ShmoopTube

00:00

and finance Allah shmoop What are three ways to estimate

00:05

costs All right people is a long one So listen

00:09

up Accounting is easy There really just two lines you

00:13

have to worry about You have revenues on top You

00:16

know that figure represents the money coming in Pretty simple

00:18

Then you've got expenses yet little different Mohr variables Mohr

00:22

Complexities and Mme Or different flavors Will The key driver

00:25

of those complexities is performance that is And while less

00:28

than perfect the first Tesla's he looked like this The

00:31

first space shuttles looked like this The first peanut butter

00:35

factory looked like Oh this and the first horseshoeing factory

00:38

looked like uh well for this Okay so let's illustrate

00:41

We'LL walk backward in business time to a hugely hot

00:44

industry in eighteen sixty eight Horse shoeing No not like

00:48

that This kind Well before they were primarily used for

00:51

good luck or explaining political ideology they were actually used

00:55

as shoes Four forces Yes that was a thing back

00:59

then People were not making this Think of it like

01:02

well you think of a car tire of the car

01:04

tire industry today more or less Okay meet Wilbur Horseshoe

01:08

entrepreneur He believes that the old way of making horseshoes

01:11

is out of date So eighteen forty well in this

01:14

process that choose their handmade one by one in a

01:17

kind of clay mould that the camera on relentlessly until

01:20

they're finally in the shape and form where they are

01:22

useful to the problem Lots of manual labor Great for

01:25

the delts though In triceps in fact all in at

01:28

close to full production of one hundred choose a week

01:30

Wilbur's cost per shoe is about a dollar and yeah

01:33

dollar back then almost bought you a house Wilbur notes

01:36

that there's always a waiting list for shoes like weeks

01:38

or months long so long that there's a kind of

01:41

a gray market for them Not quite a black market

01:44

but a gray one More and more horses are needed

01:46

to you know deliver ice and haul SAS Barilla Tio

01:50

take courting couples on a romantic rides through the streets

01:53

and to do the old fashioned work too So the

01:55

demand is therefore horseshoes Wilbur's challenges to supply that demand

01:59

with an advantage A product that is the same product

02:02

is everyone else's III his point of charity but cheaper

02:06

i e His point of difference So Wilbur seeks to

02:09

build what will eventually become This idea came from his

02:12

grandmama double waffle maker press iron oven thinking before starting

02:16

he knew it'd take him ages to craft Well he

02:19

needed a budget for the project Turns out Wilbur's grandmama

02:22

recently died but don't feel bad She made it to

02:24

the ripe old age of thirty eight and she left

02:27

some cash behind each of her eleven grandchildren including Wilbur

02:31

got one hundred dollars each A virtual fortune back then

02:34

So Wilbur made the first of three cost estimates and

02:37

figuring out what this project you know would cost i

02:40

e The engineering method of costing That's what he used

02:42

Well this process gets all detail Ian the specific to

02:45

the fundamental elements it takes to build something for the

02:48

horse shoeing machine He broke apart one hundred fifty eight

02:51

pounds of hard cast heat treated iron and clay mold

02:55

that he'd have to build or rather sculpt Yep that

02:57

was forty three bucks for that thing that he had

02:59

to load up on the really dense oak logs which

03:02

burned way hotter and longer than pine Maybe five bucks

03:04

there or maybe twenty dollars depending on the season You

03:07

know how quickly he went Then He had all kinds

03:09

of alls and chisels and Advil for his aching back

03:13

Well since it was eighteen sixty eight he just had

03:15

to chew on the bark of some tree or drink

03:17

laudanum prescribed by the itinerant Dr Pennebaker You know all

03:20

that Another twenty bucks Okay Then he went through the

03:22

eighty seven steps needed for this multi shoe waffle maker

03:25

looking horseshoe thing Ito work He'd have to melt the

03:28

steel here than have some kind of to bring it

03:30

in from the vat to the mold The mold had

03:33

to have some kind of powder in it so the

03:35

shoes would pop out easily when he was done Then

03:37

he'd have to for the molten steel in the mold

03:40

and wait the right amount of time Then press down

03:42

and voila ate shoes at a time instead of one

03:45

and they'd all pop out all uniformly formed Instead of

03:49

the manual style where each one was a little different

03:52

He'd pound on them for an eighth of the time

03:54

needed for manually made ones And instead of selling them

03:57

for a dollar will He could sell them profitably for

03:59

only thirty cents each and pretty much likely put everyone

04:02

else in town out of business After all he could

04:05

make eight hundred shoes a day for thousand a week

04:08

instead of just one hundred Everything grew cheaper in this

04:11

process not the least of which was his own labor

04:13

