Dave snowden practice without sound theory will not scale
A better method for it export - David Joyce
1. A Better Method
for IT
David Joyce
ThoughtWorks
1
Saturday, 25 September 2010
2. Systems Thinking
The means to obtain knowledge,
and act with prediction and
confidence of improvement.
John Seddon - Freedom from Command & Control
2
Saturday, 25 September 2010
3. ice NOT
Better pract
e
be st practic
No Sin gle Solution
Anything can be improved if
you know how to look.
Everything you need to know
is in your own system
W. Edwards Deming & Taiichi Ohno
4
Saturday, 25 September 2010
4. Are we just building
the wro ng thing righter?
Rather than focusing on going Agile which
may lead to being successful, instead we
should focus on the needs of our customers.
How agile you are doesn’t matter. Whether you are 50% agile,
90% agile, or agile through and through doesn’t matter.
What matters, is that your company is satisfying its customers,
stakeholders, and employees.
Dr Russell Ackoff & Esther Derby
4
Saturday, 25 September 2010
5. We need to bring considerably more We need to encourage a whole
to the table than just the technical “system” view rather than a
ability to transform user stories into locally optimised view
product in a more efficient way
We often develop
solutions based on
sub optimised status quo
are
Softw
t
Projec
Projects often focus on the
needs of a single business unit
Eric Landes, David Anderson & Dr. Peter Middleton
5
Saturday, 25 September 2010
6. Its not “The Business”,
?
Its “Our Business”
Jeff Patten & Mary Poppendieck 6
Saturday, 25 September 2010
7. If we build technology around a
wasteful process, then we are
locking in that process for longer
Dr. Peter Middleton
7
Saturday, 25 September 2010
10. Business Analysis - NO!
Custo mer An
alysis - Yes!
9
Saturday, 25 September 2010
11. Typical Organisation Hierarchy Customer View
Customers
Senior Leaders
Workers
Line Mgt
Upper
Mgt Upper Mgt
Middle Middle Mgt
Mgt
Line Mgt
Workers
Senior Leaders
10
Saturday, 25 September 2010
12. Senior Leaders
Marketing
Front Office
Finance
Sales
IT
11
Saturday, 25 September 2010
13. Senior Leaders
140
days
Marketing
Front Office
Finance
Sales
IT
Hidden costs
11
Saturday, 25 September 2010
14. Senior Leaders
180
I.T
. days
Marketing
Front Office
Finance
.
I.T .
I.T
Sales
IT
. .
I.T I.T
. . .
I.T I.T I.T
. I.T
I .T. I.T
.
Hidden costs
11
Saturday, 25 September 2010
15. Senior Leaders
Outside
in
Flow Flow Flow Flow 20
days
Marketing
Front Office
Finance
Sales
IT
Hidden costs
12
Saturday, 25 September 2010
16. Core Processes vs
Support Processes
Support
Core
Customer
Core
13
Saturday, 25 September 2010
17. Since IT “ca n” sho uld it?
There is little merit in a well
executed project that no one
wants the output from
The Standish data are NOT a good indicator
of poor software development performance.
However, they ARE an indicator of systemic failure
of our planning and measurement processes
Dr. Peter Middleton & Tripp Babbitt & Jim Highsmith 14
Saturday, 25 September 2010
18. “I recently asked a CIO whether he would prefer to deliver a
project somewhat late and over-budget, but rich with
business benefits, or one that is on-time and under-budget
but of scant business value.
He thought it was a tough call, and then went for the
on-time scenario.
Delivering on-time and within budget is part of his IT
department’s performance metrics.
Chasing after the elusive business value, over which he
thought he had little control anyway, is not.”
If you give a manager a numerical target,
he’ll make it, even if he has to
destroy the company in the process
Cutter Sr. Consultant Helen Pukszta & W. Edwards Deming 15
Saturday, 25 September 2010
19. The thing that makes technology
work is not the technology
Does this mean the end of IT?
There is a better way to approach the
use of technology.
