What to Do if Your Google Ads Campaign Is Not Converting on Cold Terms - Part 2

John: 00:00:11

So our Google ads campaign, especially our PMAX campaign is

2

: 00:00:14

not doing anything for us because

that is going to live and die.

3

: 00:00:18

I would imagine you've

probably seen performance max.

4

: 00:00:19

If we're following that correlation, it's

like when product syncs up, we look great.

5

: 00:00:24

Proxy down, we look bad because

performance max is using that

6

: 00:00:27

as a external traffic source.

7

: 00:00:29

We're not necessarily

driving any of our own users.

8

: 00:00:31

and then this thing also gets

blown up where it's like,

9

: 00:00:41

So John, if I'm seeing like, we've just

launched that YouTube campaign, and

10

: 00:00:46

we're not getting a whole lot of traffic,

but I'm seeing Shopify, we're just not

11

: 00:00:50

like top line numbers aren't moving.

12

: 00:00:52

Like orders are down or

total sales are down.

13

: 00:00:55

So I'm just like, I'm having a hard time.

14

: 00:00:58

I guess with the client, I know the power

of a top of funnel strategy on YouTube and

15

: 00:01:02

driving good quality traffic, like top of

funnel, and then converting through brand

16

: 00:01:06

or wherever they're going to convert.

17

: 00:01:08

But it's hard for me to justify, when

I'm not seeing anything like top of line

18

: 00:01:12

really move, like in terms of sales, like

sessions have increased conversion rates

19

: 00:01:17

going down, If you can help me Absolutely.

20

: 00:01:20

So here's 1 thing to note is when

you're looking at these metrics in

21

: 00:01:23

the back end, the different stream,

like gross and net or attribution.

22

: 00:01:28

All this right here, for example, is

a total amalgamation of everything,

23

: 00:01:34

which means when you drove your YouTube

campaign, so to go to the online store.

24

: 00:01:40

Yes, this is a different,

what date range are you at?

25

: 00:01:43

This is the date that you started YouTube.

26

: 00:01:45

Okay.

27

: 00:01:45

There was also a sale in between

here, which muddles things a bit.

28

: 00:01:49

Exactly.

29

: 00:01:49

Now, what happened on ProductSync

during that time that lost

30

: 00:01:52

61 percent of its revenue?

31

: 00:01:53

I think he pulled back his spend.

32

: 00:01:56

So those two things are not Are going to

cause, this to not be a reliable, metric.

33

: 00:02:03

So while we're looking at leading

indicators, like we started

34

: 00:02:07

YouTube online store did better.

35

: 00:02:09

Okay, now what was the online store?

36

: 00:02:11

Why was it getting better?

37

: 00:02:14

let's look at some traffic sources.

38

: 00:02:15

So traffic on direct is

gonna be down heavily.

39

: 00:02:18

All attribution side of here sucks.

40

: 00:02:20

The direct traffic is things inside

of the back end of Shopify by

41

: 00:02:24

saying this came from product sync.

42

: 00:02:26

where's product sync?

43

: 00:02:27

Product syncs on our site.

44

: 00:02:28

So we're sending our own

traffic to ourselves.

45

: 00:02:30

Yes, attribution Shopify is dead.

46

: 00:02:32

That's why this number here is going

to be really hard to attribute.

47

: 00:02:36

So sales by traffic source

of being down indirect.

48

: 00:02:40

Means that these here are going

to be a little bit messed up.

49

: 00:02:46

This isn't real metrics because this

means that there is an attributed

50

: 00:02:50

sale, which would mean that these

metrics here, which is funny to

51

: 00:02:55

think about all of this revenue here.

52

: 00:02:57

7, 000, 7, 000, 6, 000, 6, 000 is

only going to be able to be identified

53

: 00:03:05

and some here.

54

: 00:03:05

So it's interesting because this means

I don't know, but this means I know.

55

: 00:03:10

So you see how the brand went down 37%.

56

: 00:03:12

So our brand seems to be following

whatever their product sync outreaches

57

: 00:03:16

to Boola, Outbrain, Influencer.

