Episode 2: How Do You Know That's Actually the Problem?

Karen: 00:00:05

So I once got a call that the tech platform I worked for was down,

2

: 00:00:09

let's just call this person Jim, but he

was always a very challenging client.

3

: 00:00:14

He loved to raise his

voice, he loved to yell.

4

: 00:00:17

He was never happy.

5

: 00:00:19

You get it.

6

: 00:00:20

Anyway, he was ranting and raving and

screaming about how there was an outage

7

: 00:00:25

on our platform and the software was

totally down and nobody could access

8

: 00:00:29

it, and this was incredibly urgent.

9

: 00:00:31

And how is anyone on his team

supposed to get their job done when

10

: 00:00:35

no one can log in, blah, blah, blah.

11

: 00:00:37

I, I would normally take this very

seriously, but I'm looking around the

12

: 00:00:41

office and no one else seems to be aware

of or concerned about a so-called outage,

13

: 00:00:46

and in fact, many of them are actively

using the platform right in front of me.

14

: 00:00:50

So I started asking Jim more questions

about what was actually going on, and

15

: 00:00:54

what came to light is that two people on

his team, two people total, had forgotten

16

: 00:00:59

their passwords and couldn't log in.

17

: 00:01:01

And Jim, with all the strategic and strong

leadership skills, his title implied

18

: 00:01:06

that he possessed, decided to translate.

19

: 00:01:09

Two people can't log to the platform

is down without pausing to see if

20

: 00:01:14

those were actually the same thing.

21

: 00:01:17

When I pointed that out to him, he

said to me, and this is an actual

22

: 00:01:21

quote that an actual person said

to me, as far as I'm concerned, if

23

: 00:01:27

someone on my team can't log into the

system, it's the same as an outage.

24

: 00:01:32

Uh, no, it's not the same as an outage

because we're contractually obligated

25

: 00:01:37

to keep the platform up and running.

26

: 00:01:39

So I care a lot about an outage, but we're

not obligated in the slightest to help

27

: 00:01:44

Pam and Donna remember their passwords.

28

: 00:01:47

It's kind of like how my mom was told me

she had an emergency and I stepped out

29

: 00:01:51

of a meeting to call her back, and she

was freaking out that the computer didn't

30

: 00:01:55

work and she had very important things

to do and she couldn't type anything.

31

: 00:02:00

And after a little more probing,

we realized that, the batteries had

32

: 00:02:03

expired in her Bluetooth keyboard.

33

: 00:02:06

. So obviously that story might seem

like it's about a forgotten password

34

: 00:02:11

or maybe even just an overreaction

from a challenging personality, but

35

: 00:02:15

it's really about something bigger and

something that I think happens constantly

36

: 00:02:20

in organizations, in careers, in the

decisions we make about our own lives.

37

: 00:02:25

We are so quick to move from symptom

to conclusion without stopping to

38

: 00:02:30

ask the most important follow-up

question after what's wrong.

39

: 00:02:34

Which is, how do you

know that's what's wrong?

40

: 00:02:38

I am Karen Doak.

41

: 00:02:39

This is okay.

42

: 00:02:40

Actually, the show where we get clear, get

sorted, get going, and stay sane together.

43

: 00:02:48

Some of this is about diagnosing

before you build, but I spent decades

44

: 00:02:52

watching people solve the wrong

things with a whole lot of confidence.

45

: 00:02:57

The architecture only works

when the foundation is right,

46

: 00:03:00

and that's what we do here.

47

: 00:03:01

We really try to find the foundation.

48

: 00:03:04

I was trying to think about the

right language to use to describe the

49

: 00:03:08

process because in software, the term

you come back to a lot is root cause

50

: 00:03:12

analysis, but I ultimately felt it was

too subjective and I wanted to make

51

: 00:03:17

sure we were removing subjectivity

entirely from this conversation.

52

: 00:03:21

So I think the best term to

describe what I'm talking about

53

: 00:03:24

is the idea of ground truth.

54

: 00:03:28

I was familiar with the term from

environmental surveying and geology.

55

: 00:03:32

The real definition of ground

truth is information that's

56

: 00:03:36

known to be real or true.

57

: 00:03:38

Coming from things you're directly

observing or measuring versus

58

: 00:03:42

information that might be inferred.

59

: 00:03:46

It's not a feeling of clarity.

60

: 00:03:47

It's not a, I sat with

this and now I feel better.