Understand and improve, then ask if
technology can further improve
Dr. Peter Middleton & Tripp Babbitt 16
Saturday, 25 September 2010
20. Larger gains can be achieved through better thinking
around the design and management of work
into the wo rk as nee de d
Then pu ll technology
John Seddon 17
Saturday, 25 September 2010
22. Value vs Failure Deman d
Value Demand; what we want
from our customers and spend
time on
Failure Demand; demands we don’t want.
Failure to do something or do something
right for the customer
John Seddon 19
Saturday, 25 September 2010
23. En d to En d Flow
We often find hundreds of steps,
but very few will be valued by
the customer
People may argue over what the value steps are.
The rule is purpose defines the value work
Stuart Corrigan 20
Saturday, 25 September 2010
24. En d to En d Flow
We often find hundreds of steps,
but very few will be valued by
the customer
People may argue over what the value steps are.
The rule is purpose defines the value work
Stuart Corrigan 20
Saturday, 25 September 2010
25. En d to En d Flow
CUSTOMER
DEMAND
We often find hundreds of steps,
but very few will be valued by
the customer
People may argue over what the value steps are.
The rule is purpose defines the value work
Stuart Corrigan 20
Saturday, 25 September 2010
26. En d to En d Flow
Mapping is important, because it helps staff to
change their perspective on the work
Customers
Workers
Line Mgt
Upper Mgt
Middle Mgt
Senior Leaders
John Seddon 21
Saturday, 25 September 2010
27. Syste m Con ditions
John Seddon 22
Saturday, 25 September 2010
28. Pre dictability
Capability an d
People may argue over what we should or
shouldn’t do for our customers, but they
cannot argue over what we actually do
John Seddon 23
Saturday, 25 September 2010
29. Pre dictability
Capability an d
Upper control limit
x x x x
x x x
x x x
Mean x x x
x x
x x x x
Lower control limit
Time
24
Saturday, 25 September 2010
30. Pre dictability
Capability an d
25
Saturday, 25 September 2010
31. Pre dictability
Capability an d on Cause vs
Co m m
S pecial Cause
100
x
x
x x x x x
50
x x x
x x x x
x x
x x x
0
26
Saturday, 25 September 2010
32. Co mparison
Team 1
x Team 2
xx
x x x
x
x x x
x x
x
x x
x
Team 3
x x x
x x xx
Team 4
x x
x x
x
x
x
John Seddon 27
Saturday, 25 September 2010
33. Impro ve Improve performance without
using technology
If the current work uses
technology, leave it in place, work
with it, or treat it as a constraint
Don’t do anything to change
the technology
Instead, change to a
Systems Design Plan the redesign
Setup a clean stream and
Roll changes in
John Seddon 28
Saturday, 25 September 2010
34. When we focus on customer needs, and the organisation as
a system, many of the previous problems, that apparently
required software projects, may well have been ‘dissolved’
The improvement effort can be
targeted to where it has most benefit
Dr. Peter Middleton 29
Saturday, 25 September 2010
35. Can technology further impro ve
this process or system?
Now we can see potential
benefits, from a position of
knowledge, about the work.
We can therefore predict The result is always less
the benefits technology investment in technology, but
solutions will bring
much more value from it
IT is pulled into the work, rather
than dictating the way work
works
John Seddon 30
Saturday, 25 September 2010
36. Measure improvement results
Use o perationa
l
perfor mance d
ata
Split dat
a
after a
chang e
31
Saturday, 25 September 2010
37. A better m etho d for IT
System
Un de rstan d
Me asure
Pull Technology
Improve the work
32
Saturday, 25 September 2010
39. Case Study
Study of query management
34
Saturday, 25 September 2010
40. 35
Walking the flow
Certain data obscured for confidentiality reasons
Total customer experience involves multiple silos and hand-offs between 3rd party call
centre and BBC Divisions
Saturday, 25 September 2010
41. End to end vs SLAs
3rd Party reported following
against SLAs:
• 80% of queries resolved in
1 working day
• 100% in 5 working days
• Call answering SLA is 90%
of calls answered within 30
seconds
• Call abandoned rate is less
than 3%
Customer query end to end times in days – provides a
different perspective.
