207: Watch Me Do a Data Analyst Project in Minutes With Claude Code

Speaker: 00:00:00

A lot of aspiring data analysts ask

me, is AI going to take our jobs?

2

: 00:00:04

Are we cooked?

3

: 00:00:05

And my honest answer is,

let's watch and find out.

4

: 00:00:09

So I did that exactly in this episode.

5

: 00:00:11

I gave Claude code a real data

analyst project, the kind of analysis

6

: 00:00:15

that I used to spend hours on, and

I just kind of let a do its thing.

7

: 00:00:19

Here's what happens.

8

: 00:00:20

So here's the project.

9

: 00:00:21

I want to get some insights

on my YouTube channel.

10

: 00:00:24

What videos are doing well and why

they're doing well, so I can try

11

: 00:00:27

to make better performing videos in

the future and more helpful videos

12

: 00:00:31

for you guys who are watching, or

the you guys who are listening.

13

: 00:00:33

So what I did is I opened up

YouTube creator studio and

14

: 00:00:37

filtered by the last 365 days.

15

: 00:00:40

And got different things like the

content data, the traffic source

16

: 00:00:43

data, the geography data, and I

just exported all those as CSVs

17

: 00:00:48

straight to this folder here.

18

: 00:00:51

And so you can see I have a content

geography, new and returning viewers,

19

: 00:00:55

playlist posts, subscription status,

traffic source, and viewer age, zip

20

: 00:00:59

folders right here on my desktop.

21

: 00:01:02

And basically I wanna see what's

going well with my channel and what's

22

: 00:01:05

not doing so well with my channel.

23

: 00:01:07

If I were doing this analysis myself or

giving it to another fellow data analyst,

24

: 00:01:11

I would just kind of hope that they'd

go through all the data, explore it,

25

: 00:01:14

and generate some meaningful insights.

26

: 00:01:15

If they had prior domain knowledge

about YouTube videos, they might

27

: 00:01:18

know to look through, click-through

rates on the thumbnails and average

28

: 00:01:21

view durations of certain videos.

29

: 00:01:22

But if I gave it to just a normal data

analyst, they probably wouldn't have that

30

: 00:01:26

domain knowledge of YouTube, and so they

wouldn't know where to look necessarily.

31

: 00:01:29

In this case, I'm not going to use my

domain knowledge to give Claude any hints.

32

: 00:01:33

I just wanna see what it

comes up with on its own.

33

: 00:01:35

So let's go ahead and open up Claude.

34

: 00:01:36

So if you haven't checked out

Claude yet, I highly recommend it.

35

: 00:01:39

It's produced by a company called

Anthropic that's very similar

36

: 00:01:42

competitor to OpenAI, and Claude is

basically a competitor to chat GPT.

37

: 00:01:47

Now, Claude has gained a lot of popularity

in the last, I don't know, six months,

38

: 00:01:50

because one, it can write really well.

39

: 00:01:53

Two, it can code really well.

40

: 00:01:55

Andro has basically decided that it

doesn't want to be good at everything.

41

: 00:01:58

It just wants to be good at

writing and good at coding.

42

: 00:02:01

And here's the truth is when

you're good at writing and good

43

: 00:02:03

at coding, you're very powerful.

44

: 00:02:05

So they have this chat interface

that's very similar to Chat GPT,

45

: 00:02:08

but they also have this thing

called Claude Code right here.

46

: 00:02:11

Now, Claude Code is

basically the coding version.

47

: 00:02:14

Of Claude, it's more designated

for getting tasks done,

48

: 00:02:16

specifically coding tasks.

49

: 00:02:18

And you can use the CLI, which is

basically the command line interface.

50

: 00:02:22

Basically, that would look like you

opening up your command prompt and

51

: 00:02:25

then typing in Claude right here.

52

: 00:02:30

And then boom, Claude Code pops up here.

53

: 00:02:32

So you can have like a command line

terminal version of Claude Code.

54

: 00:02:36

Or you can just use the desktop app,

which has Claude Code right here.

55

: 00:02:40

Now, personally, I like using the desktop

app because terminals and command lines

56

: 00:02:44

still make me a little bit scared, and

it's just harder to know what's going on.

57

: 00:02:47

So I'll be using the desktop

version of Claude Code.

58

: 00:02:49

Now, I will say that I pay a

hundred dollars a month for

59

: 00:02:51

Claude's Max subscription plan.

60

: 00:02:53

Basically, that means I have the

very powerful version of Claude and

61

: 00:02:56

I can use their most powerful model.

62

: 00:02:58

Opus 4.6,

63

: 00:03:00

which is basically just their

smartest, most sophisticated,

64

: 00:03:02

best model that they have.

65

: 00:03:04

They've also recently updated

it to have a 1 million context

66

: 00:03:08

window, which is very powerful.

67

: 00:03:09

Basically, it allows you to have

a lot of context, which can be

68

: 00:03:13

good if you have a lot of data.

69

: 00:03:14

Now, if you were to change models

from Opus to something like sonnet,

70

: 00:03:17

this is their less powerful model.

71

: 00:03:19

Now, sauna is still really good, but if

you were to change it to Haiku, which

72

: 00:03:22

is really just their weakest model.

73

: 00:03:24

It probably wouldn't do as well.

74

: 00:03:26

So depending on what model you're using,

the performance will definitely change.

75

: 00:03:30

The results will definitely change.

76

: 00:03:31

So I'm just gonna use the most

powerful model just to see

77

: 00:03:33

what we're working with here.

78

: 00:03:34

Now, recently when I'm using AI

tools, uh, I have really gotten

79

: 00:03:37

into dictation, and the reason

being is I kind of suck at writing.

80

: 00:03:41

I kind of suck at typing, and if

I can just brain dump my thoughts

81

: 00:03:45

verbally, I find that to be a lot

more effective than me filtering my

82

: 00:03:49

thoughts via my typing with my fingers.

83

: 00:03:51

So I'll just go ahead and click

the record button right here.

84

: 00:03:54

In this folder, you'll find a set

of zip folders that contains the

85

: 00:04:01

analytics for my YouTube channel

directly exported from YouTube Studio.

86

: 00:04:06

I'd like you to analyze these data

sets and provide me meaningful,

87

: 00:04:10

actionable insights on what I

could be doing better on my YouTube

88

: 00:04:14

channel and what I'm doing well.

89

: 00:04:18

And that's the prompt.

90

: 00:04:19

I'm literally going to give Claude,

and I'm just gonna press this go

91

: 00:04:21

button and see what it comes up with.

92

: 00:04:24

It's gonna ask if it trusts this

workspace, and it indeed does.

93

: 00:04:27

And you'll notice that I

have that folder right here.

94

: 00:04:29

It's literally attached

to this folder right here.

