1
00:00:06,244 --> 00:00:09,610
- Before I dive in, I want
to make a comment here.

2
00:00:09,610 --> 00:00:12,220
As we've been involved with
safety implementations now

3
00:00:12,220 --> 00:00:16,940
for SAFe flag early on,
implementations for about a decade,

4
00:00:16,940 --> 00:00:19,397
we see a couple things consistently.

5
00:00:19,397 --> 00:00:23,863
We see struggles in the areas
of Lean-Agile leadership:

6
00:00:23,863 --> 00:00:26,146
convincing and teaching
people like me to take on a

7
00:00:26,146 --> 00:00:28,162
different style of
management, to really be a

8
00:00:28,162 --> 00:00:30,752
true servant leader and to
understand lean and flow

9
00:00:30,752 --> 00:00:32,255
and agile development.

10
00:00:32,255 --> 00:00:36,708
We also see issues in basic
practices and when we see

11
00:00:36,708 --> 00:00:39,516
issues in basic practices,
we typically see things like,

12
00:00:39,516 --> 00:00:41,703
"that won't work here because",

13
00:00:41,703 --> 00:00:44,105
"we tried that and that didn't work."

14
00:00:44,105 --> 00:00:46,246
And the question becomes
why didn't it work?

15
00:00:46,246 --> 00:00:48,632
Is there something wrong with
that practice or is there

16
00:00:48,632 --> 00:00:51,116
something wrong with the
implementation of the practice?

17
00:00:51,116 --> 00:00:53,149
So you constantly come
back to try to understand,

18
00:00:53,149 --> 00:00:55,201
what are the principles of SAFe?

19
00:00:55,201 --> 00:00:57,461
What are the things that,
if I do right, for sure,

20
00:00:57,461 --> 00:00:59,633
whether I adopt that
practice or have that role,

21
00:00:59,633 --> 00:01:01,143
I'm gonna be successful with.

22
00:01:01,143 --> 00:01:05,143
So I would pause that
here that this next module

23
00:01:05,992 --> 00:01:09,557
may be the most critical
module in all of the course

24
00:01:09,557 --> 00:01:12,043
because if you and your
leaders and your peers

25
00:01:12,043 --> 00:01:15,266
get this right, you'll be
armed with the thinking tools

26
00:01:15,266 --> 00:01:17,561
you need in addition
to a Lean-Agile mindset

27
00:01:17,561 --> 00:01:20,674
to really address and
adopt the right patterns

28
00:01:20,674 --> 00:01:23,311
so they can be successful
in your organization.

29
00:01:23,311 --> 00:01:24,144
So what are they?

30
00:01:24,144 --> 00:01:24,977
There's nine.

31
00:01:24,977 --> 00:01:28,234
Number one is to take an economic view.

32
00:01:28,234 --> 00:01:31,635
Number two is to apply systems thinking.

33
00:01:31,635 --> 00:01:35,853
Number three is to assume
variability and preserve options.

34
00:01:35,853 --> 00:01:38,146
That sounds interesting. Doesn't it?

35
00:01:38,146 --> 00:01:40,529
Number 4: Build incrementally with fast,

36
00:01:40,529 --> 00:01:42,098
integrated learning cycles.

37
00:01:42,098 --> 00:01:45,141
That sounds like agile-lean
all jammed together.

38
00:01:45,141 --> 00:01:48,351
Number 5: Base milestones
on objective evaluations

39
00:01:48,351 --> 00:01:49,380
of working systems

40
00:01:49,380 --> 00:01:50,685
We are going to show
you how you can do that

41
00:01:50,685 --> 00:01:53,385
even early in development
and how those milestones

42
00:01:53,385 --> 00:01:56,245
of working systems bring
a lot more value than our

43
00:01:56,245 --> 00:01:58,560
traditional kind of software
development lifecycle

44
00:01:58,560 --> 00:02:00,916
Stage-Gate milestones.

45
00:02:00,916 --> 00:02:03,590
Visualize and limit
WIP, reduce batch sizes,

46
00:02:03,590 --> 00:02:05,047
and manage queue lengths.

47
00:02:05,047 --> 00:02:07,504
You'll note we've got a pretty
power packed principal there.

48
00:02:07,504 --> 00:02:10,528
That's probably three in one
but it looks better as one,

49
00:02:10,528 --> 00:02:13,202
plus it makes the shape
nice for the principles.

50
00:02:13,202 --> 00:02:16,605
Apply cadence, and synchronize
with cross-domain planning.

