Making Better Decisions - understanding "fitness for purpose", matching strategy to objectives

[email protected] @lkuceo Copyright Lean Kanban Inc.

Presents

PresenterDavid J. Anderson

Modern Management Methods

San FranciscoMay 2014

Release 1.1

Making Better Decisionsunderstanding “fitness for purpose”,matching capability to strategy & objectives

mailto:[email protected]


Blizzard Skis



Mittersil, Austria

Blizzard Factory



Blizzard is the largest private sector employer in the Pengau Alps region of Salzburgerland,

AustriaMittersil is a factory town with over 400 people relying on the

factory either directly or indirectly for employment



• Innovator• Award winner



In 2007, Blizzard, effectively bankrupt, faced closure from

parent company, Tecnica in ItalyToday Blizzard is the most effective & efficient ski

manufacturer in the world!



What went wrong at Blizzard, a proud & leading brand in alpine

ski equipment?And what enabled a remarkable turnaround, from the brink of

extinction to a return to innovation & profitability?



Orders & Deliveries of Skis

2006 winterskis delivered

to dealers

Timeline for manufacturing, delivery and order placement for ski industry in northern hemisphere prior to 2007

timeNov Dec Jan2006

Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan2007

Feb Mar Apr

Start manufacturing

2007 skis


to dealers

Start manufacturing

2008 skis

Ordersplaced for 2007 winter



2006 was a warm winter and poor snow conditions badly affected

the ski industry as people stayed home and didn’t buy new

equipment

Blizzard dealers were left holding a lot of 2006 inventory that they

would hold & later discount during the 2007 winter



Effects of climate change in the Alps

Glacier recession is

clearly visible

1912 1933

2003



2007 was also a warm winter as climate change began to seriously

affect the AlpsGun shy from 2 bad winters,

Blizzard dealers delayed commitment on 2008 orders until

May after the ski season had finished



In 2006 business risks appear to be low


to dealers

Changing climate conditions and its affect on ski dealers dramatically shifts the risk profile of ski manufacturing

timeNov Dec Jan2006


Feb Mar Apr

Start manufacturing

2007 skis


to dealers

Start manufacturing

2008 skis




Previously lead time for delivery is 12 months


to dealers

Traditionally manufacturers have had a full year to make the skis for the following winter

timeNov Dec Jan2006


Feb Mar Apr

Start manufacturing

2007 skis


to dealers

Start manufacturing

2008 skis

Ordersplaced for 2007 winter Lead Time

to manufacture2007 deliveries



By 2007 the risk profile has changed dramatically


to dealers

Dealers still holding 2006 & 2007 inventory decide to wait until the end of the 2007 season to place reduced orders for 2008

timeNov Dec Jan2006


Feb Mar Apr

Start manufacturing

2007 skis


to dealers

Start manufacturing

2008 skis


Ordersplaced for2008 winter

Volume is low due to over-stocking of older models

The period of speculative manufacturing grows from 2

months to 6 months

Time period of building to

forecast rather than against

customer orders

Blizzard fail to anticipate falling demand and over-produce 2008

skis. Bankruptcy is a serious possibility!



Tecnica Group & Blizzard managers turn to their business school education and seek to cut

costs by consolidation & centralization

Centralizing all order processing through Tecnica HQ adds 1 month

to order times, increasing speculative build-to-forecast. As a

result costs go up!



Cutting costs will not make Blizzard “fit for purpose” !!!

What is required to be “fit for purpose” in a period of climate

change, is to defer manufacturing until firm orders are placed!

Blizzard need to cut the lead time to build skis!



By 2010 the market has a new equilibrium


to dealers

Traditionally manufacturers have had a full year to make the skis for the following winter

timeNov Dec Jan2010


Feb Mar Apr

Start manufacturing

2011 skis


to dealers

Start manufacturing 2012 skis


Ordersplaced for2012 winter

Volume is low due to over-stocking of older models

Lead Timeto manufacture2011 deliveries

Lead time to manufacture skis to order is now 6 months. No speculative build-to-forecast

To have a viable business Blizzard need a capability to make skis

twice as fast as before!