95

: 00:04:32

And that's the thing is cloud

code isn't working in the cloud.

96

: 00:04:35

It's working here locally on your machine.

97

: 00:04:37

So literally it has this YouTube

data analysis folder, the

98

: 00:04:40

contents of which are inside.

99

: 00:04:42

Now, I'm kind of curious.

100

: 00:04:44

It's gonna ask me a lot

of permission things.

101

: 00:04:45

I'm gonna literally just do, uh,

allow always for this project

102

: 00:04:48

for right now, uh, I'm curious if

it's gonna be able to unzip these.

103

: 00:04:51

It should be able to.

104

: 00:04:53

Um, but we'll see what it actually

does to, to make this go through.

105

: 00:04:56

So it's just like, Hey, do you wanna make

sure that I can run all these commands and

106

: 00:04:59

I'm gonna let it run all these commands

because I, uh, trust it for right now.

107

: 00:05:03

Now you obviously probably wanna

be a little bit more careful than

108

: 00:05:05

me when allowing these AI agents

to do things on your desktop.

109

: 00:05:09

You probably wanna make

sure you know that.

110

: 00:05:11

What's actually going on.

111

: 00:05:12

But to be honest, I'm a little

bit lazy and I just, I trust her.

112

: 00:05:15

So far so good.

113

: 00:05:16

Obviously there's been a lot of issues.

114

: 00:05:17

I know that there's been one machine

learning content creator who like deleted

115

: 00:05:21

an entire database completely with cloud

code, so you, you gotta be careful and

116

: 00:05:26

make sure you know what it's doing.

117

: 00:05:27

So right now it basically looks

like it created a user memory for

118

: 00:05:31

me and it's compiling the analysis.

119

: 00:05:33

It, read all the key files,

understand the picture.

120

: 00:05:36

It's running these bash commands.

121

: 00:05:37

It looks looking like it's

getting the CSV files.

122

: 00:05:40

And this is, yeah, look, this

is what's in the CSV file.

123

: 00:05:43

You can kinda see one of my titles,

I'd become a data analyst in: 2026

124

: 00:05:46

Then you're seeing some of the numbers,

views on a certain date or something like

125

: 00:05:49

that, and you can see that it did that

with, um, post data, the content data,

126

: 00:05:53

the playlist data, so on and so forth.

127

: 00:05:55

So you can kind of see.

128

: 00:05:56

What it's thinking about, let's

see what it's doing right now.

129

: 00:05:59

Um, it, it just created like a, a file,

a memory file for me about, you know,

130

: 00:06:03

who I am and what this project is about.

131

: 00:06:05

And it says Avery Smith is a, oh, okay.

132

: 00:06:08

Let's see what it, it

came up with some stuff.

133

: 00:06:10

I don't know what it's doing.

134

: 00:06:10

Ah, okay.

135

: 00:06:11

Here's my full analysis

of your YouTube channel.

136

: 00:06:12

Over the past year, we've had 1.7

137

: 00:06:14

million views, a hundred thousand

watch hours, 29,000 subscribers

138

: 00:06:18

gained about $8,000 in ad revenue, 30

million impressions, a click through

139

: 00:06:24

rate of 3% average view, duration of.

140

: 00:06:26

Four minutes and 442

videos in the dataset.

141

: 00:06:29

What we're doing, well, massive reach

with how to become a data analyst content.

142

: 00:06:33

Your top two videos alone account for

22% of the total, um, which is how I'd

143

: 00:06:36

become a data analyst in 2026 and how

become a data data analyst in: 2026

144

: 00:06:40

I basically just changed.

145

: 00:06:41

The 2025 video to 2026.

146

: 00:06:43

Uh, let's editor, let's,

let's rule that out.

147

: 00:06:45

If you're wondering why those video

titles are the same, it's because one

148

: 00:06:48

of 'em is actually for 2025, not for

: 2026

149

: 00:06:53

purposes, that it would be normal.

150

: 00:06:55

That way people could actually

find the video still and be

151

: 00:06:57

interested in watching it.

152

: 00:06:58

'cause although some things have changed

from 20 25, 20 26, a lot remains the same.

153

: 00:07:02

This also drove 12,000 subscribers,

which is 41% of my subscribers

154

: 00:07:06

gained two strong subscriber

conversion on career advice video.

155

: 00:07:09

Videos like I analyzed 2,893 Data

Analyst jobs don't miss these six

156

: 00:07:14

entry data levels jobs, and nine huge

lies about becoming Data analysts

157

: 00:07:17

have excellent sub to view ratios.

158

: 00:07:19

Your audience converts to

subscribers when you deliver data

159

: 00:07:22

backed actionable career insights.

160

: 00:07:24

All right, let's get to know Number three.

161

: 00:07:25

Excellent watch time on long

form interview content, your

162

: 00:07:28

podcast slash interview video.

163

: 00:07:30

So I, I like that it.

164

: 00:07:31

Figured out that I have like podcasty

style videos and interview style

165

: 00:07:34

videos versus just like YouTube videos.

166

: 00:07:36

I think that's really interesting

that it already gleaned that

167

: 00:07:38

'cause I didn't tell it that.

168

: 00:07:39

Right.

169

: 00:07:40

Um, let's see How to earn 250,000

with multiple remote data jobs.

170

: 00:07:44

13,000 views, but 2000

hours of watch time.

171

: 00:07:48

Data analyst mock interview, 25,000

views,: 2000

172

: 00:07:52

time, and an 8% click-through rate.

173

: 00:07:54

Wow.

174

: 00:07:54

See, that's really interesting.

175

: 00:07:55

Playlist driven watch time is

strong in your data career.

176

: 00:07:57

Uh, podcast playlist alone drives

658 hours of playlist watch time.

177

: 00:08:02

It's just listing a bunch of things.

178

: 00:08:03

I don't know.

179

: 00:08:04

Some of it's actionable.

180

: 00:08:05

Uh, we'll see if it suggests things.

181

: 00:08:07

Um, audience demographics,

basically 25 to 44.

182

: 00:08:10

Mostly us, India, uk, and Canada.

183

: 00:08:13

Um.

184

: 00:08:15

Browse features and suggested videos

are your top engagement sources.

185

: 00:08:19

By watch time, this actually means YouTube

is actively recommending your content.

186

: 00:08:23

The algorithm likes you.

187

: 00:08:24

Oh, I'm glad to hear that.

188

: 00:08:25

Sometimes it feels like it doesn't.

189

: 00:08:27

Alright, what needs improvement?

190

: 00:08:28

My CTR is below average.

191

: 00:08:30

YouTube considers four to 5% healthy CTR.

192

: 00:08:33

You're at 3% overall.

193

: 00:08:35

Some of your biggest videos in the 2.4,

194

: 00:08:37

2.8%,

195

: 00:08:38

this means your thumbnails

and titles are underperforming

196

: 00:08:40

relative to your impressions.