51
00:02:16,605 --> 00:02:18,988
Unlock the intrinsic motivation
of knowledge workers.

52
00:02:18,988 --> 00:02:21,453
We mentioned that briefly, now
it comes back as a principle.

53
00:02:21,453 --> 00:02:24,408
And, in the same way,
decentralize decision-making

54
00:02:24,408 --> 00:02:27,045
comes back as a leadership
and a management principle

55
00:02:27,045 --> 00:02:30,228
to help achieve fast delivery of value.

56
00:02:30,228 --> 00:02:31,061
So let's dive in.

57
00:02:31,061 --> 00:02:33,280
Let's talk, first of all, wait,

58
00:02:33,280 --> 00:02:34,785
why do we have these principles?

59
00:02:34,785 --> 00:02:37,224
Found this really cool quote
from Deming, "A common disease

60
00:02:37,224 --> 00:02:41,038
that afflicts management the
world over is the impression

61
00:02:41,038 --> 00:02:42,610
that our problems are different."

62
00:02:42,610 --> 00:02:43,527
I'll bet you've heard that.

63
00:02:43,527 --> 00:02:45,616
For those of you who are
consultants and have been

64
00:02:45,616 --> 00:02:47,955
in the field working
with other enterprises

65
00:02:47,955 --> 00:02:50,278
you know, for sure, that
the common response to

66
00:02:50,278 --> 00:02:53,261
"that won't work here because
our problems are different."

67
00:02:53,261 --> 00:02:56,228
And Deming goes on to say,
"They are different to be sure,

68
00:02:56,228 --> 00:02:58,532
but the principles that
will help to improve quality

69
00:02:58,532 --> 00:03:00,915
of product and service
are universal in nature."

70
00:03:00,915 --> 00:03:03,588
And that's where he spent
his life, coming up with

71
00:03:03,588 --> 00:03:06,308
the principles that
improve quality of service.

72
00:03:06,308 --> 00:03:08,710
So, we know that the Lean-Agile
transformation with SAFe

73
00:03:08,710 --> 00:03:11,571
can deliver many benefits
but it's a pretty big change

74
00:03:11,571 --> 00:03:13,832
and every implementation is different.

75
00:03:13,832 --> 00:03:16,207
So we might have to tailor
and we might have to tune.

76
00:03:16,207 --> 00:03:18,951
We might have to say, well,
why does that practice

77
00:03:18,951 --> 00:03:19,814
not work here?

78
00:03:19,814 --> 00:03:21,686
Why does it work here?

79
00:03:21,686 --> 00:03:24,794
Understanding why the
practices work is part of

80
00:03:24,794 --> 00:03:28,417
a leadership's mandate to
know what it is they must do.

81
00:03:28,417 --> 00:03:30,920
And most importantly, if a
practice needs to change,

82
00:03:30,920 --> 00:03:33,502
understanding the principles
will assure that the change

83
00:03:33,502 --> 00:03:36,076
moves the enterprise
in the right direction.

84
00:03:36,076 --> 00:03:37,676
We can make a change but we want to be

85
00:03:37,676 --> 00:03:39,586
more lean and agile, not less.

86
00:03:39,586 --> 00:03:42,017
Now occasionally we might
have to adapt to an existing

87
00:03:42,017 --> 00:03:44,362
circumstance or constraint
but we know that the direction

88
00:03:44,362 --> 00:03:46,828
of change has to move forward.

89
00:03:46,828 --> 00:03:49,037
And we have this incredible
body of knowledge.

90
00:03:49,037 --> 00:03:51,369
We have, I don't know
what the count is anymore,

91
00:03:51,369 --> 00:03:54,842
125 or so books that have
contributed directly to this.

92
00:03:54,842 --> 00:03:56,779
Great books like "Product
Development Flow",

93
00:03:56,779 --> 00:03:58,399
"The Lean Machine", Reinertsen's book,

94
00:03:58,399 --> 00:03:59,748
and the Lean start up series.

95
00:03:59,748 --> 00:04:01,989
I've contributed a book or two over time.

96
00:04:01,989 --> 00:04:04,427
It's a credible body of knowledge.

97
00:04:04,427 --> 00:04:08,748
Well, they all basically say
principles over practices

98
00:04:08,748 --> 00:04:10,978
and that will help us
drive these implementations

99
00:04:10,978 --> 00:04:13,129
so that's the background.

