Why I'm Moving Customers OFF of Performance Max

Kasim: 00:00:11

Hey everybody, John Moran here.

2

: 00:00:12

And today I'm going to share with

you why I'm moving almost all of my

3

: 00:00:15

customers off of Performance Max when

new customer acquisition is the goal.

4

: 00:00:20

So it's not that Performance

Max is a bad campaign type.

5

: 00:00:23

It just sucks at acquiring new customers.

6

: 00:00:25

Effectively Performance Max saves some

of the daily budget for repeat customers.

7

: 00:00:29

Also save some of the budget

for remarketing of warm traffic

8

: 00:00:32

or previous clicks in that

campaign or even other channels.

9

: 00:00:34

So when you crank up.

10

: 00:00:36

Performance max new customer acquisition

doesn't always crank up with it.

11

: 00:00:40

So I have a case study here that

we are focusing on new customer

12

: 00:00:43

acquisition because this client

has a very good lifetime value.

13

: 00:00:46

Their second, third, fourth

purchases are very valuable to them.

14

: 00:00:48

So we need new customers to buy at a high

value for the first time, but also repeat.

15

: 00:00:53

And I don't want to keep paying for them.

16

: 00:00:54

I don't want to have a three,

four, 500 cost per acquisition

17

: 00:00:57

on a returning customer.

18

: 00:00:58

That just takes my profitability,

chucks it out the window,

19

: 00:01:01

but the row has looks good.

20

: 00:01:02

So here's what I'm

going to share with you.

21

: 00:01:04

This is a client that we have since

moved off of Performance Max and

22

: 00:01:09

I've given it enough time to give

it a clear, dedicated case study.

23

: 00:01:12

What I mean by that is I waited about

a month and a half before I shared the

24

: 00:01:16

before and after so there is time where

Performance Max latency is done, the

25

: 00:01:20

conversions that are lagging or behind

are done, and there's been enough time

26

: 00:01:23

for standard shopping to make an impact.

27

: 00:01:25

So we've tested.

28

: 00:01:26

Two main types, well, actually

three types, but in two campaigns of

29

: 00:01:29

performance max, a feed only a full

build and a non feed, and they've

30

: 00:01:32

had fairly good in app success, but

new customer acquisition failed.

31

: 00:01:37

It was too expensive.

32

: 00:01:38

We needed to get below a 400, 425, if

possible below a 300 would be fantastic.

33

: 00:01:45

So what you're going to see here

is a date range of April 25th

34

: 00:01:47

to May 24th versus June 27th to

July 26th, which was yesterday.

35

: 00:01:51

And the total cost reduction is 63,

000 between those two time periods.

36

: 00:01:56

So we spent 54 percent less.

37

: 00:01:58

Now you're going to see 58

percent less conversions.

38

: 00:02:01

It looks like it was cut in half.

39

: 00:02:02

Let me share with you what's

going on inside of the backend.

40

: 00:02:05

I'm going to look at.

41

: 00:02:06

Two, one large area, and then

I'll share with you another large

42

: 00:02:09

area, but two kind of main areas.

43

: 00:02:10

One is the first order date.

44

: 00:02:12

This is something that you should be

checking, especially if you're only

45

: 00:02:15

running Google ads, because you can

kind of identify your activities in

46

: 00:02:18

Google ads, your spend, your campaign

types, and how is this affecting

47

: 00:02:20

the first time customers, the new

customers, not total value, not row

48

: 00:02:24

as new customers and develop an NCAC.

49

: 00:02:26

The other thing I'm going to

share with you is the reduction in

50

: 00:02:28

spend and actually what happened,

which is actually pretty cool.

51

: 00:02:31

Let's go back to the first part.

52

: 00:02:32

We've took 45, 000 in

the P max upgraded one.

53

: 00:02:36

This is what, which is what I was running

previously and reduce that down to zero.