197

: 00:08:42

Your highest.

198

: 00:08:43

CTR videos give you

clues about what works.

199

: 00:08:46

Okay.

200

: 00:08:47

Then it says How to export

Tableau dashboard to image 15%.

201

: 00:08:51

So this is interesting because me having

those domain knowledge and actually

202

: 00:08:56

like being a YouTube content creator

and knowing YouTube, I know that this

203

: 00:09:00

is bad advice because let's, let's go.

204

: 00:09:03

In fact, let's go to how to

exports Tableau dashboard to image.

205

: 00:09:06

And if you actually just go to

Google and you search that I'm gonna

206

: 00:09:09

be the first video that pops up.

207

: 00:09:11

Right, and so when you're

ranking number one on Google.

208

: 00:09:15

You're going to have a high click

through rate because most of this

209

: 00:09:19

video's traffic comes via search,

not via browse or suggested videos.

210

: 00:09:24

So with search, you're like already

know what you really want, right?

211

: 00:09:28

You're like, this is what I want.

212

: 00:09:29

And when someone's like, Hey, here

you go, you're gonna click on it.

213

: 00:09:32

So that's why you have a high.

214

: 00:09:33

Click through rate versus if you're just

like on, you know, YouTube in general,

215

: 00:09:38

just on the like, uh, homepage, right?

216

: 00:09:41

Mm-hmm.

217

: 00:09:41

Right here, it's like, I

don't know what I want.

218

: 00:09:43

Do I wanna watch the, this is a

new chapter, or how I built a lead

219

: 00:09:47

magnet or, um, what is Databricks?

220

: 00:09:50

Those types of things.

221

: 00:09:51

Right?

222

: 00:09:51

Oh, editor, let's redo that

because I don't want 'em to see

223

: 00:09:53

that I'm watching this video.

224

: 00:09:54

When you open up YouTube, it's like.

225

: 00:09:56

What do you wanna watch?

226

: 00:09:57

Do I wanna watch you know,

data with Barr right now?

227

: 00:09:59

Do I wanna watch Mark Lowe?

228

: 00:10:00

Do I watch Tom Scott?

229

: 00:10:01

What am I in the mood for?

230

: 00:10:03

Maybe I'm gonna play

this Burger Life game.

231

: 00:10:04

There's just a lot more distractions and

earning the click here is a lot harder

232

: 00:10:08

to do than if you're like, you know.

233

: 00:10:11

How to export to Tableau

dashboard to image.

234

: 00:10:13

Oh, okay.

235

: 00:10:14

I'm number one.

236

: 00:10:15

Click on Google for, or

YouTube for this click.

237

: 00:10:17

Like, that's the reason why that

click through rate is high is

238

: 00:10:20

because it's mostly search base

versus browse or the homepage.

239

: 00:10:23

Um, same with these videos.

240

: 00:10:25

These are all going to be,

um, higher search videos.

241

: 00:10:29

This one's not, which is interesting.

242

: 00:10:31

Um, I'll admit, becoming a data

analysis is sustainable right now.

243

: 00:10:34

Um, so that ones may be

interesting to look into.

244

: 00:10:37

Titles with specificity,

controversy, or clear utility.

245

: 00:10:40

Get clicked more.

246

: 00:10:41

You're brought how to

become a data analyst.

247

: 00:10:42

Titles get massive impressions, but lower

CTR consider AB testing thumbnails on your

248

: 00:10:46

top impression videos, even with a 0.5%

249

: 00:10:49

improvement.

250

: 00:10:50

That's 30,000 more views.

251

: 00:10:52

View your retention is short.

252

: 00:10:53

Yes, I agree.

253

: 00:10:55

Let's see.

254

: 00:10:56

Um, action.

255

: 00:10:57

Front load your value, your

short form content isn't

256

: 00:11:00

driving meaningful engagement.

257

: 00:11:01

That's why we stopped posting

short form, uh, engagement.

258

: 00:11:04

Really, to be honest, 89% of

views come from non-subscribers.

259

: 00:11:09

Let's see your attorney viewers have a 3.7

260

: 00:11:11

click through rate versus a 2.81

261

: 00:11:13

for new viewers.

262

: 00:11:15

And watch a minute, 20 longer,

add stronger CTAs for subscribing

263

: 00:11:18

mid video, not just the end.

264

: 00:11:20

Consider a subscriber CTA

in the first two minutes.

265

: 00:11:23

Okay?

266

: 00:11:24

If you guys are watching right

now, hit subscribe so that way

267

: 00:11:27

you can make Claude and me happy.

268

: 00:11:28

Right?

269

: 00:11:29

And that's, I'm doing my CTA right here.

270

: 00:11:30

Hit subscribe right now.

271

: 00:11:32

Um, okay.

272

: 00:11:33

YouTube search is high volume,

but low engagement brings

273

: 00:11:36

in a lot of YouTube search.

274

: 00:11:39

Your number one source by view count,

but it doesn't have long view duration.

275

: 00:11:43

I don't even know if that's true.

276

: 00:11:45

I'm gonna go back to the actual

raw data here, and I'm just

277

: 00:11:48

gonna go to traffic source.

278

: 00:11:50

Oh, wow.

279

: 00:11:51

See, I don't think I knew that

that search, which is this.

280

: 00:11:57

Now see, oh, I don't know.

281

: 00:11:59

Search is this, um, blue

bar right here, right?

282

: 00:12:02

And it's saying it's my number one

traffic source over the last 365 days.

283

: 00:12:06

And I don't know if it

is browse, is this green?

284

: 00:12:11

And it looks like it's above blue

most of the time, except for, um,

285

: 00:12:16

a little bit of a period last year.

286

: 00:12:20

I don't know.

287

: 00:12:20

That's really interesting.

288

: 00:12:21

I don't know if blue is always higher.

289

: 00:12:24

It's more steady, which is really nice.

290

: 00:12:26

At suggested videos and browse

videos are very similar.

291

: 00:12:29

So if you take green plus yellow,

that's always gonna be bigger than blue.

292

: 00:12:33

But I guess it's not combining

those, it doesn't know, once again,

293

: 00:12:36

it's missing that domain knowledge.

294

: 00:12:37

Browse and suggested are very

similar style type videos because

295

: 00:12:41

once again, you're, you're having

to earn the click in those.

296

: 00:12:45

Um, maybe, maybe I'm over assuming how

similar they are, but from my domain

297

: 00:12:49

experience, these two are very similar.

298

: 00:12:51

Okay, interesting.

299

: 00:12:52

Action search viewers are

looking for quick answers.

300

: 00:12:54

Consider creating dedicated answer videos.