100
00:04:13,129 --> 00:04:15,813
We synthesize these
principles that people like

101
00:04:15,813 --> 00:04:20,241
Dan Oosterwal spend x time
writing The Lean Machine

102
00:04:20,241 --> 00:04:22,797
and tells you a 5 year story

103
00:04:22,797 --> 00:04:25,678
of transforming Harley
Davidson using Lean thinking.

104
00:04:25,678 --> 00:04:28,434
He distills that down to
a couple hundred pages.

105
00:04:28,434 --> 00:04:31,105
We pull out of that what we
need to succeed with SAFe

106
00:04:31,105 --> 00:04:33,366
and that's one of the
ways we create the IP

107
00:04:33,366 --> 00:04:36,129
behind the framework because
we read, practice in the field,

108
00:04:36,129 --> 00:04:39,846
learn, synthesize, distill
and basically repurpose

109
00:04:39,846 --> 00:04:42,194
and provide that information back to you.

110
00:04:42,194 --> 00:04:44,454
Number 1, take an economic view.

111
00:04:44,454 --> 00:04:48,061
Okay, we often talk about
agile is better than waterfall.

112
00:04:48,061 --> 00:04:49,986
We believe it is, but why?

113
00:04:49,986 --> 00:04:52,074
Is it just better because it's
better, it's more fun to be

114
00:04:52,074 --> 00:04:52,907
part of an agile team?

115
00:04:52,907 --> 00:04:55,163
Well, it is more fun to
be part of an agile team.

116
00:04:55,163 --> 00:04:57,503
But I think the fundamental
economic paradigm

117
00:04:57,503 --> 00:04:59,294
is a little simpler than that.

118
00:04:59,294 --> 00:05:02,577
I think the fundamental
economic paradigm is as follows;

119
00:05:02,577 --> 00:05:06,744
If our model focused on these
things, and these phases,

120
00:05:08,013 --> 00:05:12,180
one cannot argue the fact
that value is delivered here.

121
00:05:13,308 --> 00:05:14,780
It's just a fact.

122
00:05:14,780 --> 00:05:15,751
It all comes together.

123
00:05:15,751 --> 00:05:16,584
It gets integrated.

124
00:05:16,584 --> 00:05:18,731
Now, it's easy to make fun
of waterfall development

125
00:05:18,731 --> 00:05:22,117
as if we haven't used it to
get to this point and haven't

126
00:05:22,117 --> 00:05:24,925
built incredible systems with
it but let's not think about

127
00:05:24,925 --> 00:05:27,704
making fun of it so much
as understanding it and

128
00:05:27,704 --> 00:05:30,368
understanding the basic
value proposition here is

129
00:05:30,368 --> 00:05:34,662
the fact that value happens
here, however you measure it.

130
00:05:34,662 --> 00:05:35,495
It happens here.

131
00:05:35,495 --> 00:05:37,676
Nothing happens in between.

132
00:05:37,676 --> 00:05:41,835
So, if I approached somebody
in the PMO organization

133
00:05:41,835 --> 00:05:45,678
who is literally chartered
with evaluating programs

134
00:05:45,678 --> 00:05:49,672
with respect to the waterfall
model and ask 'em a question,

135
00:05:49,672 --> 00:05:52,774
I say, if I have a choice,
if I pick this point in time

136
00:05:52,774 --> 00:05:56,134
and said if I have a choice
between delivering you

137
00:05:56,134 --> 00:05:59,237
a comprehensive set of
requirements documents or

138
00:05:59,237 --> 00:06:03,264
delivering you the first part
of a system that reflects

139
00:06:03,264 --> 00:06:06,437
our best understanding of
that time as to how the system

140
00:06:06,437 --> 00:06:08,735
should behave, which would you choose?

141
00:06:08,735 --> 00:06:11,093
Every time they'll say, "well
I'd rather have the system

142
00:06:11,093 --> 00:06:12,767
but I didn't know I could
have it because I didn't know

143
00:06:12,767 --> 00:06:13,693
what the requirements were.

144
00:06:13,693 --> 00:06:15,619
We have enough. We have a vision.

145
00:06:15,619 --> 00:06:16,863
We'll take a first step.

146
00:06:16,863 --> 00:06:21,192
And, if we just, migrate
through time doing that,

147
00:06:21,192 --> 00:06:24,449
and we don't worry so much
about when design is complete

148
00:06:24,449 --> 00:06:27,204
because, gosh, this part of
the design wasn't complete

149
00:06:27,204 --> 00:06:30,147
until here so it's awfully
tough to say oh my design

150
00:06:30,147 --> 00:06:31,200
has to be complete here.

