Inside Cover or Main Title Page, with Logo (add carriage return if only one line)

1
Analytical Problem Solving
Gemini Skills Workshop
August 1998
2
Learning objectives

• Understand the key steps in the problem solving
  process.
• Learn tools and techniques that are availablefor
  each step of the process.

3
Analytical Problem Solving
The significant problems we face cannot be
solved at the same level of thinking we were at
when we created them- Albert Einstein
The difficult is what takes a little time the
impossible is what takes a little longer -
Fridtjof Nansen
4
As consultants we need to have a structured
approach to problem-solving

• We work in groups.
• We work with complex problems.
• Other consultants or clients may have to continue
  our work (e.g. in later phases of the project or
  when implementing the solution).
• Need to know where we are and what has been
  done
• Need to understand the process that leaded to the
  result/recommendation
• A structured approach helps the client follow
  where we are
• Our solution will be shared with people that did
  not take part in the problem solving process.
• It is easy to miss a step.
• Current steps often seem less important than
  future steps.

5
Key steps in the problem solving process
1. Determine criteria 2. Determine decision
process
1. Determine what we know and what we
dont 2. Gather information 3. Identify
constraints 4. Determine if you should proceed
1. Measure expected benefits 2. Collect
feedback 3. Incorporate feedback into ongoing work
1. Identify possible causes 2. Design
tests 3. Perform tests 4. Determine causes or
re-test 5. Determine to proceed
1. Determine solution approach 2. Develop
solutions
1. Compare with decision criteria 2. Decide on
solution(s) 3. Validate
1. Prepare action plans 2. Prepare follow-up plan
and measures 3. Implement
There are many variations of this process, but
these are the basic steps you should follow.
6
Clarify the problem
STEP 1 CLARIFY PROBLEM
Determine DecisionCriteria
Follow-upandMeasure
Implement Solution
Evaluate Solutions
Identify Solutions
InvestigateCauses
ClarifyProblem
7
Clarifying the problem is the most important step
in this process
STEP 1 CLARIFY PROBLEM

• A problem can be defined as a gap between where
  we are and where we want to be. This gap should
  be measurable.
• Be aware that perception is reality. Although
  some client problems we encounter are very
  logical and factual, such as machine breakdowns,
  most client issues are based on peoples
  perceptions of problems, such as poor customer
  service.
• Because of this, most problems will not require
  an optimal solution, but will have many adequate
  solutions.
• Ensure that the problem statement accurately
  depicts the client situation. It will determine
  your entire course of action.

8
These are the steps involved in clarifying the
problem
STEP 1 CLARIFY PROBLEM - EXAMPLE
1. Determine what we know and what we dont.
Using a table with What, Who, When, Where, How,
and Why can help define what information needs to
be gathered. Example Urgent customer requests
are not being addressed.

• What We Know What We Dont Know
• What Customers are complaining about Are
  requests being lost, forgotten, or not lack of
  responsiveness answered initially?
• Who Only customers with urgent E-mail,
  voicemail,or telephone requests or rush
  orders communications?
• Where Headquarters Customer place of origin
• When Problem only in last two months
• Why No documentation of requests or System
  error? Not recorded by service orders agent?
  Message not received?
• How

9
Steps to clarify problem (continued)
STEP 1 CLARIFY PROBLEM
2. Gather the information you need in order to
define the problem statement. You may begin to
identify possible causes, but that should really
be done at a later step. 3. Identify constraints
- Who is the client for this problem and what is
important to that client? Consider time frame
(short-term vs long-term), costs, resources
required, level of effort vs value-added,
etc. 4. Define the problem statement. Validate
the problem with the client. Do we agree that
this is really the problem at hand? 5. Determine
how to proceed. Seriously consider if the time
and effort involved creates enough benefit or if
this problem will disappear as the result of
other activities.
Most importantly, this step frames the
investigation before we begin tackling the causes.
10
One very easy way to understand and define a
problem is to ask Why? 5 times (5 Whys)
STEP 2 INVESTIGATE CAUSES - EXAMPLE

• Real Client Example
• Why are we shipping orders late? Because we
  cant meet our production schedule.
• Why cant we meet our production schedule?
  Because we are constantly changing it.
• Why do we change it?To accommodate late orders
  from our customers.
• Why do we have late orders?Because many of our
  customers dont know what their orders are by the
  order cut-off date.
• Why do we have a cut-off date?So we can create a
  production schedule and meet our shipping dates.