58

: 00:03:18

So our Google ads campaign,

especially our PMAX campaign

59

: 00:03:21

is not doing anything for us.

60

: 00:03:23

Because that is going to live and die.

61

: 00:03:25

I would imagine you've

probably seen performance max.

62

: 00:03:27

If we're calling a correlation, it's like

when product sinks up, we look great.

63

: 00:03:32

Proxy down, we look bad because

performance max is using that

64

: 00:03:35

as the external traffic source.

65

: 00:03:37

We're not necessarily

driving any of our own users.

66

: 00:03:40

and then this thing also gets

blown off where it's like, ah,

67

: 00:03:43

like awesome Easter sale people.

68

: 00:03:45

Oh, my God.

69

: 00:03:47

Look at the Delta between these two.

70

: 00:03:49

So this Delta is the same as that

Delta, which is the same as that Delta.

71

: 00:03:52

so the blue has stayed about the same.

72

: 00:03:54

The blue was only lifted up

by the purple, pushing it up.

73

: 00:03:57

A lot of returning customers.

74

: 00:03:59

New customer growth is if the blue is

starting to crawl away from the purple,

75

: 00:04:03

but we already established that this

company makes majority of its first

76

: 00:04:07

time sales through attentive and Clavio.

77

: 00:04:10

Where's it coming from?

78

: 00:04:11

Taboola, Outbrain, not Pmax, but when

we drive good old fashioned standard

79

: 00:04:15

shopping, as an example, and remember

my glass ceiling meeting from everybody,

80

: 00:04:20

a few weeks ago when I said traffic

sources will have different glass

81

: 00:04:22

ceiling, standard shopping might

hit and media efficiency ratio crash.

82

: 00:04:27

With 10 grand a month, that might

be the only 2 percent of available

83

: 00:04:30

traffic that we have in direct response

that wants to buy our product today.

84

: 00:04:33

So only 2 percent of the available

market that wants to buy your

85

: 00:04:35

product at any given time that

may 10 grand might be 20 grand.

86

: 00:04:38

I don't know.

87

: 00:04:39

But when we start to push past that, we

see NCAC drop or NCAC keep climbing up.

88

: 00:04:43

Okay.

89

: 00:04:43

We hit a point to make

sure in terms of pullback.

90

: 00:04:45

This though is going to be so

heavily reliant on this dude

91

: 00:04:48

here telling us what the hell he

is doing on that channel there.

92

: 00:04:55

If you're going to have 200, 000 swings

in 30 days, nothing that we do is going

93

: 00:05:00

to be helpful, but we should see though.

94

: 00:05:03

NCAC, boom, fluctuate 10, 20%, but at

crazy volumes, 40 NCAC, 10, 000 sales.

95

: 00:05:10

50 NCAC, 60, 000 sales, 20 NCAC, 12 sales.

96

: 00:05:15

So those things are, the NCAC is going to

fluctuate based on the amount of volume,

97

: 00:05:19

usually higher volume, slightly higher

NCAC, lower volume, slight lower NCAC.

98

: 00:05:22

So when we're looking at these big

swings, do we know if there's a

99

: 00:05:24

cost associated that he has here?

100

: 00:05:26

No, I don't know.

101

: 00:05:27

Okay.

102

: 00:05:27

And you said that he's running Google

and Meta, but he also runs Tubula,

103

: 00:05:31

Outbrain, those other types of things.

104

: 00:05:32

Yeah, those do have

costs associated to them.

105

: 00:05:34

Yeah.

106

: 00:05:35

And like affiliate.

107

: 00:05:36

So that would obviously,

yeah, it would have a cost.

108

: 00:05:38

Yeah, so that's what we have to

do is we have to get, hey, what

109

: 00:05:42

are you spending everywhere?

110

: 00:05:43

Because we're going to be driving some

high quality, good old fashioned traffic.

111

: 00:05:47

you're doing a meta.

112

: 00:05:48

But it's also going to look like meta.

113

: 00:05:49

It's going to look crappy, but if we

can see that we can have a scalable

114

: 00:05:53

achievable NCAC, because once they get

to your site, sir, it was a guy, right?