time In the old days he could make maybe two

04:15

shoes an hour and the cost of labor was a

04:17

big percentage of the total cost of things But in

04:19

this case in an hour with a waffle maker sure

04:22

thinking one hour of labour produce something like well one

04:25

hundred shoes maybe more The big idea in this first

04:28

review of cost the engineering method was that everything was

04:32

kind of theoretical That is this method of costing took

04:35

every little element of the build process analyzed it assuming

04:38

it all worked perfectly the first time from day one

04:41

and that the company was in business the next day

04:44

it assumed that every process worked optimally so well What's

04:48

happening Yeah and he's starting to run out of money

04:52

Wilbur here realizes that he should have jettisoned the perfect

04:55

looking engineering method and instead gone straight to the account

04:58

analysis method of costing when making his budget in the

05:01

account analysis it's the processes that are analyzed not the

05:04

discrete units of the build Like any engineering method the

05:08

focuses just on each unit in a vacuum working it

05:11

assumes no structural variability or errors That storm coming that

05:16

blows a bunch of dust into the molten steel Yes

05:19

for gotten the account analysis method looks at the linkages

05:22

between elements like this So in Wilbur's case each process

05:25

is put into a bucket like getting the loads of

05:27

heating elements for the mold is one thing falling in

05:30

the oak keeping it dry being sure nobody steals it

05:34

And then the termites are generally kept away and then

05:36

heating them up so that the wind doesn't blow them

05:39

out or blow the fire to them And we'LL catch

05:41

the rest of the barn on fire And that's just

05:44

one thing of many What about the process of pouring

05:47

the molten steel into the mold Those tubes leak and

05:50

that's a problem The leaks cost a fortune heat loss

05:53

deal loss a mess on the floor and then repair

05:55

of the tubes and most expensively the huge number of

05:58

orders that don't get filled so angry customers just leave

06:02

and go to the old school competitors So tons of

06:05

variables here tons of moving parts in this way of

06:07

looking at costs account analysis methods of costing take all

06:11

these dark scenarios into account all these outlier situations that

06:15

aren't supposed to happen But they dio yeah Murphy's law

06:18

baby So think about what each of them cost just

06:21

in relation to expected output Like if a spill or

06:25

a break in the tubing or poor management of Oakwood

06:28

Supply fails And there just isn't enough heat to make

06:31

the molten steel or even a flood happens up the

06:33

road and makes the barn unworkable for shoeing for awhile

06:36

And Wilbur is down an entire day What does that

06:39

cost Well of Wilbur normally makes eight hundred choose a

06:42

day that he sells for thirty cents and that cost

06:44

him twelve cents in direct cost Then he's giving up

06:47

eighteen cents times eight hundred shoes or one hundred forty

06:51

four dollars in contribution every day of lost productivity That's

06:55

a fortune more than Grandmama even left him So downtime

06:59

costs a ton here and yet downtime is a regular

07:02

part of any business so it has to be accounted

07:04

for an estimated And the real question here then is

07:07

whether it makes sense for Wilber to spend the money

07:09

to hire someone whose job it is full time to

07:12

do nothing other than worry about all these fixes meaning

07:15

that his chief operating shooing officer her CEO Oso does

07:20

nothing other than inspect the equipment and keep replacements on

07:22

hand watching over the supply line of oak and then

07:25

raw steel deliveries Keeping three sets of repair tubing available

07:29

at all times and having the eighteen sixty eight equivalent

07:32

of a Swiffer on hand to clean up after spills

07:34

and you know so on So if Wilbur hires this

07:37

guy for say three hundred dollars a year can this

07:40

guy save him from having to full down days If

07:43

he can Well then he's worth hiring So that's the

07:46

account analysis method of costing It takes into account the

07:49

processes and views production or service not indiscreet unit elements

07:53

but rather in groups of moves that all come together

07:57

to make the product okay moving on the third type

07:59

of costing system is called statistical cost estimation and shockingly

08:04

it uses statistics as the key driver and figuring out

08:07

optimal configurations of resource is in and along the production

08:10

line So why you stats here Like how does the

08:13

application of stats make for better costing estimates Well simply

08:17

put things change Like when Wilbur Shoe Waffler is first

08:20

getting off the ground A lot of problems come up

08:22

but over time once it's put together and actually producing

08:25

well maybe there simply aren't the same high error rates

08:27

as there were in the beginning The process gets better

08:30

Mohr efficient The problem with account analysis is that decisions

08:34