151
00:06:31,200 --> 00:06:33,151
Well I haven't really proven much yet.

152
00:06:33,151 --> 00:06:37,363
Well swap, anytime, this
set of proxy elements

153
00:06:37,363 --> 00:06:38,895
for the real system.

154
00:06:38,895 --> 00:06:40,488
Well what happens to economics?

155
00:06:40,488 --> 00:06:42,331
Let's say it's a think that
could actually deliver.

156
00:06:42,331 --> 00:06:44,412
Even if I can't deliver it,
I'm getting value because

157
00:06:44,412 --> 00:06:46,327
I'm getting feedback, but
let's pretend like it's a small

158
00:06:46,327 --> 00:06:47,566
think I can deliver.

159
00:06:47,566 --> 00:06:50,140
Well, immediately, in
the first, let's just say

160
00:06:50,140 --> 00:06:52,991
these are iterations,
and the first interation,

161
00:06:52,991 --> 00:06:54,280
well, I worked pretty
hard for a couple weeks

162
00:06:54,280 --> 00:06:56,071
I learned some things
and we communicated well

163
00:06:56,071 --> 00:06:57,340
but I didn't really give you anything.

164
00:06:57,340 --> 00:06:58,348
I didn't give you a new feature.

165
00:06:58,348 --> 00:07:02,613
I didn't deploy this one, small
toolkit that we delivered.

166
00:07:02,613 --> 00:07:04,648
I didn't deploy even
the first set of slides.

167
00:07:04,648 --> 00:07:06,203
I don't have any guidance for it.

168
00:07:06,203 --> 00:07:09,165
I don't have an enablement
video but I put the slides on

169
00:07:09,165 --> 00:07:11,560
the platform where SPCs can use it.

170
00:07:11,560 --> 00:07:12,856
Is that the full value delivery?

171
00:07:12,856 --> 00:07:13,689
No, but guess what?

172
00:07:13,689 --> 00:07:15,103
They can start using it.

173
00:07:15,103 --> 00:07:16,879
So let's just say they
start using those slides

174
00:07:16,879 --> 00:07:18,614
and we come back later on
and say well the slides

175
00:07:18,614 --> 00:07:22,063
weren't perfect and we've now
added to that the instructive

176
00:07:22,063 --> 00:07:23,673
presenter's notes for the slides.

177
00:07:23,673 --> 00:07:24,506
Well that's pretty cool.

178
00:07:24,506 --> 00:07:27,514
We've added video to that
so this person delivered

179
00:07:27,514 --> 00:07:28,398
the slides here.

180
00:07:28,398 --> 00:07:29,361
They delivered the slides here.

181
00:07:29,361 --> 00:07:32,431
They delivered the slides here
and the value is accumulating

182
00:07:32,431 --> 00:07:34,211
so I'm giving this stair step.

183
00:07:34,211 --> 00:07:36,911
Now, if I didn't take
that, pretend like I erased

184
00:07:36,911 --> 00:07:40,079
all that and start over and
say well we don't do that.

185
00:07:40,079 --> 00:07:40,996
We do this.

186
00:07:41,836 --> 00:07:45,328
Then no value is delivered
until the end state,

187
00:07:45,328 --> 00:07:48,367
and even then, only if it's perfect.

188
00:07:48,367 --> 00:07:50,241
If the deadline is
perfect, I get value here.

189
00:07:50,241 --> 00:07:52,348
So what is this?

190
00:07:52,348 --> 00:07:56,662
This is the integral
value of agile development

191
00:07:56,662 --> 00:07:58,254
versus waterfall.

192
00:07:58,254 --> 00:08:00,073
That's basic economics.

193
00:08:00,073 --> 00:08:02,315
That says, even if you
did waterfall perfect,

194
00:08:02,315 --> 00:08:06,042
you can't compete with the
economics of agile development.

195
00:08:06,042 --> 00:08:07,957
Now that requires some
thinking and you have to think

196
00:08:07,957 --> 00:08:09,263
differently about the problem.

197
00:08:09,263 --> 00:08:11,262
That's when I go oh, if that's the case,

198
00:08:11,262 --> 00:08:12,306
I want to rethink this.

199
00:08:12,306 --> 00:08:14,300
I'm not sure I want to
think so much about that.

200
00:08:14,300 --> 00:08:17,012
I want to think about enough
design, enough requirements,

201
00:08:17,012 --> 00:08:19,753
enough architecture to be able to deliver

202
00:08:19,753 --> 00:08:21,928
that small piece of value.