115

: 00:05:57

Yeah.

116

: 00:05:58

Once we get to your

site, sir, they buy that.

117

: 00:06:00

They go through the quiz and then you

text them and boom, we got a new customer.

118

: 00:06:04

Or then they start Googling Cora things.

119

: 00:06:06

is it good?

120

: 00:06:07

And then Cora converts

them under native ads.

121

: 00:06:09

I don't know, but all of the conversion

paths that these people have, that he

122

: 00:06:12

has figured out how to take, we want to

increase the amount of traffic to it.

123

: 00:06:15

We just can't, say PMAX followed

product sync and now we're good.

124

: 00:06:18

Yeah.

125

: 00:06:18

We've known this for a

long time with the PMAX.

126

: 00:06:20

It's just that even when we run standard

shopping, like standard shopping We'll

127

: 00:06:23

also stick with the branded terms.

128

: 00:06:25

So we can't get standard shopping

to go cold or even show like

129

: 00:06:29

impressions or whatever on cold

terms, peptide cream or whatever.

130

: 00:06:33

Like we get little volume there.

131

: 00:06:34

So that also becomes hard.

132

: 00:06:36

The only traction on like top funnel, cold

stuff was P max from, even though it's

133

: 00:06:42

half free marketing, and branded terms.

134

: 00:06:45

have you run a non

branded center shopping?

135

: 00:06:47

Yes.

136

: 00:06:48

repair and release.

137

: 00:06:50

they all should be non branded,

but they'll still branded

138

: 00:06:52

terms will squeak in there.

139

: 00:06:54

this is good.

140

: 00:06:55

because like really good amount

of quality down here, but, almost

141

: 00:06:59

like next to zero conversions,

which is like very high in CPA.

142

: 00:07:03

Yep.

143

: 00:07:03

And that's the same as.

144

: 00:07:05

I was just going to say, we didn't

see like a huge difference in

145

: 00:07:07

Shopify, but we could have been

looking at the wrong metrics and

146

: 00:07:11

that's totally on us if we were, but.

147

: 00:07:13

Yeah.

148

: 00:07:14

Yeah.

149

: 00:07:15

That's the thing too, is like this one

might've crushed it or might be really

150

: 00:07:18

broad, but peptides for skin, like that

seems to be a good, like when we can

151

: 00:07:22

attribute something that looks healthy.

152

: 00:07:23

Yeah.

153

: 00:07:24

It was a good term when

we could attribute it.

154

: 00:07:25

Yeah.

155

: 00:07:26

Yeah.

156

: 00:07:26

So if we say peptides,

that's 100 right there.

157

: 00:07:30

that's fantastic.

158

: 00:07:30

And we probably, we don't even

know what really came from this,

159

: 00:07:33

but peptides was like, yeah, boom.

160

: 00:07:34

And it was pretty consistent for

a hundred bucks, which we look at,

161

: 00:07:37

an average order value of one 26.

162

: 00:07:39

That's a, probably a really

shitty attribution, but

163

: 00:07:43

we're still giving them a 1.

164

: 00:07:44

25, even though we can attribute.

165

: 00:07:46

But yeah, so this is a good

indicator of good quality traffic.

166

: 00:07:49

what I would say is a good idea is to

go through and see like the themes.

167

: 00:07:52

So it definitely seems like a

peptide, neuropeptide, peptide,

168

: 00:07:55

even though that was called signal

does was peptide, skin camera,

169

: 00:07:58

peptides, like all of these are good.

170

: 00:08:00

So we turn this into a, try

to use a peptide, and do it

171

: 00:08:04

like a, do it T row as of 20%.

172

: 00:08:07

this has got good click.

173

: 00:08:08

The rates got really good.

174

: 00:08:11

this would be a really good indicator

to start to push into, to see how that

175

: 00:08:14

affects our NCAC totals, because this is

where we can see the interest, whether

176

: 00:08:19

they immediately buy on the site majority

of our ad spend should be going towards,

177

: 00:08:25

increasing that good quality traffic.