In 2010 a Lean initiative was started in the factory. This was followed later with a Kanban

initiative in IT and Quality Assurance

Blizzard becomes the first Lean ski manufacturer in the world!

CIO, Eric-Jan Kaak wins Austrian CIO of the Year 2013

and is promoted within Tecnica Group



Meanwhile, Major Ski Resorts Deploy Snow Cannon to Reduce Risk for Warm Winters



Lesson 1



Blizzard initially made a bad decision because they didn’t

understand the dynamics of their external environment

Once they realized that only manufacturing skis faster &

deferred commitment would make them “fit for purpose” did they

focus improvement efforts where they could be most effective



Understanding“fitness for purpose”



What makes a pizza delivery service“fit for purpose” ?



Meet Neeta - a project manager• Delivery time =

approximately 1 hour• Non-functional quality =

tasty & hot• Functional quality (order accuracy) =

doesn’t matter if small mistakes are made, geeks will eat any flavor of pizza

• Predictability =+/- 30 minutes is acceptable

• Safety =so long as health & safety in food preparation is good, it’s fine



Neeta is also a working mom!

• Delivery time =20 minutes

• Non-functional quality =doesn’t matter too much, it’s pizza!!!

• Functional quality (order accuracy) =it must be cheese pizza! No other flavor is acceptable! (even if you take the pepperoni off)

• Predictability =+/- 5 minutes maximum!!!

• Safety =only mommy worries about that stuff!



Lesson 2



To be “fit for purpose” there is a product component & a

service delivery componentWe need to offer a selection of

different recipes which are tasty & popular. However, we must also

deliver with speed & predictability



Modern creative & knowledge worker businesses often

obsess with product definition & strategy

Operational excellence and service delivery excellence are often overlooked or treated as

inferior management skills



Lesson 3



Neeta has 2 identities –Mother and Project ManagerEach of Neeta’s identities

represents a different market segment for the pizza delivery

service



We need a different set of thresholds for our fitness criteria for each market

segment

Our business needs the ability to “sense” changing customer tastes. As time goes by the

criteria & thresholds for a given market segment may change



Our pizza delivery service can be “fitter for purpose”

by offering different classes of service for each market

segmentBut, do we have the capability to

deliver on customer expectations?



Understanding & improving service delivery capability with Kanban



TestReady

FF

FFF

F F

Commitment Frequency

H

E

C A

I

G

D

Replenishment

Discarded

I

Pull

IdeasDev

Ready

5Ongoing

Development Testing

Done3 35

UATReleaseReady

∞ ∞

The frequency of system replenishment should reflect

arrival rate of new information and the transaction &

coordination costs of holding a meeting

Frequent replenishment & commitment is more

agile.

On-demand commitment is most agile!



TestReady

FF

FFF

F F

Defining Kanban System Lead Time

H

E

C A

I

G

D

Pull

System Lead Time

Discarded

I

IdeasDev

Ready

5Ongoing

Development Testing

Done3 35

UATReleaseReady

∞ ∞

The clock starts ticking when we accept the customers order, not

when it is placed!

Until then customer orders are merely available options

Kanban system lead time ends when the

item reaches

the first ∞ queue



TestReady

FF

FFF

F F

Delivery Frequency

H

E

C A

I

G

D

Delivery

Discarded

I

Pull

The frequency of delivery should reflect the transaction &

coordination costs of deployment plus costs &

tolerance of customer to take delivery

IdeasDev

Ready

5Ongoing

Development Testing

Done3 35

UATReleaseReady

∞ ∞

Frequent delivery is more agile.

On-demand delivery is most agile!



Service Delivery Agility

Service Agility

Commitment frequencyLead TimeDelivery

Frequency Lead T

ime

Short

Long

Deliv

ery

Service Agility

Com

mit

ment

Frequent

Seldom

Frequent

Seldom

MoreAgile

LessAgile

Kanban system dynamics



UnderstandingCost of Delay Risk



Data from Corbis, Seattle, April 2007

Lead times for IT system software change requests deployed during April 2007

Lead Time Distribution

0

0.5

1

1.5

2

2.5

3

3.5

Days

CR

s &

Bu

gs

This is multi-modal data!