61

: 00:03:52

Ground truth is verified.

62

: 00:03:54

It's a load-bearing fact

underneath everything else.

63

: 00:03:58

It's not the story, it's not

the presenting symptom, and

64

: 00:04:02

it's not what you wish was true.

65

: 00:04:05

I was watching the BBC series Vera,

and first of all, no spoilers.

66

: 00:04:10

Uh, although I don't think spoiling

a: 2011

67

: 00:04:14

available on Brit box is like a thing,

but still, I promise no spoilers.

68

: 00:04:20

In the first season, there's an

episode with an environmental

69

: 00:04:23

surveyor who uses the phrase ground

truth in a completely literal sense.

70

: 00:04:29

She's talking about excavation,

the verified condition of what's

71

: 00:04:33

actually happening underneath

before you can build on it.

72

: 00:04:37

Not assumed, not estimated,

confirmed on site.

73

: 00:04:43

And I stopped the episode because that's

exactly what I've been trying to name.

74

: 00:04:48

Obviously in her case, the

ground she was evaluating

75

: 00:04:51

ultimately held some dead bodies.

76

: 00:04:53

And I don't think that's a spoiler either,

because this is a murder mystery series.

77

: 00:04:58

Um, and frankly, your ground truth

might also contain some dead bodies.

78

: 00:05:03

Meta metaphorically, of

course, where I hope, I hope.

79

: 00:05:08

Ground truth isn't when

you feel confident.

80

: 00:05:11

It's when you've tested the

thing enough to build on it.

81

: 00:05:13

And those are not the same moment.

82

: 00:05:16

I like to call the process plum with

a B, but when you think about the

83

: 00:05:19

plum line, that's the instrument

builders have been using since ancient

84

: 00:05:23

Egypt to establish what's actually

vertical before construction begins.

85

: 00:05:28

It's a weight.

86

: 00:05:29

It uses gravity in one string

to tell you what you can build

87

: 00:05:32

on and what's actually true.

88

: 00:05:34

Not advanced technology, but certainly

reliable enough for the Pharaohs.

89

: 00:05:39

I was at a software company where 90%

of our revenue and customer churn was

90

: 00:05:44

attributed to product issues, and yet

we weren't making any improvements to

91

: 00:05:48

the product, and we were still being

asked why we weren't improving churn.

92

: 00:05:53

Everything else that we

were doing is essentially a

93

: 00:05:55

bandaid on the wrong problem.

94

: 00:05:57

Instead of really building what works.

95

: 00:06:01

We would have meetings to review

the data and it was crystal clear,

96

: 00:06:05

except the product team didn't have

any part of its success tied to

97

: 00:06:08

retention metrics and my team was

100% accountable for churn and had no

98

: 00:06:14

ability to actually fix the product.

99

: 00:06:17

We knew everyone on the call

knew, and that was the thing.

100

: 00:06:21

The ground truth wasn't hidden.

101

: 00:06:23

It was sitting in the feedback

data plain as day, but the

102

: 00:06:27

product wasn't going to get fixed.

103

: 00:06:29

And so everything we did on the

support and customer success side

104

: 00:06:32

was us trying to solve a problem.

105

: 00:06:34

We didn't have the ability to solve.

106

: 00:06:36

We were running root cause analysis on

a situation where we already had the

107

: 00:06:40

root cause and nobody wanted to hear it.

108

: 00:06:43

And that's its own particular kind of

stuck because at least when you haven't

109

: 00:06:48

found ground truth yet, you have the

search to occupy and entertain you when

110

: 00:06:52

you have it and you can't act on it.

111

: 00:06:54

You're just watching these

band-aids slowly peel off and float

112

: 00:06:59

in the depressing pool of doom.

113

: 00:07:02

I wanna walk you through the

four questions I use to test

114

: 00:07:06

whether I've actually gotten to

ground truth or whether I'm just

115

: 00:07:09

comfortable with where I've landed.

116

: 00:07:12

You don't have to use all four

every time, but I'd start with

117

: 00:07:15

the first one and see how far you

get before the answer stops you.

118

: 00:07:19

The four questions are,

does it change shape?

119

: 00:07:22

Can you break it?

120

: 00:07:24

Is this the source or the signal?

121

: 00:07:26

And does it hold alone?

122

: 00:07:28

I'm gonna walk through each one and

I'll put all four in the show notes

123

: 00:07:31

so you have them and don't have to

take notes in case you were okay.