Client problems may require several iterations of
the 5 Whys.
11
Strengths, Weaknesses, Opportunities, and Threats
(S.W.O.T.) can clarify complex issues and define
direction

• S.W.O.T. analysis is a general tool that can be
  used across key areas
• Product mix
• Profit/pricing
• Promotions
• Space management
• Supply chain
• Definitions

12
Investigate causes ( or perform analysis)
STEP 2 INVESTIGATE CAUSES
Determine DecisionCriteria
Follow-upandMeasure
Implement Solution
Evaluate Solutions
Identify Solutions
InvestigateCauses
ClarifyProblem
We use two general methods to approach problems -
using analogies to previous experience and
breaking the problem down into smaller pieces.
13
Step 2 is to investigate causes of the problem
STEP 2 INVESTIGATE CAUSES

• 1. Identify possible causes. We need to do this
  in order to carve out a manageable piece of work
  by narrowing the scope of the problem to the most
  probably causes.
• There are many tools we can use to investigate
  causes. The two basic ways to analyse problems
  for causes are
• Use analogies. We naturally relate the current
  problem to our previous experiences. As experts,
  we should be able to develop plausible hypotheses
  to explain the problem.
• Break the problem into smaller subsets of
  problems (chunking). In conjunction with our
  hypotheses we can also dissect the problem into
  its variable components and determine which of
  these components is most likely to be causing the
  problem.

14
Now we are ready to test our possible causes
STEP 2 INVESTIGATE CAUSES

• 2. Design tests or analytics. Tests can include
  surveys, interviews, process flows, pareto
  analyses, control charts, etc. it is unlikely
  that you will have to create an entirely new
  analytic because so many already exist, in Gemini
  and externally. Well-designed tests should
  directly prove or disprove hypotheses and should
  consider one problem variable at a time.
• 3. Perform tests. Ensure that the test will not
  be a burden for the client and that they want to
  do it. Otherwise, the results may not be
  accurate.
• 4. Determine causes based on test results. (Or
  re-test, if necessary)
• 5. Determine how to proceed. Is the cause within
  our sphere of influence? How does it compare
  with our constraints?

