20:04
saying some of these words but it hasn't
00:20:06
really become mainstream yet we're
00:20:08
hoping that this paper not because it's
00:20:10
coming out of marketing but just because
00:20:11
of the of of the rigor involved on the
00:20:14
data side on the analysis on the
00:20:16
substantive implications is going to win
00:20:18
over broad interest from folks in all
00:20:20
different business disciplines and say
00:20:22
you know what the marketing folks really
00:20:23
do have something to add to the Finance
00:20:25
and Accounting conversation and of
00:20:27
course Finance and Accounting have a lot
00:20:28
to add to marketing if we can break down
00:20:30
stereotypes if we can break down silos
00:20:32
and make everybody across the
00:20:34
organization not only smarter but but
00:20:36
looking at the same kinds of metrics
00:20:38
then that's got to be a great thing for
00:20:43
shareholders so all the work thus far
00:20:45
has basically been to value a single
00:20:47
company so we lay out a framework and
00:20:49
then we apply it to a company we want to
00:20:51
show in the next paper that essentially
00:20:53
you can apply this across many many many
00:20:55
companies along two Dimensions you know
00:20:57
one it's not just rehashing the same
00:20:59
analysis that we did for one company you
00:21:01
know a 100 times over it's really that
00:21:03
there's a whole broad spectrum of data
00:21:06
availability these different companies
00:21:07
some that have disclosed a lot some that
00:21:09
have disclosed very little and there's
00:21:12
actually a whole heck of a lot of
00:21:13
information that you can learn about
00:21:15
those companies that may not have
00:21:16
disclosed very much when you kind of
00:21:18
look at the patterns across companies so
00:21:20
we want to bring all those companies
00:21:21
together in what they call a basian
00:21:23
hierarchical model and be able to fill
00:21:25
in a lot more of the blanks for
00:21:26
companies where the data is not so good
00:21:28
and apply this in a systematic way
00:21:30
across all of them
00:21:32
simultaneously another big question is
00:21:34
what metrics should be disclosed right
00:21:37
now we're limited by what companies such
00:21:39
as Dish Network choose to disclose but
00:21:40
are those the right kinds of metrics
00:21:42
that investors should be demanding and
00:21:44
would it be the same kinds of metrics
00:21:45
for every kind of business so an
00:21:47
important caveat of this work is that
00:21:49
we've only looked at companies that
00:21:51
operate in a contractual setting where
00:21:53
we sign customers up they pay on some
00:21:55
regular basis and then they go and the
00:21:57
company knows that they left it's
00:21:59
observable churn so this model applies
00:22:01
to pretty much any contractual setting
00:22:03
and that's great but there's many
00:22:05
companies out there such as Amazon or a
00:22:07
grocery store where there's no formal
00:22:09
contract where people just kind of buy
00:22:11
things on an occasional basis and then
00:22:13
there's this long Gap and the company is
00:22:15
left wondering is that customer gone or
00:22:19
are they just taking a snooze between
00:22:21
purchases if you're in a non-contractual
00:22:23
setting what would be the right kind of
00:22:25
data that should be disclosed again if
00:22:27
you're an investor in a non-contractual
00:22:29
company and you're pounding on the door
00:22:31
to that company uh what data are you
00:22:32
going to be demanding for them to
00:22:33
disclose it's going to be different
00:22:35
kinds of metrics so we want to have a
00:22:37
broader understanding about what are the
00:22:39
right kinds of metrics that should be
00:22:40
disclosed in different kinds of business
00:22:41
settings and then of course what kind of
00:22:43
analysis would we lay on top of it in
00:22:45
order to do the corporate valuation
00:22:58
[Music]