301

: 00:12:57

Under five minutes optimize

for search while keeping your

302

: 00:12:59

longer career advice for browse.

303

: 00:13:01

Interesting.

304

: 00:13:01

Your Tableau export video.

305

: 00:13:04

Of a 15.64%

306

: 00:13:06

CTR proves this model.

307

: 00:13:09

Well, okay, it proves that we have

a high click-through rate, but do

308

: 00:13:12

I get anything from that video?

309

: 00:13:14

I'd argue not really.

310

: 00:13:16

Let's open it up and look at the

data a little bit more closely.

311

: 00:13:18

Okay.

312

: 00:13:18

I have that video popped up.

313

: 00:13:19

As you can see, it has

37, almost 38,000 views.

314

: 00:13:23

473, uh, watch hours.

315

: 00:13:26

It's only gotten me

$135 and 73 subscriber.

316

: 00:13:30

So although this video is getting

a lot of views, like how am

317

: 00:13:33

I really benefiting from it?

318

: 00:13:35

I've made, let's see, I guess,

how many dollars a day do I make?

319

: 00:13:39

135.

320

: 00:13:40

I make like 10 cents a day

from this video, right?

321

: 00:13:43

1 36 divided by 14, 16 days.

322

: 00:13:49

Yeah, I make 9 cents a day with

this video, um, which isn't

323

: 00:13:55

nothing, I'm grateful for it, but

like, that's not really the point.

324

: 00:14:00

Like even though this has an

unreal, uh, click through rate,

325

: 00:14:03

like this, this click through rate

is really, really high, right?

326

: 00:14:06

Like at 13%, I guess, since published.

327

: 00:14:09

When you compare it to a video

like this, how to become a data

328

: 00:14:11

analyst in 2026, like, look at this.

329

: 00:14:14

So many more subscribers, so

much more money, you know?

330

: 00:14:17

And even though our click through rate.

331

: 00:14:19

Is really low.

332

: 00:14:20

It's probably like 2% for this video.

333

: 00:14:22

Yeah, look at the

click-through rate's only 2.8%.

334

: 00:14:24

Even though the click-through rate kind

of stinks, like the overall number of

335

: 00:14:29

subscribers and revenue is way better.

336

: 00:14:31

So I don't know why it's kind

of fixed on, like the idea that.

337

: 00:14:35

Clickthrough rate is everything.

338

: 00:14:36

Now clickthrough rate's really important,

but just 'cause your clickthrough

339

: 00:14:38

rate is high doesn't mean that it's

good for you as a YouTube creator.

340

: 00:14:41

But I do kinda like the idea of

creating dedicated answer videos under

341

: 00:14:44

five minutes for optimized search.

342

: 00:14:46

That is an interesting concept.

343

: 00:14:47

India viewers have low

engagement shorts aren't working.

344

: 00:14:51

Yep.

345

: 00:14:51

Uh, either invest seriously

in short to the clear strategy

346

: 00:14:54

or stop spending time there.

347

: 00:14:55

And that's unfortunately.

348

: 00:14:56

Uh, what I've done is we

just haven't done any shorts.

349

: 00:14:59

Now, we might start doing shorts in a

little bit, but um, we'll have to see.

350

: 00:15:03

End screens only drove 2,800

views and video cards,: 1592

351

: 00:15:09

That's a very small amount,

so yeah, you're right.

352

: 00:15:11

We don't use end screens nearly as well as

we probably should, which is unfortunate.

353

: 00:15:15

Okay, top strategic recommendations,

double down on data backed career

354

: 00:15:19

insights, AB test thumbnails.

355

: 00:15:20

We're already doing that.

356

: 00:15:21

I like this idea.

357

: 00:15:23

Create a quick answer series.

358

: 00:15:24

I like that.

359

: 00:15:25

Front load hooks in the first 30 seconds.

360

: 00:15:26

I mean, I try, but it's harder

than it, I guess, than it looks

361

: 00:15:30

and leverage your podcast.

362

: 00:15:31

More interview content

drives disproportional.

363

: 00:15:33

Watch time in the Data

Career podcast playlist.

364

: 00:15:35

Use your strongest pay, consider

promoting it more actively

365

: 00:15:38

in your recommended videos.

366

: 00:15:40

Okay.

367

: 00:15:40

Interesting.

368

: 00:15:41

Um.

369

: 00:15:42

So this is like all the analysis it did.

370

: 00:15:44

I don't even know.

371

: 00:15:45

Okay.

372

: 00:15:45

It did extract all of those ZIP files and

to get all those separate CSVs, and for

373

: 00:15:51

each one of these CSVs, it usually has the

table data, the totals and the chart data.

374

: 00:15:55

Okay.

375

: 00:15:56

It kind of like crunched a bunch of these

numbers, but I didn't give much direction.

376

: 00:15:59

And one thing I actually really wanted

to actually see is like a dashboard.

377

: 00:16:03

So I'm gonna actually ask

it to create a dashboard.

378

: 00:16:07

Can you create a dashboard that

monitors these metrics and shows me.

379

: 00:16:11

The key things that I should be looking

at as a YouTube content creator.

380

: 00:16:16

Once again, I'm leaving it pretty.

381

: 00:16:18

Open-ended in general because

I wanna see what it does.

382

: 00:16:20

I'm gonna go ahead and click go.

383

: 00:16:22

And you can see it's currently puzzling.

384

: 00:16:24

It's currently thinking it's

booping, it's gonna try to

385

: 00:16:26

create some sort of a dashboard.

386

: 00:16:28

Now, Claude Code really likes JavaScripts,

so my guess is it's going to create

387

: 00:16:34

some sort of a JavaScript dashboard.

388

: 00:16:36

Um, these JavaScript

dashboards are usually probably

389

: 00:16:40

going to be done in React.

390

: 00:16:43

React is a JavaScript library

that's really good for like.

391

: 00:16:46

Creating websites basically,

and in turn data visualizations.

392

: 00:16:51

Um, let's see what it's thinking.

393

: 00:16:52

So it's entered planning mode.

394

: 00:16:54

Whoa.

395

: 00:16:55

Okay, here's the plan.

396

: 00:16:57

I guess it already planned.

397

: 00:16:57

So it said, uh, Avery's exported

these CSVs, and this is the data.

398

: 00:17:03

We're gonna build a single

self-contained HT mal file with

399

: 00:17:06

embedded CSS and JavaScript.

400

: 00:17:08

Use chart js for the charts, parse all CSV

data and embed it directly as JavaScript

401

: 00:17:12

objects so the dashboard opens instantly

to the browser with no server needed.

402

: 00:17:16

It's gonna have KPI

cards, views over time.

403

: 00:17:19

Top 15 videos.

404

: 00:17:20

Traffic sources, audience, age,

geography, new versus returning viewers,

405

: 00:17:25

subscriber versus non-subscriber.