203
00:08:21,928 --> 00:08:24,981
And, if you wanna see how
this works in kind of a fun,

204
00:08:24,981 --> 00:08:26,973
interactive way, hit pause and go

205
00:08:26,973 --> 00:08:28,592
to the marshmallow challenge.

206
00:08:28,592 --> 00:08:30,230
You can see it here on youtube.

207
00:08:30,230 --> 00:08:33,336
It's a short, seven minute
video about experimentation

208
00:08:33,336 --> 00:08:37,594
and prototyping and how to
build a system with everybody

209
00:08:37,594 --> 00:08:41,467
from kindergarteners
to MBAs to executives.

210
00:08:41,467 --> 00:08:42,916
It's really fun and that'll
teach you a little bit

211
00:08:42,916 --> 00:08:44,968
about agile development
so, if you don't mind,

212
00:08:44,968 --> 00:08:47,829
take that seven minutes,
grab a glass of water,

213
00:08:47,829 --> 00:08:51,602
and watch that video and then come back.

214
00:08:51,602 --> 00:08:53,313
Isn't that a cool video?

215
00:08:53,313 --> 00:08:55,454
Didn't you like the part where
the kindergarteners seemed

216
00:08:55,454 --> 00:08:58,540
to do best because they
made no assumptions about

217
00:08:58,540 --> 00:08:59,836
the final output?

218
00:08:59,836 --> 00:09:02,157
I think you also appreciated
the fact that, hey,

219
00:09:02,157 --> 00:09:04,002
engineers and architects
did pretty well, too.

220
00:09:04,002 --> 00:09:05,893
I think we're thankful for
that because they do think

221
00:09:05,893 --> 00:09:07,250
about how to build the system.

222
00:09:07,250 --> 00:09:09,996
And you learned about how
things, like, executives

223
00:09:09,996 --> 00:09:13,996
do terrible unless they're
working with an admin

224
00:09:15,153 --> 00:09:17,882
or somebody who helps them
coordinate their activities.

225
00:09:17,882 --> 00:09:19,374
Gosh, that sounds familiar.

226
00:09:19,374 --> 00:09:22,653
But a really neat way to
think about additive learning.

227
00:09:22,653 --> 00:09:25,281
Take the marshmallow challenge.

228
00:09:25,281 --> 00:09:27,031
Now, I described how,

229
00:09:28,770 --> 00:09:30,462
that entered a process.

230
00:09:30,462 --> 00:09:32,749
Having a marshmallow be this
high, it delivers value.

231
00:09:32,749 --> 00:09:36,142
Having just a set of
slides, go out to an SPC,

232
00:09:36,142 --> 00:09:38,833
without all the full products
around provides value.

233
00:09:38,833 --> 00:09:41,506
Let's talk about that from a
slightly different perspective.

234
00:09:41,506 --> 00:09:43,117
I would like you to do this exercise.

235
00:09:43,117 --> 00:09:45,908
So let me set it up and then
I'll do the conclusion for you.

236
00:09:45,908 --> 00:09:47,797
You can do this by yourself or in a group.

237
00:09:47,797 --> 00:09:49,019
Here's the exercise.

238
00:09:49,019 --> 00:09:52,923
You have two scenarios so
you're gonna create a chart.

239
00:09:52,923 --> 00:09:55,916
Right? Of scenario one and scenario two.

240
00:09:55,916 --> 00:09:58,696
On the x axis is time.

241
00:09:58,696 --> 00:10:00,613
On the y axis is value.

242
00:10:03,133 --> 00:10:04,971
I'm giving you the easy
part, time and value.

243
00:10:04,971 --> 00:10:08,418
Then we're gonna have
months, we're going to have

244
00:10:08,418 --> 00:10:11,677
four months like that and value like this.

245
00:10:11,677 --> 00:10:13,827
So this is one unit of
value, two units of value,

246
00:10:13,827 --> 00:10:14,708
three units of value.

247
00:10:14,708 --> 00:10:16,042
So that's the set up.

248
00:10:16,042 --> 00:10:19,201
So now when we look at this
exercise, you have three

249
00:10:19,201 --> 00:10:21,206
features, you can deliver them serially or

250
00:10:21,206 --> 00:10:22,585
you can do them in parallel.

251
00:10:22,585 --> 00:10:23,773
Now these are important projects.

252
00:10:23,773 --> 00:10:25,509
Features, projects, whatever.

253
00:10:25,509 --> 00:10:26,342
It doesn't really matter.