air driven off of whatever happened yesterday or at least

08:36

in the last period of analysis for whatever processes failed

08:40

with statistics applied broadly longer Previous periods can then be

08:45

measured and not just for the cataclysmic failure situation but

08:48

for optimizing dated a productivity For example the burning dense

08:52

oak wood has to produce a melting temperature of at

08:54

least four hundred degrees When the wood is delivered wet

08:57

and warm me it takes more wood to produce the

09:00

same heat The wood crackles a lot and increases the

09:03

need to keep loose strands of dry Hey off the

09:06

floor right because they light up fast account analysis might

09:09

say Always check the wood Yeah these finer hazards air

09:12

dangerous It's expensive to burn extra wood when it's wet

09:15

So in all circumstances put someone in charge of inspecting

09:18

the wood right But statistical analysis would uncover the fact

09:21

that the on ly fires the barn had in the

09:24

last say three years happened right after the delivery of

09:27

wet wood in February Well now you have evidence that

09:30

inspection is only necessary one month out of the year

09:33

you just saved a small fortune that one observation driven

09:36

by a statistical viewpoint could save days or weeks of

09:39

downtime It might also lead to a process change that

09:42

further fixes the problem maybe combining incoming wet wood with

09:45

previously stored dry wood so that there was always a

09:49

blend and thus less risk of starting a barn fire

09:52

Well there are all kinds of narrowing lenses through which

09:54

to view the statistical reviews of elements of the business

09:57

You might look at the wetness of the woods You

09:59

might look at things like days since last death inducing

10:02

accident to a whole set of ratios like cords of

10:05

wood needed produce a thousand shoes and stuff like that

10:08

All those numbers are put into a calculator with the

10:10

intent to optimize Management of precious resource is in the

10:13

process Well one core element with anything stat C You

10:17

have to only look at relevant data or data that

10:20

is within relevant range of things you can control For

10:24

example if there's one hundred year storm that last from

10:26

November to May and all the wood one season comes

10:29

in wet Well then that data has to get thrown

10:32

out because that one season event is so unusual that

10:35

if it's numbers air included in making predictions well throws

10:38

everything else off It's like trying to base anything physically

10:41

done by humans in the water off of whatever stats

10:44

Michael Phelps tests would give you right He's outside of

10:47

the relevant range of most humans and some dolphins so

10:50

you toss out all of his numbers visually Much of

10:53

these observations of data get presented in scatter plots are

10:56

scattered graphs where a fitted Linus hopefully useful in predicting

11:00

future resource allocations so that you go from a trend

11:02

that looks like this into a guesstimate set over here

11:06

And it basically just tells you that you likely need

11:09

a better storage system and covering area to keep your

11:11

hot oakwood dry Maybe it even gives you a mega

11:14

observation like you should buy your wood a year in

11:17

advance and let it dry an entire year And then

11:20

you only burned dry wood with way less fire spark

11:24

risk Well how would you know to find that data

11:27

Well you'd have some minimum number of observations meaning that

11:30

if you only had one or two crackly sparkler fire

11:33

things of hey catching from wet wood explosions Well that's

11:36

not enough in samples to really make an observation or

11:39

a least not one that you can count on being

11:41

repeated in the future But if you had like eighty

11:43

observations well first off Yikes that's a lot of fire

11:46

risk things but the upside of burning down your barn

11:49

eighty times well it's probably enough of a sample size

11:52

to confidently make some predictions and commit to spending like

11:55

spending whatever it cost to buy a tarp to keep

11:58

your wood dry So you've got three ways to estimate

12:00

cost First the engineering method A theoretical look at the

12:04

different aspects of production It's like a plan for how

12:06

things should work It provides a good sketch but doesn't

12:09

take into account much leeway for things not going as

12:12

planned Then you've gotta count analysis It looks at the

12:15

various processes for producing your product It lets you look

12:17

at things that could break down along the production process

12:20

But it can be narrowly focused And finally you've got

12:23

statistical cost estimation which allows you to take a long

12:26

view by crunching various stats You can also find hidden

12:30

efficiencies that might not immediately be obvious with other methods

12:33

of costing Well of course the statistical method is hard

12:36

for Wilber at this point After all his only employees

12:39

Main Mast skill is you know counting to ten And 00:12:42.431 --> [endTime] for that he even needs his feet

Full Transcript