406

: 00:17:26

Top playlists and content

performance scatter.

407

: 00:17:31

Interesting style will be YouTube

studio inspired dark theme.

408

: 00:17:36

And, uh, okay, let's go ahead and

approve that and see how it does.

409

: 00:17:40

Now, just for reference, so most of you,

I'm assuming, don't have domain knowledge

410

: 00:17:44

of what YouTube studio looks like.

411

: 00:17:47

Um, this is kind of what YouTube

Studio looks like right now.

412

: 00:17:49

Um, it has my latest video, which was

she became a data analyst in 67 days.

413

: 00:17:54

Um, how many views that has,

what's the click through rate?

414

: 00:17:56

What's the average view duration?

415

: 00:17:58

My current subscribers, which is 65,000

and a summary of the last 28 days.

416

: 00:18:04

Um, it's not very graphics based.

417

: 00:18:09

It's not very chart based.

418

: 00:18:10

Like, I don't think there's

any charts on this page.

419

: 00:18:13

Right.

420

: 00:18:14

Um, there's like a lot of cards

with like their own like almost

421

: 00:18:20

ads for different YouTube things.

422

: 00:18:22

Um, if you go to analytics, this

one's like a little bit more of

423

: 00:18:25

what we're trying to look for.

424

: 00:18:27

Um, this is basically what would

be beating on the overview page.

425

: 00:18:31

You basically have your views

over the last 28 days, your watch

426

: 00:18:35

time, your subscribers, and your

revenue of the last 28 days.

427

: 00:18:38

You have a little realtime

counter for your subscribers and

428

: 00:18:41

your views in the last 48 hours.

429

: 00:18:43

And then you have your top content

in this period, which is interesting.

430

: 00:18:47

My, my shorts, there's a few of my

shorts who actually, that actually

431

: 00:18:50

do really well, which is interesting.

432

: 00:18:52

So this is kind of what we have to beat.

433

: 00:18:54

Um, it's very tab based,

which I kind of don't love.

434

: 00:18:57

It'd be nice if you could

just like customize your own

435

: 00:18:59

version of this, I guess.

436

: 00:19:00

It has your audience channels, your

audience watches, those types of things.

437

: 00:19:05

So we'll see if cloud code can kind

of beat this and what it looks like.

438

: 00:19:11

Uh, right now you can kind of see that

it's just extracting and preparing

439

: 00:19:14

all the CSV datas for embedding.

440

: 00:19:16

And then the next step will be

to build the H TM L dashboard

441

: 00:19:19

and then verify dashboard

renders correctly in the browser.

442

: 00:19:22

It's been working for about.

443

: 00:19:24

Three minutes and it's still

thinking we'll go ahead and

444

: 00:19:28

fast forward until Claude tells

me something more interesting.

445

: 00:19:31

Right.

446

: 00:19:32

Alright.

447

: 00:19:32

I just finished the dashboard here

and, uh, it has the KPI cards and

448

: 00:19:37

then the 10 interactive sections.

449

: 00:19:39

It has this little preview for

me over the right hand side.

450

: 00:19:42

I'm not.

451

: 00:19:43

Optimistic.

452

: 00:19:44

I don't think it's gonna look that good,

but let's go ahead and see how it looks.

453

: 00:19:48

I'm gonna go back to my analysis

page and then I'm gonna open up

454

: 00:19:51

the dashboard here separately.

455

: 00:19:53

Okay.

456

: 00:19:54

It doesn't look as bad as I thought.

457

: 00:19:56

Lots of scrolling on the dashboard.

458

: 00:20:00

Um, but at least it has

a lot of information.

459

: 00:20:03

So as total views, watch hours,

subscribers gained revenue

460

: 00:20:06

impressions, view, duration.

461

: 00:20:09

Um, the daily views versus

the seven day moving average.

462

: 00:20:15

Peak periods, which is interesting.

463

: 00:20:18

Top 15 videos by views and the

revenue, the traffic source.

464

: 00:20:23

So search is 31%.

465

: 00:20:25

Oh wow.

466

: 00:20:26

So 31.4

467

: 00:20:27

versus 31.

468

: 00:20:28

So I guess search by itself

actually is my number one source.

469

: 00:20:33

Um, but browse and suggested I feel

like are so similar and that would be.

470

: 00:20:37

52.

471

: 00:20:38

Um, and I've always linked these

together, so that's interesting.

472

: 00:20:42

Okay.

473

: 00:20:42

Then you do see that this is the

interesting thing where my search

474

: 00:20:46

average duration is quite low.

475

: 00:20:49

Um, oh, and I guess this is red.

476

: 00:20:51

'cause it's low and green if it's good.

477

: 00:20:56

So that's interesting.

478

: 00:20:57

So external, okay.

479

: 00:20:58

Top country views.

480

: 00:21:00

The age distribution for views and

watch time subscribed versus non

481

: 00:21:06

subscribed, new versus returning users.

482

: 00:21:09

And then CTR versus view.

483

: 00:21:11

Each bubble equals a video.

484

: 00:21:13

Size is the watch time,

color is the subscriber.

485

: 00:21:17

Rate impressions, and

then click through rate.

486

: 00:21:19

So this has a ton of impressions, a ton of

impressions, but low click through rate.

487

: 00:21:23

But the green.

488

: 00:21:24

Is lots of subscribers.

489

: 00:21:27

So what is this one?

490

: 00:21:28

This one is, Tableau is

easier than you think.

491

: 00:21:29

Oh, it's a short, so there's

no subscribers coming from it.

492

: 00:21:34

There's a lot of views and a decent

amount of a click through rate.

493

: 00:21:37

Okay.

494

: 00:21:38

Interesting.

495

: 00:21:38

I'd almost.

496

: 00:21:40

Remove shorts from this because they

don't really compare top playlist.

497

: 00:21:44

This is interesting.

498

: 00:21:45

So SQLs data analyst, this is something

I've been really interested in.

499

: 00:21:48

This playlist is doing quite well.

500

: 00:21:51

Uh, average due duration.

501

: 00:21:53

So shout out to my Kenyans.

502

: 00:21:54

It looks like you guys

watch my videos the longest.

503

: 00:21:56

Other than us and India, why aren't

you guys watching the videos longer?

504

: 00:22:00

Same with Indonesia and Pakistan.

505

: 00:22:02

Come on guys.

506

: 00:22:03

Come on.

507

: 00:22:03

Maybe it's 'cause I talk crazy.

508

: 00:22:05

I don't know.

509

: 00:22:05

This dashboard is okay.

510

: 00:22:08

I don't think it's anything amazing.

511

: 00:22:10

I mean, obviously I didn't have to make

it and it was quick, so I appreciate that.