124

: 00:07:36

Question one.

125

: 00:07:38

Does it change shape?

126

: 00:07:39

Does your explanation

about the problem change?

127

: 00:07:42

Depending on who's asking?

128

: 00:07:44

Are you going to describe the

situation differently to your partner

129

: 00:07:48

versus your coach versus your boss?

130

: 00:07:51

Real ground truth doesn't have

an audience dependent version.

131

: 00:07:55

It just is, and the fact that

you're editing it is extremely

132

: 00:07:59

important information for you.

133

: 00:08:02

As a personal example, my youngest is

graduating college next month, and I've

134

: 00:08:06

been incredibly stressed about whether

she has a plan, whether she's gonna

135

: 00:08:11

find a job, whether she's gonna be okay.

136

: 00:08:14

And I can tell you with complete sincerity

that if you ask me what I'm worried about,

137

: 00:08:18

I will say her future, her trajectory,

whether she's gonna be set up for success.

138

: 00:08:23

But if I'm being honest, if I'm running

the ground truth test on my own anxiety.

139

: 00:08:28

A big portion of what's actually driving

the stress is that I'm worried if

140

: 00:08:32

she moves home, she will never leave.

141

: 00:08:34

And that's the real thing.

142

: 00:08:35

And I would not say that to her.

143

: 00:08:37

I would absolutely not say that to her.

144

: 00:08:39

Which means my explanation changes

completely depending on the audience,

145

: 00:08:44

which means by definition I'm

not talking about ground truth.

146

: 00:08:48

I found the version I'm

comfortable saying out loud.

147

: 00:08:51

So that's the audience test.

148

: 00:08:54

But even when you're being honest

about what the problem is, there's

149

: 00:08:58

a second question worth asking, and

it's one that gets at whether your

150

: 00:09:01

diagnosis is actually testable.

151

: 00:09:05

Question two is can you break it?

152

: 00:09:08

Can you actually poke holes

in your own conclusion?

153

: 00:09:11

Can you find the scenario

where you're wrong?

154

: 00:09:15

And this is the one that I think

separates a real diagnosis from

155

: 00:09:18

just, you know, a confident

declaration, a diagnosis is testable.

156

: 00:09:23

It has conditions under which

you, you know, you were wrong.

157

: 00:09:26

If someone asked my favorite client,

Jim, what would really make him

158

: 00:09:29

reconsider whether this is an outage?

159

: 00:09:32

He had no answer because

he wasn't diagnosing.

160

: 00:09:34

He was just, you know, making statements

fuming and ranting a real diagnosis.

161

: 00:09:40

Sounds like, I think it's x.

162

: 00:09:42

But if I found out this other information,

I'd look somewhere else first.

163

: 00:09:48

If you can't name what would change

your mind, then you started with the

164

: 00:09:52

conclusion and you built backwards.

165

: 00:09:54

It's like the problem

we all have with WebMD.

166

: 00:09:58

You type in your symptoms and

it tells you it might be cancer.

167

: 00:10:01

Not because that's a diagnosis, but

because cancer matches the symptom

168

: 00:10:05

list and WebMD has no mechanism for

testing whether it's actually true.

169

: 00:10:10

Nobody asks what would have to

be false for this to be cancer.

170

: 00:10:13

It just serves up results.

171

: 00:10:15

We see cancer, we spiral.

172

: 00:10:18

Or for a more personal example, I talked

last week about a job situation that

173

: 00:10:23

wasn't working and for a long time my

diagnosis was exactly that this job isn't

174

: 00:10:27

working, which sounds like a diagnosis.

175

: 00:10:30

But it's actually just a conclusion

I made with nowhere to go.

176

: 00:10:34

So I ran this test on myself.

177

: 00:10:36

I asked what would have to be true

for this job to actually work.

178

: 00:10:40

And when I started writing everything

down, what I realized is that every

179

: 00:10:44

single thing on the list required my

boss to operate in a fundamentally

180

: 00:10:49

different way or for some other major

business changes to happen, all of

181

: 00:10:54

which I had no ability to change.

182

: 00:10:56

So the real diagnosis wasn't.

183

: 00:10:58

This job isn't working, it's

this job will never work.

184

: 00:11:02

And those might sound almost

identical, but they're not because

185

: 00:11:06

one suggests there's a fix.

186

: 00:11:08

One has that glimmer of optimism

and the other tells you what

187

: 00:11:13

you're actually deciding.