36
Saturday, 25 September 2010
43. BBC Customer Support System - IVR
Current IVR Query Routing vs Top 5 Query Reasons
IVR questions did not match customer demand
38
Saturday, 25 September 2010
44. Reco m men dations
1. Total system visibility - map end to end experience,
identifying improvements
2. Call handling analysis - establish end to end times
and value/failure demand
3. Align IVR route options with customer needs
4.Supplement SLAs with end to end times and &
value/failure demand metrics
5. Website for customers to easily update their details
39
Saturday, 25 September 2010
45. Case Study
Results
40
Saturday, 25 September 2010
46. Case Study
IT is THE bottleneck
41
Saturday, 25 September 2010
47. !"!#
$"!#
%"!#
&"!#
'"!#
(!"!#
($"!#
)*+,-./#
0123*45#
67.89+:#
!"#"#$
0;<#=>+25+*#
Saturday, 25 September 2010
;7:#?2@9#6A#
AB7C97D93#
!"#"#$
E2C@7@+2#
F+*85G+>#
)9722CC2:#
=.G3,-93##
E2C@7@+2#
E2C@7@+2#
Value Stream Mapping
H5@I7@+2#
<3.G#JC5C+2#
0;<#K-2,C2:#
L3M-35/#
F*C/3#=/+*C35#
=/+*N#
H5@I7@+2#
A..3>/72.3#
Mapped the value stream from idea to completion
O*C/3*C7#
#L3BC31#=/+*C35#
1C/G#P3B39+>3*5#
L37,N#Q+*#
P3B39+>I32/#
R-3-3#
!"#"#$
=>*C2/#)9722C2:#
SC.8+T#
P3B39+>#U#<35/#
L3:*355C+2#U#
;+7,#<35/#
L393753#
)+5/#J3*CV.7@+2#
/35@2:#
F75/3#
W78C2:#J79-3#
48. Case Study
Results
43
Saturday, 25 September 2010
49. Case Study
Measures
44
Saturday, 25 September 2010
50. If you find that despite lots of measures, you don’t really
know much about what matters to customers, then it can be a
powerful starting place for change
Clarity of purpose, and measures that
relate to purpose, are prerequisites to
learning and improvement
Measures should be in the
hands of those doing the work
John Seddon 30
Saturday, 25 September 2010
51. How do you study purpose? Go out into the system and study
demand, what matters to your customer. Look at Value
Demand vs Failure Demand, go and talk to your customers!
Jeremy Cox 34
Saturday, 25 September 2010
52. The worker has to be
responsible, and have the means
to control their own work
When people understand what is happening
where they work, they are able to contribute
more in improving the work.
This results in greater control and flexibility
John Seddon 35
Saturday, 25 September 2010
53. Case Study
Results
48
Saturday, 25 September 2010
54. Case Study
Understanding Capability and
Increasing Capacity for New
Development
49
Saturday, 25 September 2010
55. Value Demand Failure Demand Demand
50
Saturday, 25 September 2010
57. Mean reduced from 30 to 10 days (67%)
Large drop in the spread in variation.
# Days Blocked Recent outliers have proved to be special cause,
waiting for a 3rd party.
52 Data split from financial year end
Saturday, 25 September 2010
58. Case Study
Results
53
Saturday, 25 September 2010
59. Case Study
IT Help Desk
54
Saturday, 25 September 2010
60. Variation
reducing,
but average
# calls not
by much
Low # calls
is xmas period
55
Saturday, 25 September 2010
61. Jun 08 - Sep 08 Oct 08 - Dec 08
UCL: 76 Mean: 16 Longest: 500 UCL: 66 Mean: 14 Longest: 360
Jan 09 - Mar 09
UCL: 63 Mean: 14 Longest: 300
56
Saturday, 25 September 2010
64. Case Study
Results
59
Saturday, 25 September 2010
65. Beyond the BBC
Applying the Method
60
Saturday, 25 September 2010
66. Deming: I find few people in industry Ohno: The important thing with
know what constitutes a system improvement, is to think of new
work methods, not to make new
Drucker: There is nothing so tools or equipment
useless as doing efficiently, that
which should not be done at all
61
Saturday, 25 September 2010