512

: 00:22:13

There's not like any crazy insights in

terms of like, is it better than, than

513

: 00:22:19

the YouTube, you know, studio mode.

514

: 00:22:22

Probably not.

515

: 00:22:23

There's some things I like

maybe more and some things.

516

: 00:22:26

Uh, I don't like, I mean, I don't love

scrolling on dashboards really that often.

517

: 00:22:30

But then the other equivalent is

you have to tab it like this, right?

518

: 00:22:33

So are you gonna tab or

are you gonna scroll?

519

: 00:22:35

It just, I guess, depends on how

you like your information viewed.

520

: 00:22:38

I'd rather tab to be honest.

521

: 00:22:39

Um, because then I can actually like

choose where I'm going to versus if I want

522

: 00:22:43

to get to, you know, the traffic sources.

523

: 00:22:45

I have to go by the

top 15 videos by views.

524

: 00:22:48

So on and so forth.

525

: 00:22:49

This isn't bad.

526

: 00:22:50

It's not great.

527

: 00:22:51

It's not bad.

528

: 00:22:52

I'm actually just gonna ask Claude.

529

: 00:22:53

I'm gonna say review the dashboard,

find the pros and cons of your

530

: 00:23:00

data displayed and the, um, UI and

create a second better version.

531

: 00:23:08

So I'm basically telling.

532

: 00:23:10

Claude, you go, Hey, go look at

your dashboard, see what you did.

533

: 00:23:13

Well see what you didn't do well,

and then recreate it based off

534

: 00:23:16

of, you know, your findings.

535

: 00:23:17

Um, this is something I found that's

really interesting is when you have AI

536

: 00:23:20

grading ai, it often works better if

you give it to like a different model.

537

: 00:23:24

Like I gave this to, you

know, chat GPT or GPT 5.2

538

: 00:23:28

or something like that.

539

: 00:23:29

Or Gemini or something like that,

it might do a little bit better.

540

: 00:23:32

Um, but I'm actually interested to see how

it does, it looks like it's struggling.

541

: 00:23:37

It's trying to open the dashboard,

uh, in Chrome and it's not working.

542

: 00:23:43

Let's see what it's trying to do now.

543

: 00:23:45

Okay, let's see.

544

: 00:23:48

It's been trying to open up the dashboard

for quite a bit here and it looks like

545

: 00:23:51

it's really struggling to, when this

sort of thing happens, it's a big red

546

: 00:23:55

flag to me because it keeps trying to

do this and eventually it's gonna give

547

: 00:23:59

up and it's going to just probably

guess what it looks like instead of

548

: 00:24:02

actually knowing what it looks like.

549

: 00:24:03

Although it should be able to like

read the HTML that it created.

550

: 00:24:07

But I get nervous when it like starts

to do these like failures and repeat

551

: 00:24:11

itself over and over again because

eventually it's gonna give up and

552

: 00:24:13

it's just gonna make stuff up and I.

553

: 00:24:14

Done this enough.

554

: 00:24:15

I've spent hundreds of hours analyzing

data with Claude that I know eventually,

555

: 00:24:19

uh, it will make something up.

556

: 00:24:20

Basically, it's like, this is what I think

it looks like, and you have to be really

557

: 00:24:23

careful because you, unless you're paying

really close attention, you can actually

558

: 00:24:27

see what's going on in these log files.

559

: 00:24:29

You might not notice it,

it might not tell you.

560

: 00:24:31

So that's something I have a red flag.

561

: 00:24:33

I'm gonna be looking really

closely to see what it says and

562

: 00:24:36

actually make sure that it's.

563

: 00:24:37

Talking normal versus this

is gonna make something up.

564

: 00:24:40

Alright, so after about five minutes,

it did a full review of V one.

565

: 00:24:45

Some of the V one, uh, cons

were three outta the seven

566

: 00:24:48

charts are completely broke.

567

: 00:24:49

Uh, no month over month.

568

: 00:24:51

Trends missing, computed efficiency

metrics, no upload frequency analysis,

569

: 00:24:56

post data completely unused, no

content category grouping, all

570

: 00:24:59

these different things, right?

571

: 00:25:00

Well, I mean, like it is working.

572

: 00:25:02

We just saw the graph, right?

573

: 00:25:03

Like all this.

574

: 00:25:04

Stuff is working.

575

: 00:25:05

All of this is here, so I don't

know why it thinks it's broken.

576

: 00:25:10

The truth is it doesn't actually

see it, it doesn't actually know

577

: 00:25:13

what's going on, is my guess.

578

: 00:25:15

Or it's just not being rendered

correctly when it's trying to to

579

: 00:25:18

view it, it can actually view it.

580

: 00:25:20

And so it tried to, you know, uh,

it says it's building V two now, but

581

: 00:25:24

it's not, it's literally stopped.

582

: 00:25:26

So maybe I'll say, you know,

keep going and we'll see if.

583

: 00:25:30

It actually does build version two, but

at this point I'm not super optimistic.

584

: 00:25:35

We'll, we'll see how it does.

585

: 00:25:36

Okay.

586

: 00:25:37

After some coaxing, I think it got

this new version of dashboard two here.

587

: 00:25:41

Let's see how it looks.

588

: 00:25:43

Okay.

589

: 00:25:44

Oh, and look, it did go to a tabbing.

590

: 00:25:46

Oh, it's both tabbing.

591

: 00:25:47

Wow.

592

: 00:25:48

It's like it was listening

to our conversation.

593

: 00:25:49

I was like, I like tabs

more, and I added tabs.

594

: 00:25:52

This is interesting.

595

: 00:25:54

Uh, this is like the

number of views on a day.

596

: 00:25:58

I don't get why this is here.

597

: 00:26:00

That's interesting.

598

: 00:26:02

Um, let's see.

599

: 00:26:04

Subscriber efficiency per 1000 views.

600

: 00:26:07

Oh see this is actually interesting.

601

: 00:26:09

So this, this is normalizing it by views.

602

: 00:26:12

So what video brought in the most

subscribers per: 1000

603

: 00:26:17

And that's really

interesting because like.

604

: 00:26:20

Yeah, you know, something like

this one, I analyze this many jobs.

605

: 00:26:25

Like this is getting 37.

606

: 00:26:27

Let's see, it says, oh yeah,

subscribers per: 1000

607

: 00:26:31

See, that's interesting.

608

: 00:26:32

I think this is like the most

interesting graph it's created so far.

609

: 00:26:36

Um, let's see here.

610

: 00:26:38

This one's interesting.

611

: 00:26:40

Watch time, duration by source.

612

: 00:26:43

Um, and then this is the average

duration and this is the watch time.

613

: 00:26:47

I don't think that's

very useful comparison.

614

: 00:26:51

Um, yeah.

615

: 00:26:53

Here's my funnel.