254
00:10:26,342 --> 00:10:29,317
We are just trying to show
you how to think through this

255
00:10:29,317 --> 00:10:31,710
problem and, most importantly,
to get together your

256
00:10:31,710 --> 00:10:33,718
key stakeholders and do this
exercise because this is

257
00:10:33,718 --> 00:10:34,923
really an eye opener.

258
00:10:34,923 --> 00:10:38,667
So what it asks you to do
is to plot the difference

259
00:10:38,667 --> 00:10:40,854
between value delivery

260
00:10:40,854 --> 00:10:43,743
with serial and parallel
approaches to projects.

261
00:10:43,743 --> 00:10:46,172
In other words, starting
three things at once or just

262
00:10:46,172 --> 00:10:47,658
doing one at a time.

263
00:10:47,658 --> 00:10:50,475
So take about five
minutes or so to do that.

264
00:10:50,475 --> 00:10:52,291
I hope you had fun with that exercise.

265
00:10:52,291 --> 00:10:53,286
When we do it in class,

266
00:10:53,286 --> 00:10:54,960
people tend to struggle a little bit.

267
00:10:54,960 --> 00:10:57,127
They typically start off
thinking, oh there's some value

268
00:10:57,127 --> 00:10:58,955
here but, you know what,
I've got to get at least

269
00:10:58,955 --> 00:11:01,681
a full slide deck out there
or maybe it's the feature,

270
00:11:01,681 --> 00:11:03,687
haven't been able to deliver that.

271
00:11:03,687 --> 00:11:05,921
So when you take the project
serially, we go, well nothing

272
00:11:05,921 --> 00:11:09,007
really happens here so we
have to basically agree

273
00:11:09,007 --> 00:11:10,807
no value in that first iteration.

274
00:11:10,807 --> 00:11:14,931
But, wow, right here, I get a
unit and that's pretty cool.

275
00:11:14,931 --> 00:11:17,279
And what happens at the
end of the next month?

276
00:11:17,279 --> 00:11:18,666
I get another unit.

277
00:11:18,666 --> 00:11:20,896
That's pretty cool as well.

278
00:11:20,896 --> 00:11:22,688
And what happens at the
end of the next month?

279
00:11:22,688 --> 00:11:25,061
I get another unit.

280
00:11:25,061 --> 00:11:27,390
This diagram is starting to look familiar.

281
00:11:27,390 --> 00:11:28,667
It's like what we had earlier.

282
00:11:28,667 --> 00:11:31,358
So this is that
incremental value delivery,

283
00:11:31,358 --> 00:11:34,708
kind of dot dot dot, dot
dot dot, two units of value,

284
00:11:34,708 --> 00:11:35,970
three units of value.

285
00:11:35,970 --> 00:11:38,557
Pretty good stuff within a
very short period of time,

286
00:11:38,557 --> 00:11:41,499
a single sprint or a
single PI or a few days,

287
00:11:41,499 --> 00:11:43,371
whatever you're context is.

288
00:11:43,371 --> 00:11:45,495
You've been able to deliver value.

289
00:11:45,495 --> 00:11:46,656
That's the serial case.

290
00:11:46,656 --> 00:11:47,612
So this is the serial case.

291
00:11:47,612 --> 00:11:49,195
It looks like this.

292
00:11:51,426 --> 00:11:53,100
For those of you who are
thinking ahead and have some

293
00:11:53,100 --> 00:11:55,127
familiarity with Lean you
also recognize that's a

294
00:11:55,127 --> 00:11:57,323
working process limit of one project.

295
00:11:57,323 --> 00:11:59,023
We are only gonna do
one project at a time.

296
00:11:59,023 --> 00:12:00,203
We are not going to start three.

297
00:12:00,203 --> 00:12:01,668
So these things are gonna start to come.

298
00:12:01,668 --> 00:12:04,861
Now let's do the parallel
case next and people start

299
00:12:04,861 --> 00:12:06,976
thinking about it and they go,
well you know it might look

300
00:12:06,976 --> 00:12:09,702
like this and we go, oh it
doesn't really look like that.

301
00:12:09,702 --> 00:12:11,214
And I ask a question,
if you're working on all

302
00:12:11,214 --> 00:12:13,393
three projects, what value
is delivered at the end

303
00:12:13,393 --> 00:12:14,517
of the first month?

304
00:12:14,517 --> 00:12:17,541
Nothing because I picked
the funky way to do it.

305
00:12:17,541 --> 00:12:20,018
How about the second month over here?

306
00:12:20,018 --> 00:12:22,249
Nothing. Well that's not helpful.