The work is of two types: Change Requests (new

features); and Production Defects

This is multi-modal data!

The work is of two types: Change Requests (new

features); and Production Defects

Lead Time Histogram



85% at10 days

Mean5 days

98% at25 days

Change R

equest

s

Pro

duct

ion D

efe

cts

85% at60 days

Mean 50 days

98% at150 days

Mode

Median 45 days

Filter by Type or Class to get Single Modal Data



Understanding Cost of DelayR

oom

nig

hts

sold

per

day

Actual rooms sold

Cost of delay

Estimated additional rooms sold

When we need it When it arrived

Cost of delay is difference in integral between the two curves

timeJan Feb Mar Apr

First sketch the market payoff function for the total lifetime of the opportunity. In this example, a Spring Break promotion for a hotel chain.



Cost of Delay for a limited shelf-life opportunity follows an s-curve shape

Cost of delay function for a Spring Break marketing campaign delayed by 1 month from mid-January based on the difference of 2 integrals on previous slide

time

impa

ct

Total costof delay



When should we start something?

imp

act

When we need it

85th percentile

Ideal StartHere

Commitment point

timeJan10

Nov11

If we start too early, we forgo the option and opportunity to do something else that may

provide value.

If we start too late we risk incurring the cost of delay

If we pull the work into our kanban system on Nov 11 we

have a 6 out of 7 chance of on-time delivery



We can study sensitivity to different start dates

imp

act

When we need it

50th percentile

Later StartHere

Commitment point

timeJan10

Nov25

If we start as late as November 25 we only have a 50% chance

of on-time delivery

However, the cost of delay incurred if we deliver within 60

days is relatively small. We have an 85% chance of achieving delivery with acceptable cost of delay

85th percentile



What is the latest we could start?

imp

act

When we need it

0th percentile

Very latestart

Commitment point

timeJan10

Dec19

If we start as late as December 19 we have 0% chance of on-

time delivery

We have about a 10% chance of a total loss delivering the

promotion beyond the expiry date of the opportunity

85th percentile

total loss



To be certain of delivery without incurring any cost of delay is expensive

imp

act

When we need it

98th percentile

Early Start

Commitment point

timeJan10

Aug11

If we are conservative and do not wish to carry any risk of late delivery or any risk of

incurring an opportunity cost of delay, then we must start as

early as August 13th.

We must commit to our Spring Break 2015 promotion during

Summer 2014!!!



Lesson 4



Lead time is perhaps the most important metric we gain from

kanban systemsLead time coupled with cost of

delay sensitivity analysis is a key enabler of deferred commitment and consensus on when to make

commitments



What if we planned to do a series of marketing

promotions for seasonal opportunities?

Such as, Halloween, Thanksgiving, New Year’s,

Valentine’s,Spring Break, Cinco de Mayo

& 4th of July?



How many service teams do we need?im

pact

time

HalloweenThanksgiving

New Year’sSpring Break

Cinco de Mayo4th of July

Valentine’s

WIP

= 4

To avoid risk of late delivery altogether, we need a WIP of at

least 4.

If current WIP >= 4 then workers will experience

significant periods of slack and utilization will be low.

Commitments must be made 150 days in advance of deliver and up to 240 days prior to the

event

Additional WIP beyond 4 must be of a lower class of service to insure that these marketing promotions can be delivered

without incurring a cost of delay



What if we were willing to accept more risk?im

pact

time




Valentine’s

With a 50% chance of on-time delivery we only need 3 WIP.

We can take on more lower class of service work, or we can

reduce the WIP limit and reduce staffing on the system

by as much as 25%.