616

: 00:26:54

I guess this is the number of impressions.

617

: 00:26:56

This is the number of views.

618

: 00:26:57

This is watch greater than one minute.

619

: 00:26:58

This is the subscribers.

620

: 00:27:00

That's kinda interesting.

621

: 00:27:01

I like that idea.

622

: 00:27:02

This is views versus average duration.

623

: 00:27:05

I don't know.

624

: 00:27:06

Okay.

625

: 00:27:07

Overall, like this is, this is just fine.

626

: 00:27:10

I think this is nothing

amazing, nothing terrible.

627

: 00:27:14

What I think where it

gets really powerful.

628

: 00:27:16

Is where instead of me just saying,

analyze this, where me as like a

629

: 00:27:20

data analyst, me as a domain expert

come in and be like, I have things

630

: 00:27:24

that I actually want you to look at.

631

: 00:27:26

I have my brain, I know it's important.

632

: 00:27:29

Help me to do the actual

dirty work of the analysis.

633

: 00:27:33

So for example, um, one thing that I

think is, is really powerful or would

634

: 00:27:37

be really interesting to see also what

monthly heat map grid 12 monthly cells.

635

: 00:27:44

I don't even see where that's at.

636

: 00:27:45

Did you guys see a heat map?

637

: 00:27:46

Am I blind?

638

: 00:27:47

Uh, is it like in the old version?

639

: 00:27:50

If I refresh the old version,

I don't see it there either.

640

: 00:27:53

So I'm actually gonna call

Claude out real quick for that.

641

: 00:27:56

I'm gonna say, um, where,

where is the monthly heat map?

642

: 00:28:02

Oh, it's calling.

643

: 00:28:04

Oh, okay.

644

: 00:28:04

It's calling this thing a heat map, which

it's hardly a heat map, but that's fine.

645

: 00:28:09

Uh, it's calling.

646

: 00:28:11

This a heat map.

647

: 00:28:12

I mean, that's a terrible

heat map if I'm being honest.

648

: 00:28:16

So one thing I think would be really

powerful or be interesting actually

649

: 00:28:18

for me to see as someone who's, you

know, invested in this data set is

650

: 00:28:22

something like it tried to make with

this, uh, bubble chart right here.

651

: 00:28:26

I really like bubble charts 'cause it

can show you a lot of variables at once.

652

: 00:28:29

Uh, but I don't think this is quite what

I want in terms of the bubble chart.

653

: 00:28:33

So I'm actually just gonna say,

I'm actually just gonna tell

654

: 00:28:37

Claude what I'd like to see.

655

: 00:28:39

Please make just a standalone bubble chart

of the click-through rate on the x axis.

656

: 00:28:44

The views on the Y axis where each circle

is equal to a video, the size is equal

657

: 00:28:51

to the number of new subscribers, and

the color is the percentage of those

658

: 00:28:55

viewers that came by the search traffic.

659

: 00:28:59

This way, I'm able to see the click

through rate, the views, the number of

660

: 00:29:02

new subscribers, and the percentage.

661

: 00:29:04

Coming from search all in one place,

and that'll let me see outliers

662

: 00:29:08

a little bit better visually.

663

: 00:29:10

Um, because for me data is really hard

to understand unless I can visually

664

: 00:29:15

see it and visually understanding

it, uh, makes it really helpful.

665

: 00:29:20

Okay.

666

: 00:29:20

Traffic source data is only available at

the channel level by date, not per video.

667

: 00:29:24

Hmm.

668

: 00:29:24

That's interesting.

669

: 00:29:25

Didn't earlier, didn't you tell

me, Claude, that there was a video?

670

: 00:29:30

I guess it's by aggregate.

671

: 00:29:31

So it turns out that even though on

YouTube studio, you can see how viewers

672

: 00:29:36

found this video and see that search

was, you know, 60%, that that is not

673

: 00:29:39

available in the dataset that we have.

674

: 00:29:42

There is, that is not included

in the export in YouTube studio.

675

: 00:29:46

So in order to get that data, we'd

have to like take a screenshot,

676

: 00:29:49

uh, or, or jot down these numbers

right here, or use the YouTube API

677

: 00:29:53

and that's for a another video.

678

: 00:29:55

So we're not gonna do that today.

679

: 00:29:56

So instead we're gonna just go ahead and

kind of create a similar bubble chart.

680

: 00:30:00

Um, but instead of the color

being the percent of search.

681

: 00:30:04

Will make it the view time,

the average view duration.

682

: 00:30:07

So it's really not that much different

than this chart, to be perfectly honest.

683

: 00:30:10

I would've liked to have

been a little bit different.

684

: 00:30:12

Um, but I guess we

didn't give it the data.

685

: 00:30:14

However, I did notice on the first version

right here, it gave like this little

686

: 00:30:18

optimization down here where you have

high impressions, low CTR, which basically

687

: 00:30:25

it says thumbnails and tidal problem.

688

: 00:30:27

Top left quadrant.

689

: 00:30:28

While high impressions,

low CTR are down here.

690

: 00:30:31

It's not the top left quadrant right here

and it, but that does mean that there I

691

: 00:30:36

could improve the title and thumbnail.

692

: 00:30:37

So I just got the quadrant area wrong.

693

: 00:30:39

High CTR, low views.

694

: 00:30:42

Well, we don't have views on

this chart anywhere actually.

695

: 00:30:45

So high CTR, low views.

696

: 00:30:47

It doesn't even make sense.

697

: 00:30:48

Right.

698

: 00:30:49

It says it's the bottom right,

but that doesn't make sense.

699

: 00:30:53

Uh, videos in the top right

are your proven winners.

700

: 00:30:55

I don't have any with high

impressions and high clickthroughs,

701

: 00:30:58

so I have no winners, I guess.

702

: 00:30:59

Uh, anyways, let's go ahead and

say, okay, instead of doing the

703

: 00:31:03

percent by search traffic, make

it the average video duration.

704

: 00:31:06

Also, please exclude.

705

: 00:31:08

All shorts, or I guess rather

make two separate charts, one for

706

: 00:31:12

shorts and one for longer videos.

707

: 00:31:14

Because previously it

had put those together.

708

: 00:31:16

Right.

709

: 00:31:16

And that's where you, I mean,

this is a short right here.

710

: 00:31:19

This is a short right here.

711

: 00:31:21

And this is a short right here.

712

: 00:31:23

So basically all of the red.

713

: 00:31:24

Uh, dots on this page were shorts, so

it doesn't really make sense to have,

714

: 00:31:29

you know, shorts and long videos on the

same page because they're quite different

715

: 00:31:33

products and quite different audiences and

quite different purposes, to be honest.

716

: 00:31:37

Okay, I'm gonna hit allow and

let's see what it creates here.