54

: 00:02:39

And then we took a shopping campaign,

a standard shopping campaign, and

55

: 00:02:43

we went from zero up to 25 grand.

56

: 00:02:46

So we went and actually removed about

20, 000 out of the shopping network.

57

: 00:02:51

At the same time, we also removed

36, 000 out of the search network.

58

: 00:02:55

So I.

59

: 00:02:56

Tanked this campaign and spend,

and I did that on purpose.

60

: 00:02:59

One is because we're

going into a slow season.

61

: 00:03:01

It is going to get expensive.

62

: 00:03:03

And I've been working with this client

for this, my third year, every time during

63

: 00:03:06

the summertime, we tank new customers.

64

: 00:03:09

So I needed to pull back,

but I said, you know what?

65

: 00:03:11

I'm going to get more efficient.

66

: 00:03:12

I'm going to focus on new customer

acquisition, not just simply counting

67

: 00:03:14

new brand and repeat and counting it

as row as now, two things you're going

68

: 00:03:18

to say, John, why didn't you upload

the customer list and bid only for new

69

: 00:03:21

customers or bid higher for new customers?

70

: 00:03:22

Did that each time I've tried that and

actually I tried about 14 other accounts,

71

: 00:03:26

I could not get above a 35 percent of

where I was previously in activity.

72

: 00:03:30

So I cut my ad spend and cut my activity

down by 65 percent without trying.

73

: 00:03:35

All I did was turn on bid for new

customers and it would not spend over a.

74

: 00:03:39

Over 35 percent of where it was before.

75

: 00:03:41

It doesn't surprise me when 50,

60, 75 percent sometimes of a

76

: 00:03:45

performance max campaign is not

new customers and not cold traffic.

77

: 00:03:49

So that was number one.

78

: 00:03:49

Number two, you're gonna say,

why didn't you exclude the brand?

79

: 00:03:52

Did that as well.

80

: 00:03:52

I had no brand in PMAX.

81

: 00:03:54

There was not one branded term

in my performance max campaigns.

82

: 00:03:56

It did not show up.

83

: 00:03:58

The CPC did drop a bit.

84

: 00:04:00

The ROAS stayed high and new

customers did not come in when

85

: 00:04:03

I start to increase spend.

86

: 00:04:04

So I think that the two channels,

the search and shopping inside of

87

: 00:04:07

performance max were denied entry by

my existing customers and ran traffic.

88

: 00:04:11

However, the YouTube GSB discover

display and basically kind of

89

: 00:04:15

remarketing aspect of those networks

were still going after that.

90

: 00:04:18

Warm existing customer audience to

get a one cent click for a thousand

91

: 00:04:22

dollars sale, which is about our AOV.

92

: 00:04:24

those two things I tested multiple

times in other accounts tested

93

: 00:04:28

and here as well still failed.

94

: 00:04:29

we know that we've taken out 63, 000

out of Google ads and we know that

95

: 00:04:33

we've reduced 36, 000 out of the search.

96

: 00:04:36

We reduced about 20,000 in shopping.

97

: 00:04:39

So what happened in between

or the last 30 days?

98

: 00:04:41

So in this last 30 days,

I have 181 new customers.

99

: 00:04:44

Okay, so let's see what our cost is.

100

: 00:04:46

It's 53 8 6 1.

101

: 00:04:48

So 53 8 6 1 is giving me a $297

cost per acquired new customer.

102

: 00:04:54

We're under the $300 mark, which

is finally where we need to be.

103

: 00:04:57

Last year during this time, we were

456, so we were down now to 2 91.

104

: 00:05:02

And also last year when

we reduced, we only spent.

105

: 00:05:05

I think 44, 000, we

still had a higher cost.

106

: 00:05:09

So we spent less and then we actually

had a higher cost for acquisition.

107

: 00:05:12

So now we're at 300.

108

: 00:05:14

What was the time period before?