Staff will still experience periods of slack if we don’t

introduce lower class of service work

Commitments need only be made 45 days in advance. A 105 day gain from the risk

averse plan

WIP

= 3

If we do reduce the WIP limit and/or adjust staffing we can

expect the lead time distribution to change – hopefully left-shifted

with a shorter tail



What happens if we have a much more agile organization?If observed lead time capability was significantly shorter and predictability greater, what

benefits do we gain?



Change R

equest

s

85% at44 days

Mean 33 days

98% at68 days

Median 26 days

Improved Service Delivery CapabilityShorter tail, much more predictable



Now, how many service teams do we need?im

pact

time




Valentine’s

To completely avoid risk of late delivery we need a WIP of 3.

If WIP >= 3 when data was collected we have capacity. There will also be slack for lower class of service work

Commitments are needed 70 days in advance.

If we were prepared to reduce demand by just 1 project, WIP

= 2 is sufficientWIP

= 3



What if we were willing to accept more risk?im

pact

time




Valentine’s

With a 50% chance of on-time delivery we only need WIP = 2.

Potentially allowing us to reduce staff by 33%.

As a result the lead time distribution will also be left-

shifted further improving agility.

There will still be some slack for lower class of service work.

Commitments need only be made 26 days in advance.

WIP

= 2



Lesson 5



We have to conclude that for a reasonably acceptable

economic outcome, we must accept some risk of late

delivery & incurred cost of delay

To guarantee on-time delivery and no lost opportunity cost due to delay, we must have 33-50% more WIP, staff to work it and

accept large amounts of idle time



Aligning goals with capability



Go on, you can lift it!



Kanban system dynamics

Shelf-life(of business opportunities)

Is your service delivery fit for purpose?

Short(days, weeks,

months)

Medium(months,quarters,1-2 years)

Long(years,

decades)

Lead T

ime

Short

Long

Deliv

ery

Service Delivery Agility

Reple

nis

hm

en

t

Frequent

Seldom

Frequent

Seldom

Pre

dic

tabili

ty

High

Low

Is your service delivery

predictability & agility fit enough for

your business strategy?

If you plan to pursue short shelf-life opportunities, you must measure predictability, lead time, replenishment

& delivery frequency as fitness criteria. Does the capability exist to pursue the chosen strategy

effectively?



Lesson 6



Improve your capabilities before pursuing market

segments or strategies that require levels of service

delivery beyond your reach

With coaching & incremental development a child can grow to

dead lift a large bar bell. Impatience & over-reaching is

likely to end in tears!

With patience, education and a focus on evolutionary

change, your organization can grow its service delivery

capability



Conclusions



1. Understand your external environment before deciding what to change

2. “Fitness for Purpose” has both a product component & a service delivery component

3. Each market segment will have its own fitness criteria and threshold values



4. Lead time coupled with cost of delay sensitivity analysis is crucial for determining start dates

5. For acceptable economic outcomes we must have a mix of work with different risks and expect some late delivery & incurred cost of delay

6. Improve your capabilities before pursuing market segments or strategies that require service delivery you currently cannot achieve



Thank you!



About

David Anderson is a thought leader in managing effective 21st Century businesses that employ creative people who “think for a living” . He leads a training, consulting, publishing and event planning business dedicated to developing, promoting and implementing new management thinking & methods…He has 30 years experience in the high technology industry starting with computer games in the early 1980’s. He has led software teams delivering superior productivity and quality using innovative agile methods at large companies such as Sprint and Motorola.

David is the pioneer of the Kanban Method an agile and evolutionary approach to change. His latest book, published in June 2012, is, Lessons in Agile Management – On the Road to Kanban.

David is a founder of the Lean Kanban Inc., a business dedicated to assuring quality of training in the Lean Kanban Method for managers of those who must “think for a living.”



I’d like to thank Eric-Jan Kaak and the staff at Blizzard for providing access to produce the story of their Lean transformation.

Software maintenance lead time data courtesy of Corbis.

Acknowledgements




Making Better Decisions - understanding "fitness for purpose", matching strategy to objectives

Leadership & Management

lkuceocopyright

pursuing market

product component

measure things

delay sensitivity

pizza delivery

late delivery

lead time