717

: 00:31:39

So finally, after about seven

minutes, uh, I think it finished.

718

: 00:31:43

Um, I did notice there was

a few funny things going on.

719

: 00:31:46

Like for instance, there

was some issues with the.

720

: 00:31:48

Number of data we were trying to look at.

721

: 00:31:51

It looks like, like in terms of context

windows, that makes me a little bit

722

: 00:31:54

nervous and it was having a hard

time actually screenshotting them.

723

: 00:31:58

Um, because there is a lot of, uh,

different, uh, bubbles going on, a

724

: 00:32:02

lot of different data being displayed.

725

: 00:32:04

Um, let's see how it went.

726

: 00:32:07

Okay, so here is our click through versus

bubble chart, and we have it for shorts

727

: 00:32:12

and we have it for long form views.

728

: 00:32:14

So.

729

: 00:32:15

Um, this is interesting to me.

730

: 00:32:19

Um, you obviously have like, almost,

uh, I forgot the, what this is

731

: 00:32:24

called, but like this like shape

where it's like l almost, right,

732

: 00:32:27

which just goes down and then, right.

733

: 00:32:29

And then it looks like we just

have some huge outliers and shorts.

734

: 00:32:33

Um, this interest, this is

interesting to me 'cause now.

735

: 00:32:36

I can like, interpret this data.

736

: 00:32:37

In fact, let me ask Claude

how it interprets this data.

737

: 00:32:40

Um, okay.

738

: 00:32:42

How do you interpret this data?

739

: 00:32:46

What action should I take?

740

: 00:32:49

Uh, we'll see what it says

while I, I give you my thoughts.

741

: 00:32:51

So what, what is the size of the

bubble that's number of subs?

742

: 00:32:54

So even though, like I was saying

earlier, even though these, these

743

: 00:32:56

have huge click through, right?

744

: 00:32:57

They have low subs, um, this one's

the closest one, so data analyst,

745

: 00:33:01

mock interview, we're still getting

a decent amount of subscribers, but

746

: 00:33:05

not even really then we're, we don't

get meaningful number of subscribers

747

: 00:33:10

till about this video right here.

748

: 00:33:13

Um, and then these are obviously where the

subscribers are, are quite substantial.

749

: 00:33:17

Uh, the color is the video duration.

750

: 00:33:21

Um, I don't think I asked

for the that, did I?

751

: 00:33:25

If I did, I didn't mean

to, what did I say up here?

752

: 00:33:28

I said that the, oh yeah.

753

: 00:33:30

I make it the average video duration.

754

: 00:33:32

I meant a VD uh, wait, that is.

755

: 00:33:36

Average view, duration,

not video duration.

756

: 00:33:38

Ah, that's my fault for, for

saying that, but eh, okay.

757

: 00:33:43

So this video right

here, this is 45 minutes.

758

: 00:33:46

Okay.

759

: 00:33:46

Yeah, that makes sense.

760

: 00:33:48

Okay.

761

: 00:33:48

So, um, let's see.

762

: 00:33:50

So the color's basically meaningless.

763

: 00:33:52

I mean, I guess we can say that

the, the interviews, which are

764

: 00:33:57

usually the longer videos, um.

765

: 00:34:00

Don't get a ton of views.

766

: 00:34:02

I guess that's one thing we can look

at it and none of them get a particular

767

: 00:34:06

large amount of sub subscribers.

768

: 00:34:09

Okay.

769

: 00:34:10

Um, let's look at shorts.

770

: 00:34:11

So we have Tableau, remote Data,

job, Tableau, Harvard Saturated data

771

: 00:34:20

sets, Google Analytics certificate.

772

: 00:34:25

And what's the difference between these?

773

: 00:34:27

Okay.

774

: 00:34:28

So it's almost like we have like a low

line, a little bit higher line, and then

775

: 00:34:33

like these three outliers, uh, over here.

776

: 00:34:36

So I think I could just basically

take anything that's above.

777

: 00:34:40

This like bottom line and try

to make more shorts like that.

778

: 00:34:43

That would be kind of my takeaway.

779

: 00:34:44

And pay attention to the titles on these.

780

: 00:34:47

Um, we're missing such a key metric.

781

: 00:34:49

The being able to know like what

the traffic's from is so important.

782

: 00:34:53

So we'd wanna try to get

that data for the future.

783

: 00:34:56

But since we don't have it

right now, that's my takeaways.

784

: 00:34:58

I would just try to make more

videos like these titles in the

785

: 00:35:02

shorts and on this one over here.

786

: 00:35:04

I mean, really the important thing

is anything I would say that's like.

787

: 00:35:10

In this circle, uh, right here, we should

probably include, try to include the

788

: 00:35:15

click through rate on most of these.

789

: 00:35:17

See if we could improve it.

790

: 00:35:19

Um, and on these.

791

: 00:35:22

So that's my takeaway.

792

: 00:35:22

Let's see what Claude says.

793

: 00:35:23

Claude is saying, um, your

biggest videos have the worst CTR.

794

: 00:35:28

Yeah, we've known about that.

795

: 00:35:30

Higher CTR is good, obviously.

796

: 00:35:33

Um, not a whole lot there.

797

: 00:35:35

Shorts two massive outliers.

798

: 00:35:37

We talked about that.

799

: 00:35:37

Shorts don't convert to subscribers.

800

: 00:35:39

We talked about that to.

801

: 00:35:40

Yeah, I don't think this is

necessarily very meaningful, but

802

: 00:35:42

I didn't really come up with that

much better analysis on my end.

803

: 00:35:45

I think really, in order for me to

get really meaningful data outta this,

804

: 00:35:48

we need to have the traffic source

involved, so hopefully that helps.

805

: 00:35:51

You see how I personally

use Claude Code as a.

806

: 00:35:55

Helpful tool as a data analyst,

it makes my work so much faster.

807

: 00:35:59

Doing all this previously would've taken

me so much time to get through everything.

808

: 00:36:03

It does a great job of

creating graphs for me.

809

: 00:36:05

It does a great job of coming up with

some sort of suggestions or some sort

810

: 00:36:08

of actual analysis and, uh, insights.

811

: 00:36:11

That being said.

812

: 00:36:12

I still need to prompt it.

813

: 00:36:14

I still need to ask it what to do.

814

: 00:36:15

I need to, you know, obviously be a domain

expert to try to know what all this stuff

815

: 00:36:20

means and to ask it the right question.

816

: 00:36:22

So I don't really foresee it

replacing any data analyst.

817

: 00:36:25

I kind of just see it being as

the new tool for data analysts to

818

: 00:36:29

actually use to do their analysis.

819

: 00:36:31

But let me know what you think

in the comments down below.

820

: 00:36:34

I appreciate you guys watching

or listening, and I'll see

821

: 00:36:36

you in the next episode.