307
00:12:22,249 --> 00:12:24,976
What about end of the second
month, end of the third month?

308
00:12:24,976 --> 00:12:26,073
Still nothing.

309
00:12:26,073 --> 00:12:27,793
It's getting depressing, isn't it?

310
00:12:27,793 --> 00:12:30,071
We're doing all this work
month in and month out and

311
00:12:30,071 --> 00:12:30,935
nothing is happening.

312
00:12:30,935 --> 00:12:32,878
Now what about at the
end of the fourth month?

313
00:12:32,878 --> 00:12:36,838
Oh! Finally, I get a dot right here. Nope.

314
00:12:36,838 --> 00:12:38,819
What about that 20 percent tax?

315
00:12:38,819 --> 00:12:41,561
That multiplexing tax real
so if I had the multiplex

316
00:12:41,561 --> 00:12:43,767
from project to project, I
mean, do you do it every day,

317
00:12:43,767 --> 00:12:46,194
every hour? Gosh only knows.

318
00:12:46,194 --> 00:12:50,500
Is there a time, cost
and context switching?

319
00:12:50,500 --> 00:12:51,765
You know there is

320
00:12:51,765 --> 00:12:53,764
if you had to leave that
project and go to another one.

321
00:12:53,764 --> 00:12:57,014
So if it's 20 percent, maybe it's here.

322
00:12:58,193 --> 00:12:59,687
Holy cow.

323
00:12:59,687 --> 00:13:02,214
So this guy, even, passed this boundary.

324
00:13:02,214 --> 00:13:05,166
So they have all this value
being delivered in the serial

325
00:13:05,166 --> 00:13:07,999
case and now I reach this deadline

326
00:13:09,280 --> 00:13:11,548
and I haven't delivered anything yet.

327
00:13:11,548 --> 00:13:12,609
Well that's pretty painful.

328
00:13:12,609 --> 00:13:14,517
That's terrible economics.

329
00:13:14,517 --> 00:13:18,943
So simply, serializing the
work, limiting work in process,

330
00:13:18,943 --> 00:13:21,474
delivering incrementally,
is far better economics

331
00:13:21,474 --> 00:13:23,517
and it's a little bit worse than that.

332
00:13:23,517 --> 00:13:24,961
Let me tell you why.

333
00:13:24,961 --> 00:13:28,378
What happens to the requirements fidelity

334
00:13:33,424 --> 00:13:34,909
over the course of time?

335
00:13:34,909 --> 00:13:37,114
If I say, you know, I really
understand the requirements

336
00:13:37,114 --> 00:13:39,382
at this point, is that steady?

337
00:13:39,382 --> 00:13:42,412
It's absolutely flat or is
it more likely that each

338
00:13:42,412 --> 00:13:45,546
period of time, the market
changes a little bit,

339
00:13:45,546 --> 00:13:47,590
requirements change,
it decays a little bit.

340
00:13:47,590 --> 00:13:49,210
So let's just pretend,
for the sake of argument,

341
00:13:49,210 --> 00:13:53,377
that we have five percent
decay here on each of these.

342
00:13:54,259 --> 00:13:56,444
And if that's the case,
there's this tiny little delta.

343
00:13:56,444 --> 00:13:57,291
Well that's no big deal.

344
00:13:57,291 --> 00:13:58,748
I can fix it here pretty readily.

345
00:13:58,748 --> 00:13:59,847
It's just a quick correction.

346
00:13:59,847 --> 00:14:02,214
Wow, we got feedback at the
demo and they said we need to

347
00:14:02,214 --> 00:14:03,047
change that.

348
00:14:03,047 --> 00:14:04,482
We got feedback at the demo.

349
00:14:04,482 --> 00:14:05,581
Let's take the other case now.

350
00:14:05,581 --> 00:14:08,425
Let's take the case where
this five percent, we start

351
00:14:08,425 --> 00:14:11,132
on the parallel and we
start over here and assume

352
00:14:11,132 --> 00:14:13,779
that we have that same five
percent decay over time.

353
00:14:13,779 --> 00:14:17,946
Well, when I end up at the
end, I have, fifteen percent

354
00:14:22,177 --> 00:14:25,829
decay in requirements fidelity
so I have a much bigger

355
00:14:25,829 --> 00:14:27,912
correction to make later.

356
00:14:28,898 --> 00:14:31,472
Wow. What that means is
I need to think about my

357
00:14:31,472 --> 00:14:35,125
requirements as live and
changing and evolving.