109

: 00:05:15

Well, when we look at April 24th

through May, I'm sorry, April 25th

110

: 00:05:19

to May 24th, which is the date

range here, we're spending 117, 000.

111

: 00:05:23

And now, what was that CAC?

112

: 00:05:25

That was a 434, 000 CAC because we had

000 in spend, 117, We had 270 customers,

113

: 00:05:32

which is here at 200 and there it

is 270 customers and we're at 434.

114

: 00:05:37

So reducing the spend by 54 percent did

not reduce the new customers by 54%.

115

: 00:05:45

It reduced it less, reduced by about 35%.

116

: 00:05:48

Good.

117

: 00:05:49

So what ended up happening is during

the cold season, I've actually

118

: 00:05:52

been able to drop my CAC down.

119

: 00:05:54

From 4.

120

: 00:05:55

34 down to 2.

121

: 00:05:57

97.

122

: 00:05:57

Yes, I did lose volume, but it

came in at a much cheaper CAC.

123

: 00:06:00

The interesting thing, though,

is that in app shows this.

124

: 00:06:03

Watch, watch this.

125

: 00:06:04

Inside of my branded alpha

campaign, which is just, just brand.

126

: 00:06:07

It's called alpha because we

had a beta and yada, yada, yada.

127

: 00:06:09

But that's brand.

128

: 00:06:10

That's my brand campaign.

129

: 00:06:11

I spent 400 less and then got 6.

130

: 00:06:14

5, there we go, 6.

131

: 00:06:17

5 more conversions and my

conversion value went up 11, 000.

132

: 00:06:20

So I spent 400 less, made 11, 000 more.

133

: 00:06:24

Why?

134

: 00:06:24

Well, Performance Max would have taken

that and would have taken that credit.

135

: 00:06:28

Are they returning?

136

: 00:06:29

Yes, there's some returning.

137

: 00:06:30

Is there new?

138

: 00:06:31

Yes, there's actually more new.

139

: 00:06:32

And how do we know?

140

: 00:06:33

Well, what we saw here is that inside

of the YouTube remarketing, we were

141

: 00:06:38

able to take out 3, 700 from my YouTube

remarketing that is exclusive to...

142

: 00:06:43

Only warm traffic on the site,

but not existing customers.

143

: 00:06:46

I lost one sale for 71% less, which means

my YouTube remarketing got more efficient.

144

: 00:06:52

So what's interesting is my

YouTube remarketing started

145

: 00:06:54

to convert much better.

146

: 00:06:56

We can see that my in-app cost per

conversion here was reduced by 70%.

147

: 00:07:01

My brand traffic was reduced by 25%.

148

: 00:07:04

Now, the fun part though is, is that

the performance Max used to get.

149

: 00:07:09

a $232 cost per conversion,

and now my standard shopping

150

: 00:07:13

campaign is getting me $1,100.

151

: 00:07:16

Cost per conversion.

152

: 00:07:17

Whoa.

153

: 00:07:18

200 to 1100.

154

: 00:07:19

John, you're crazy.

155

: 00:07:21

Look at how much you increased

the cost per conversion.

156

: 00:07:23

It's not true because the amount of sales

that are now unattributed from performance

157

: 00:07:28

max that are now properly landing in the

remarketing and the brand is coming in as

158

: 00:07:32

such a much cheaper cost per conversion

than what the fluffiness that, that

159

: 00:07:37

performance max was simply to take credit

for, to continue to spend the money.

160

: 00:07:42

So long story short, I didn't go

from, from a, what was the one here?

161

: 00:07:47

Sorry.

162

: 00:07:47

The 232 costs per conversion up to a 1146.

163

: 00:07:52

That was them separated.

164

: 00:07:54

When you count new customer acquisition,

I was probably still getting a 242

165

: 00:07:57

inside of performance max, but also

getting a whole bunch of overspend.

166

: 00:08:00

I was spending remember 50, 60 percent

of just remarketing existing display,

167

: 00:08:05

YouTube, all the other stuff I saw just.