It seems logical that the next stop would be to
develop solutions. but to make our time more
effective, we should plan ahead to determine what
a good solution looks like.
15
The Fishbone or Cause and Effect Diagram is
simple tool for investigating causes
STEP 2 INVESTIGATE CAUSES - EXAMPLE
CAUSES FOR A CAR NOT TO START
Man
Machine
Dead battery
No gas
Left lights on Bad Switch Electrical
Bad choke
Lost keys
Worn out
Wrong fuel
Lemon
No oil
Vapor Lock
Car will not start
Don't know how to start
Out of tune
Too cold
Rental
Parts Stolen
No antifreeze
Repo'd
Gets wet in the rain
Methods
Material
Milieu (Environment)
5 M fishbone - Man, Machine, Materials, Methods
and Milieu (Environment)
16
Supply Chain fishbone is an mental model for
looking at a business
STEP 2 INVESTIGATE CAUSES - EXAMPLE
Raw material packaging
Supply Chain Effectiveness
Order
Order size
EDI
Frequency
Cost order
Bar coding
Pricing
Segmentation
Know your customer
Verification
Perfect orders
Lead time on orders
Transportation costs
Order processing
ECR
Payment terms
Promotions
Distribution/logistics
Accounts receivables
Order error rate
Inventories
Discounts
Credit control
Vendor mgt.
Order status record
Customer
Connectivity to other core processes
Performance Measures
Order Fulfilment
Distribution Logistics
17
A DuPont tree is a structured way to look at
causes
STEP 2 INVESTIGATE CAUSES - EXAMPLE
18
We rely heavily on surveys and interviews to gain
information quickly - Example Culture Survey
Please complete the following questions by
circling the appropriate number on a scale from 1
to 6, where 1strongly disagree and 6strongly
agree. Answer the statements according to the
way it is in your organisation at the moment on
the left column. Then move to the right-hand
column and respond according to the way it
should be in your opinion. There are no right or
wrong answers. Please answer all questions from
your own perspective.
Strongly disagree Disagree Agree
Strongly Agree 1 2 3 4 5 6
Please circleThe way it is 1 2 3 4
5 6 1 2 3 4 5 6 1
2 3 4 5 6 1 2 3
4 5 6 1 2 3 4 5 6
1 2 3 4 5 6
Please circleThe way it should be 1 2 3
4 5 6 1 2 3 4 5
6 1 2 3 4 5 6 1 2
3 4 5 6 1 2 3 4
5 6 1 2 3 4 5 6
1. People here act when they see things that
need to be done 2. People in this organisation
try to resolve their differences
constructively 3. People here readily accept
responsibility for their actions and
decisions 4. Cost is one of the most important
factors taken into account before decisions are
made here 5. Inconsistencies exist between what
is said and done. Managers dont walk the
talk 6. People here constantly explore new and
better ways of doing things 7. People here
listen to customers and respond to their
needs 8. There is a concerted effort to perform
better than the competition
19
We use focus interviews on every Analysis and
Design project
Focus Interview Guide General Information 1) Year
s with Company _______________________________
Years in current position _____________________
__________ Number of reports
_______________________________ 2) a) What is
your understanding of your companys top three
business objectives? b) What is your
companys vision? 3) a) What are your groups
top three business objectives? b) How are
they (or will they be) measured? efer back to
the scope graphic on page 1 if the concept of GTS
is not clear.) 4) What do you consider to be
your groups three greatest strengths?
Your Group
Overall
20
Paretos Law states that 20 of the sources cause
80 of the problem
Customer Service Complaints Pareto Analysis
35M
30
25
20
Number of Occurrences/month
15
10
5
2
2
1
1
1
0
No one Answers Phone
Routed to wrong person
Dont know the answer
Dont return calls
Unhelpful
Discourteous
Gum chewing
Hard to understand
Customer Complaints about our Customer Service
We can use this principle to determine where to
focus our improvement efforts
21
Determine Decision Criteria
STEP 3 DETERMINE DECISION CRITERIA
Determine DecisionCriteria
Follow-upandMeasure
Implement Solution
Evaluate Solutions
Identify Solutions
InvestigateCauses
ClarifyProblem
22
The third step is to determine our decision
criteria
STEP 3 DETERMINE DECISION CRITERIA

• Determine the decision criteria. Refer back to
  the constraints. Consider
• Needs vs. wants
• Long-term vs. short-term
• Interim steps
• Risks
• Costs
• Timing
• Desired benefits
• Ranking or prioritizing the decision criteria
  (most important to least important)
• Determine decision process. Who needs to be
  involved in the decision? Who has final say?
  What method will we use - voting, client chooses,
  numerical rankings, a dart board?

Doing this now avoids looking foolish later.
23
Quadrant Analysis provides a framework for making
decisions

• Quadrant analysis is a general tool that can be
  used across different levels of analysis
• Corporate portfolios
• Customers
• Products

• You can compare any two axes relevant to your
  problem
• Quality vs. cost
• Market share vs. market potential
• Volume vs. margin

MarketShare
MarketGrowth
24
And then helps determine possible actions based
on your findings - Example Quadrant Decision
Matrix
Quadrant
PossibleCategory Roles
Possible Actions
Opportunity gaps (Higher growth/low share)
Profit contributor Variety image

• Review planograms - are category and fastest
  movers underspaced?
• Review pricing mix - is pricing of key items too
  high versus market?
• Review promotions - are category and key items
  under-promoted versus market?
• Review product mix - is mix wrong for target
  customer segments? Any new, faster-moving items
  not being carried?
• Tie-in promotions with higher margin consumption
  items

Winners (High share/higher growth)
Traffic draw Cash generator Price image

• Continue current programs
• Increase promotional support
• Review space management to ensure minimal
  out-of-stock potential
• Add good performing items not carried but
  available in market
• Be first with new items
• Review pricing and gross margins to see if
  selected price reductions can enhance image and
  increase growth and share

Profit contributor Transaction builder

• Review product mix versus

Sleepers (Good share/lower growth)
Questionables (Low share/lower growth)
Profit contributor Variety image

• Review product mix versus market (variety index)
• Delete poorest performance items (brands,
  flavors, sizes)
• Raise prices if appropriate
• Promote category to meet market
• Minimize space allocated