358
00:14:35,125 --> 00:14:38,737
And I need to build a model
that allow me to determine

359
00:14:38,737 --> 00:14:42,084
the requirements at the last
responsible moment because

360
00:14:42,084 --> 00:14:45,495
here the last responsible moment
was the time when I started

361
00:14:45,495 --> 00:14:46,611
all projects.

362
00:14:46,611 --> 00:14:50,192
Here, the last responsible
moment is each time I started

363
00:14:50,192 --> 00:14:53,811
a new project, so it's even
worse than the raw math

364
00:14:53,811 --> 00:14:56,464
would indicate because of
the multiplexing cost and the

365
00:14:56,464 --> 00:14:58,165
fact that requirements fidelity kicks in.

366
00:14:58,165 --> 00:15:01,525
So they understand that in
manufacturing, Lean manufacturing

367
00:15:01,525 --> 00:15:04,189
and that's called just in time.

368
00:15:04,189 --> 00:15:09,129
Do not give me fifty little
widgets when I only need ten

369
00:15:09,129 --> 00:15:12,085
because by the time I go to
manufacture that, there's been

370
00:15:12,085 --> 00:15:14,737
an engineering change order
applied against that widget

371
00:15:14,737 --> 00:15:18,326
that basically says, that's
not exactly the widget anymore

372
00:15:18,326 --> 00:15:19,460
that you want to build with.

373
00:15:19,460 --> 00:15:21,503
So a really critical and important lesson.

374
00:15:21,503 --> 00:15:25,105
The basic economics of
agile development versus

375
00:15:25,105 --> 00:15:27,596
waterfall development is
not waterfall is kind of

376
00:15:27,596 --> 00:15:29,780
old-fashioned, sucky stuff
that doesn't really work

377
00:15:29,780 --> 00:15:31,195
and agile is hip and cool.

378
00:15:31,195 --> 00:15:33,889
It's that we deliver
incrementally and that's where the

379
00:15:33,889 --> 00:15:35,274
economics come from.

380
00:15:35,274 --> 00:15:37,515
If you like the economics
topic, you can dive into

381
00:15:37,515 --> 00:15:40,376
Reinertsen's work, "Principles
of Product Development Flow,"

382
00:15:40,376 --> 00:15:43,297
one of my favorite pieces and
you'll learn more about things

383
00:15:43,297 --> 00:15:46,038
that are even deeper
thoughts, sequence jobs for

384
00:15:46,038 --> 00:15:48,563
maximum benefit, we're
gonna talk about that.

385
00:15:48,563 --> 00:15:50,912
It's called weighted, shortest job first.

386
00:15:50,912 --> 00:15:53,371
Do not consider money already spent.

387
00:15:53,371 --> 00:15:56,893
Remember pivot without
mercy or guilt, same thing.

388
00:15:56,893 --> 00:15:59,276
We spent a million
dollars on this project.

389
00:15:59,276 --> 00:16:02,586
It's another million to complete
it but that same million

390
00:16:02,586 --> 00:16:05,317
dollars spent on a different
initiative could provide

391
00:16:05,317 --> 00:16:07,143
a higher overall cost.

392
00:16:07,143 --> 00:16:08,743
People are going, but
wait, we spent the million,

393
00:16:08,743 --> 00:16:10,110
we have to recover that.

394
00:16:10,110 --> 00:16:11,257
You can't recover it.

395
00:16:11,257 --> 00:16:14,216
Now if that million means
that the next million produces

396
00:16:14,216 --> 00:16:17,358
great value, compared to
another larger project, great,

397
00:16:17,358 --> 00:16:18,191
then you'll proceed.

398
00:16:18,191 --> 00:16:19,635
But if not you're gonna pivot.

399
00:16:19,635 --> 00:16:21,246
What did we learn about
our economic choices?

400
00:16:21,246 --> 00:16:22,604
It has to happen all the time.

401
00:16:22,604 --> 00:16:24,637
We have to build that
in kind of real time.

402
00:16:24,637 --> 00:16:28,167
Empower local decision making
and then one key Reinertsen

403
00:16:28,167 --> 00:16:30,236
takeaway and one that we're
gonna want to build on,

404
00:16:30,236 --> 00:16:34,152
if you only quantify one thing,
quantify the cost of delay.

405
00:16:34,152 --> 00:16:38,330
I wouldn't be much of an
instructor if I didn't put some

406
00:16:38,330 --> 00:16:40,741
real understanding behind
that key principal and I'll

407
00:16:40,741 --> 00:16:42,991
do that in a later module.