188

: 00:11:17

You can't get to that clarity without

asking what would have to be false

189

: 00:11:21

and being honest about the answer.

190

: 00:11:23

You have to be able to distinguish

between a true diagnosis and a story

191

: 00:11:28

that has good posture, but isn't

necessarily true and related to that

192

: 00:11:33

sometimes the issue isn't that we drew

the wrong conclusion, it's that we were

193

: 00:11:37

looking in the wrong place entirely.

194

: 00:11:40

Which brings us to question three.

195

: 00:11:43

Is this the source or the signal?

196

: 00:11:46

Symptoms present so loudly and

causes sit quietly underneath that

197

: 00:11:51

forgotten password made noise.

198

: 00:11:53

Or at least Jim.

199

: 00:11:55

Jim made a lot of noise,

but the platform was fine.

200

: 00:11:58

So you have to ask yourself, if I

solve this exact thing, would the

201

: 00:12:03

underlying situation actually change?

202

: 00:12:05

And if the answer is not really.

203

: 00:12:08

You are treating a symptom which matters

because symptoms have consequences and

204

: 00:12:12

symptoms can be painful, but it's not

the same thing as finding the source.

205

: 00:12:18

My toy poodle pepper is 14.

206

: 00:12:20

She's my soulmate, and in my objective

opinion, she's perfect except for all of

207

: 00:12:26

the terrible things she does every single

day, but she's also 14, which means that

208

: 00:12:32

on a fairly regular basis, she presents

with some new and alarming symptom.

209

: 00:12:36

That sends me on a spiral.

210

: 00:12:38

There's a limp.

211

: 00:12:39

There's a lump.

212

: 00:12:40

She's breathing funny.

213

: 00:12:41

She's snoring louder.

214

: 00:12:43

She seems lethargic, and I'm

convinced something is deeply wrong.

215

: 00:12:47

I'm Googling, I'm asking Claude.

216

: 00:12:50

I am treating whatever the thing is

that's visible to me as the problem

217

: 00:12:54

to be solved, and my extremely patient

vet has said to me so many times in

218

: 00:12:59

the kindest way possible, she's 14.

219

: 00:13:03

That's it.

220

: 00:13:04

That's the whole answer.

221

: 00:13:06

The limp is real and the symptom is

real, but the source isn't the limp.

222

: 00:13:10

The sources that she's 14 and I

can't fix the passing of time.

223

: 00:13:16

I can manage it.

224

: 00:13:17

I can make her comfortable.

225

: 00:13:19

I can go to the vet at the first

sign of anything, and obviously

226

: 00:13:23

I'll continue to do that.

227

: 00:13:25

I can spend a little more time

than is probably rational or

228

: 00:13:29

healthy on the website of Tom

Brady's pet cloning company.

229

: 00:13:34

But I'm not addressing the

problem in any of that.

230

: 00:13:36

I'm managing the reality and

knowing that actually knowing it

231

: 00:13:41

not just intellectually changes

what you're asking of yourself.

232

: 00:13:45

You're not failing to fix something.

233

: 00:13:47

You're just doing what's

actually possible.

234

: 00:13:51

I think this is probably

the most common fail point.

235

: 00:13:54

So often we jump to what we can

see for that very obvious reason,

236

: 00:14:00

and we need more perspectives

to point out what's not visible.

237

: 00:14:03

If you don't have those outside

perspectives at the ready, just try for

238

: 00:14:07

a minute to put on the hat of someone

who doesn't have your vested interest

239

: 00:14:11

In the current explanation, would

they answer the question the same way?

240

: 00:14:17

So those first three questions are really

about the problem, but I think the last

241

: 00:14:21

one is more about you and as a result,

it might be the most important one.

242

: 00:14:27

Question four is, does it hold alone?

243

: 00:14:30

Does this still feel true

when no one's watching?

244

: 00:14:33

And nothing needs to sound good when

you're alone, when you're writing in

245

: 00:14:38

your journal, when it's two in the

morning and you wake up with a panic

246

: 00:14:41

attack, does this still feel true?

247

: 00:14:45

Because at the end of the workday when

I have a glass of wine in hand and I'm

248

: 00:14:49

making dinner with my husband, Jeff,

I'm revved up while I'm storytelling.

249

: 00:14:54

I'm on a roll.

250

: 00:14:55

My justice complex and my need

for attention are doing all the

251

: 00:14:59

driving, but that's not ground truth.