25
Key steps in the problem solving process
Determine DecisionCriteria
Follow-upandMeasure
Implement Solution
Evaluate Solutions
Identify Solutions
InvestigateCauses
ClarifyProblem
26
Finally, we can develop solutions
STEP 4 IDENTIFY SOLUTIONS

• 1. Determine solution approach. This can be
  almost anything, like
• Brainstorming
• Benchmarking
• Best practices
• Modelling techniques, e.g., decision modelling,
  business modelling, process modelling
• Vision engineering
• Organisation design
• Any number of Gemini methodologies
• 2. Develop solutions. It is good practice to
  develop alternative scenarios, if applicable.

27
Benchmarking can provide a gauge of what good
looks like - Purchasing KPI Benchmarks
Average Benchmark
Benchmark Range
Key Performance Indicator

• Purchasing function expense as a percentage of
  sales
• Purchasing function expense as a percentage of
  purchase value
• Purchasing headcount as a percentage of total
  company headcount
• Active suppliers per purchasing employee
• Percentage of all active suppliers that account
  for 90 of total purchase value
• Average actual time to develop/negotiate a
  contract
• Percent of deliverables received on-time within
  the most recent 12 month period

0.3 1.2 1.1 50 20 9 wks 88
0.06 -gt 3.0 0.7 -gt 7.0 0.3 -gt 4.5 6 -gt
182 5 -gt 75 2 -gt 26 wks 63 -gt 98
1
2
1 Median 392 Median 91
Source Center for Advanced Purchasing Status
28
We can also benchmark best practices - Purchasing
Best Practices
Company

• Best Practice/Strength

• Partnering establishes 3-5 year contracts with
  suppliers
• Partnering 70 of North American Automotive
  Operations contracts are long term (3-5 years)
• New product development with preferred
  suppliers, Ford uses black box design
  responsibility. Ford specifies a parts function
  and lets suppliers figure out the best way to
  manufacture it
• Partnering suppliers participate in developing
  design guidelines for new products
• Supplier management communications sector
  trimmed its supplier base from 4,200 in 1985 to
  1,155 in 1989
• Technology supports a common global database for
  critical material purchases
• Supplier management has created Motorola
  University, an education and training center in
  Schaumburg, Illinois where suppliers and its own
  employees brush up on basic quality concepts as
  well as learn the more advanced techniques in
  quality control

• Outboard Marine Corporation
• Ford Motor Company
• Motorola

Source Purchasing
29
Key steps in the problem solving process
STEP 5 EVALUATE SOLUTIONS
Determine DecisionCriteria
Follow-upandMeasure
Implement Solution
Evaluate Solutions
Identify Solutions
InvestigateCauses
ClarifyProblem
30
Evaluating solutions becomes easy because we have
already laid the groundwork
STEP 5 EVALUATE SOLUTIONS

• 1. Follow the decision process and compare with
  decision criteria
• 2. Decide on solution(s)
• 3. Validate solutions with initial constraints
  and your sphere of influence

31
An Impact/Effort Matrix is a useful tool for
prioritizing work and identifying early wins
High
Level of impact (results)
High
Low
Level of effort required
32
Key steps in the problem solving process
Determine DecisionCriteria
Follow-upandMeasure
Implement Solution
Evaluate Solutions
Identify Solutions
InvestigateCauses
ClarifyProblem
33
Step 6 is to implement the solution(s)
STEP 6 IMPLEMENT SOLUTIONS

• 1. Prepare action/implementation plans. Include
  responsibilities and time frames
• 2. Prepare follow-up plan and measures
• 3. Do it!

34
Key steps in the problem solving process
Determine DecisionCriteria
Follow-upandMeasure
Implement Solution
Evaluate Solutions
Identify Solutions
InvestigateCauses
ClarifyProblem
35
Dont forget to measure and follow-up!
STEP 7 MEASURE AND FOLLOW UP

• 1. Measure improvements and compare with expected
  benefits
• 2. Collect feedback
• 3. Incorporate feedback into on-going work

You will learn more about performance measurement
in later sessions
36
In everything we do, Plan (think) - Do - Review
THINK
THINK
REVIEW
DO
DO
THINK
DO