178

: 00:08:26

I think that PMAX is just trying to

get the returning users or people that

179

: 00:08:29

are, Googling things about their brand.

180

: 00:08:31

. Yeah.

181

: 00:08:32

That is what's happening.

182

: 00:08:33

Okay.

183

: 00:08:34

So John, in the sequencing campaign

in YouTube, for the same account, it

184

: 00:08:38

is, showing like it conversion in last

seven days, but in North it is zero.

185

: 00:08:42

Can you just help me why it is.

186

: 00:08:44

Like

187

: 00:08:45

Absolutely.

188

: 00:08:45

when you're looking in the last 14

days, what do you wanna measure?

189

: 00:08:47

Is the time lag between

conversions and conversion?

190

: 00:08:49

By time, what The go away stupid thing.

191

: 00:08:52

So check this out.

192

: 00:08:53

This is the last 14 days.

193

: 00:08:56

Let's just count this here.

194

: 00:08:57

At day one ish, the

blue is the conversions.

195

: 00:09:02

The red is the conversion

by conversion time.

196

: 00:09:05

So there is a time lag here between

when they started to engage with our

197

: 00:09:08

content versus when they actually

pulled the trigger on something.

198

: 00:09:11

And you'll see that this is, on

day two, this started to work.

199

: 00:09:16

You just didn't see it until day six.

200

: 00:09:18

So that top of funnel.

201

: 00:09:20

Is saying, okay, we're starting

to hit some cold traffic.

202

: 00:09:22

They're starting to

engage with our content.

203

: 00:09:24

Yes.

204

: 00:09:25

When did they buy?

205

: 00:09:25

we spent the money on day one.

206

: 00:09:27

They bought on day six ish.

207

: 00:09:28

I don't see the dates here, unfortunately,

but those are the things that we look

208

: 00:09:32

for is those kind of, Patterns, I guess

I would say those are the patterns that

209

: 00:09:35

we're looking at it's gonna be very much

like meta we're hitting that cold traffic.

210

: 00:09:39

They're not just going to see YouTube.

211

: 00:09:40

I'd be like, Oh, my God.

212

: 00:09:41

And then by, we're warming them up.

213

: 00:09:43

That's why high frequency is good on

YouTube, which is what campaign does is

214

: 00:09:47

this focuses on a high frequency because

the structure of the sequence means you

215

: 00:09:52

have to go through a high frequency.

216

: 00:09:53

That's what's really good here is

they're seeing at least four and if

217

: 00:09:57

they skip, they still see, skip one.

218

: 00:09:59

It's view base and the

whole story is different.

219

: 00:10:02

Like they might look same from

thumbnail, but they are totally

220

: 00:10:06

different videos, different story

and all of the other things.

221

: 00:10:09

like everything below this line,

impression skips, as our impression steps,

222

: 00:10:13

what this means is that, if they view,

and then skip this one, if they also

223

: 00:10:18

skip this one, they don't see that one.

224

: 00:10:20

So what I would say is take these here

and change this to, we can't rebuild.

225

: 00:10:26

Don't rebuild it.

226

: 00:10:27

It's a big pain in the ass to rebuild us.

227

: 00:10:29

But if it's usually a, impression

based skip, what it means is that

228

: 00:10:34

they can view this one, skip,

and then they don't see that one.

229

: 00:10:37

They don't see that one.

230

: 00:10:38

So usually at any point in time

that they skip, these are usually

231

: 00:10:40

just impressions after this, because

I'm going to force you to watch the

232

: 00:10:43

next ones, whether you skip or not,

233

: 00:10:45

especially here, it's if I skip

that one, they're not going

234

: 00:10:47

to see anything down here.

235

: 00:10:49

Because they have to view this

one in order to continue on, they

236

: 00:10:51

have to view this one, they'll go

through this one, which is great,

237

: 00:10:54

but see how you did this one.

238

: 00:10:55

Okay.

239

: 00:10:55

So you skip the 1st 1, you skip this 1.

240

: 00:10:57

Now you're going through all of it.