252

: 00:15:02

That's just a great story.

253

: 00:15:04

I'll be mid-sentence, making a point

that seems airtight to Karen, the dinner

254

: 00:15:08

prep performer, but something underneath

me will go, is that actually right?

255

: 00:15:13

And that's the question

worth stopping for.

256

: 00:15:16

If you're trying to get to ground truth,

it's absolutely worth stopping for.

257

: 00:15:21

If you're really trying to solve

a problem, you cannot ignore

258

: 00:15:23

that feeling or that resistance

that's kind of holding you back.

259

: 00:15:28

None of these are necessarily

comfortable tests, and they ask you

260

: 00:15:32

to be really honest with yourself.

261

: 00:15:34

That's the point though, because if

ground truth was easy to surface,

262

: 00:15:38

you'd have already found it.

263

: 00:15:40

The discomfort is not a sign

that you're doing it wrong.

264

: 00:15:42

It's usually a sign that

you're getting closer.

265

: 00:15:45

So just to recap, the four questions

which you can find in the show

266

: 00:15:48

notes are, does it change shape?

267

: 00:15:51

Can you break it?

268

: 00:15:52

Is this the source or the signal?

269

: 00:15:54

And does it hold alone?

270

: 00:15:57

So let's say you've run

through those questions.

271

: 00:15:58

Let's say you sat with the discomfort,

you found it, you found the thing that's

272

: 00:16:03

actually true, not the thing you've

been performing or saying is true.

273

: 00:16:08

Here's what I want you to

know about that moment.

274

: 00:16:11

Ground truth doesn't feel triumphant.

275

: 00:16:14

It can actually feel very

deflating because now you

276

: 00:16:17

have to reckon with the gap.

277

: 00:16:19

The gap between where you've

been spending all your energy and

278

: 00:16:23

where the actual problem lives.

279

: 00:16:25

But it's also the only place that

architecture and solutions can

280

: 00:16:28

begin anything else Is a bandaid

on the wrong problem instead

281

: 00:16:32

of really building what works.

282

: 00:16:34

You can't skip just to the good part.

283

: 00:16:37

When clients of mine get to ground

truth for the first time, they're

284

: 00:16:40

not necessarily doing a jig.

285

: 00:16:42

They often go quiet because it

means they've been working on

286

: 00:16:45

something that wasn't right, and

that can be very discouraging.

287

: 00:16:49

And now they have to start somewhere new.

288

: 00:16:51

But that quiet, that is the work and

it's starting that, that's gravity

289

: 00:16:56

on the plumb line doing this job.

290

: 00:16:59

And I'm not gonna pretend that's easy.

291

: 00:17:02

Starting again when you thought

you were already underway.

292

: 00:17:04

It takes something, but the

alternative is staying busy on the

293

: 00:17:07

wrong thing, and that's really,

really expensive misdirection.

294

: 00:17:13

I think there's ultimately something

so freeing and saying, okay, actually

295

: 00:17:18

I'm gonna start spending my time

now on the thing that matters.

296

: 00:17:22

Now.

297

: 00:17:23

At least you know you're not doing

a bunch of performative bullshit.

298

: 00:17:26

You're at least getting

to the heart of it.

299

: 00:17:29

Or maybe you're not ready to work

on it or solve it, but at least

300

: 00:17:31

you're being honest about what's not

working and you can go from there.

301

: 00:17:35

Deciding to put something in a box

and not deal with a problem is not

302

: 00:17:39

necessarily the worst thing in the world.

303

: 00:17:42

Making sure you've actually gotten

clear on what the problem is before

304

: 00:17:46

you decide to ignore it, at least

then you're being honest with half

305

: 00:17:49

of it, and that's both a step forward

and a step in the right direction.

306

: 00:17:55

Before I let you go this week, I

would love it if you could take one of

307

: 00:17:59

those four questions and apply it to

something you're currently working on.

308

: 00:18:02

I'd like to know what you're thinking.

309

: 00:18:03

You can find my email in the

show notes, but please reach out.

310

: 00:18:07

My hope for everyone is that all

of your platform outages turn,

311

: 00:18:10

turn out just to be forgotten.

312

: 00:18:12

Passwords, I'm Karen Doak This is okay.

313

: 00:18:16

Actually.

314

: 00:18:17

Hopefully we're getting clear

because that's the focus of

315

: 00:18:20

ground truth, so that we can get

sorted, get going, and stay sane.