241

: 00:10:58

So this is probably going to have

the biggest path is step to skip 1

242

: 00:11:02

step to skip 1 and then everything else.

243

: 00:11:05

has less impressions because this

is the path that they are taking

244

: 00:11:09

now because of that structure.

245

: 00:11:10

Yeah, this is one that

says, okay, did you skip?

246

: 00:11:12

Yeah.

247

: 00:11:12

Okay.

248

: 00:11:13

This one's going to be it.

249

: 00:11:13

Everything else is going

to be you watched it.

250

: 00:11:15

You watched it.

251

: 00:11:16

You watched it, which is good.

252

: 00:11:17

It's okay.

253

: 00:11:18

but this will just skew those metrics

into being like, okay, if you skip the

254

: 00:11:21

perfect scenario and I force a video

at every section there on out, do they

255

: 00:11:27

come back on the 3rd or 4th video?

256

: 00:11:29

But we can't see that because

we're just forcing him through.

257

: 00:11:31

Okay.

258

: 00:11:32

You can skip everything

and this is the path.

259

: 00:11:34

So all of our non engage users have

to go through this path and we don't

260

: 00:11:38

see if we ever pick them back up yet.

261

: 00:11:40

That makes sense.

262

: 00:11:41

But, yeah, this is honestly in

this scenario, this is perfect.

263

: 00:11:43

This is what we want.

264

: 00:11:44

a 443 cost for conversion on impression

based attribution, which is I'm

265

: 00:11:51

imagining is probably going to be it.

266

: 00:11:53

Out of a type, gauge view, step two,

add sequence two, five, engage views.

267

: 00:11:57

Great.

268

: 00:11:58

This is okay.

269

: 00:11:59

as long as we can confirm this is fairly

cold traffic, which we can look at.

270

: 00:12:05

Here's one thing that I always

do in my YouTube sequences.

271

: 00:12:07

is build another ad group with the

people that, visit the site and

272

: 00:12:12

then clone that, from the other one.

273

: 00:12:14

So it's a big pain in the

butt, actually an easier way.

274

: 00:12:17

Just clone the campaign and in

the campaign here, have your

275

: 00:12:20

exclusions be website traffic.

276

: 00:12:22

So this one stays cold.

277

: 00:12:24

If they ever click,

they jump into the next.

278

: 00:12:26

The other campaign, which

is the remarketing one.

279

: 00:12:28

Then what's nice is if you want

to take that two 50 a day and then

280

: 00:12:31

the remarketing, let's say is that

50 a day, it's like, all right,

281

: 00:12:34

scale, I can scale a two 51.

282

: 00:12:37

Not necessarily scaled.

283

: 00:12:38

That could be going towards warm traffic.

284

: 00:12:40

That's on the site.

285

: 00:12:41

Sequencing stays pretty

cold though, which is okay.

286

: 00:12:43

So we're okay for now.

287

: 00:12:44

That's just yeah.

288

: 00:12:45

All right.

289

: 00:12:45

Everything that we're

doing is actually good.

290

: 00:12:46

We're starting to push, starting to scale.

291

: 00:12:48

I wouldn't just dump all of our money

into this one to scale it because

292

: 00:12:51

this will turn more warm later on.

293

: 00:12:53

that's what Google does.

294

: 00:12:54

so it'd be good to potentially

divide those into a prospecting

295

: 00:12:57

only and remarketing only that

have the same videos that when they

296

: 00:13:00

click, they jump into this campaign.

297

: 00:13:02

And so you can raise this up

20%, 30%, 40%, 50%, which means,

298

: 00:13:06

a 200 hour campaign and a 50.

299

: 00:13:07

He had 10%.

300

: 00:13:09

A 200 goes into 2050 goes to 55,

so they can still keep the same

301

: 00:13:13

amount of engagement, but you're

able to scale something that, is

302

: 00:13:15

prospecting traffic to hold that in.

303

: 00:13:17

This is fun.

304

: 00:13:18

Thanks, John.

305

: 00:13:19

I appreciate it.

306

: 00:13:20

Thank you so much.

307

: 00:13:21

You got it.

308

: 00:13:22

Good stuff.