168

: 00:08:09

Throwing all that money out the

window, that was increasing my CAC.

169

: 00:08:11

Remember, when you're putting your money

in a, in a, in a pool, but you're only

170

: 00:08:14

measuring one half, if that pool is

unaffected by the amount of, of change

171

: 00:08:20

between new versus returning, you're

going to see differences in those CACs.

172

: 00:08:24

So returning CAC and new CAC.

173

: 00:08:25

So it's nice because now I have a

pool where I can dump money knowing

174

: 00:08:27

it's going only after new customers

rather than dumping in more and more

175

: 00:08:30

and more money and only being 50

percent as effective to that traffic.

176

: 00:08:33

I put a dollar in a campaign

and half goes to cold, that's

177

: 00:08:37

going to double my CAC, right?

178

: 00:08:38

Because I'm not putting money

towards the area that I'm measuring.

179

: 00:08:41

Very simple stuff.

180

: 00:08:42

our cost per conversion went up 29%.

181

: 00:08:46

No, it did not go from

a 3 93 up to a 5 0 9.

182

: 00:08:52

It actually went from

a 4 34 down to a 2 97.

183

: 00:08:57

So Google Ads shows a 30% increase.

184

: 00:09:00

Not true.

185

: 00:09:01

Standard shopping just

sucks at attributing.

186

: 00:09:03

We already know this.

187

: 00:09:04

If you've been following me for

some time, this is not new stuff.

188

: 00:09:07

So I went down $135.

189

: 00:09:10

Google says, I went up 115.

190

: 00:09:12

If you measure an app, you're wrong.

191

: 00:09:13

Period.

192

: 00:09:14

now let's look a little bit

more recent time period.

193

: 00:09:16

If I look at the last 30 days here

compared to the previous period, we're

194

: 00:09:22

gonna see something really cool.

195

: 00:09:24

you're gonna see me take

$79,000 and bring it down to 53.

196

: 00:09:28

I pulled 32% more money outta my campaigns

and I took it down another $25,000.

197

: 00:09:35

You'll see where I took it out.

198

: 00:09:36

I took standard shopping out 15

K and I took general out 7, 800.

199

: 00:09:41

What happened to my brand?

200

: 00:09:43

Well, my brand spent 11 percent

less and made, whoa, sorry.

201

: 00:09:46

My brand spent 11 percent less and

made 132 percent more conversions.

202

: 00:09:51

And now look at the cost.

203

: 00:09:52

I spent 155 less and made 32, 000 more.

204

: 00:09:58

So the conversion value, though,

is decreasing cost 38%, lost

205

: 00:10:02

I'm gonna do this, lost 16,

decreased it to 22, lost 27.

206

: 00:10:05

Google thinks I am just you know,

hemorrhaging cash right now.

207

: 00:10:09

What's cool about this, though, is

because the branded traffic spiked up so

208

: 00:10:12

much, which I was gonna get anyway, and

I didn't have to pay 400 per conversion

209

: 00:10:16

in PMAX, or 300 per conversion in

PMAX, I took 32 percent of my cost out.

210

: 00:10:20

I made 18 percent more conversions

and 3, 000 more in conversion value.

211

: 00:10:24

But what actually happened?

212

: 00:10:26

Well, because I'm in a standard shopping

and I'm so much more efficient month

213

: 00:10:29

over month because of negative keywords

and, and also the products and, and

214

: 00:10:32

obviously just good optimization.

215

: 00:10:34

In the last 30 days,

saving myself 25, 000 last.

216

: 00:10:38

30 days compared to previous period,

I have 8 percent more orders.

217

: 00:10:43

I took out 13 percent of

my traffic by pulling back.

218

: 00:10:45

I made 8 percent more orders.

219

: 00:10:47

My returning customer rate didn't blow

up because brand is not new customer.

220

: 00:10:53

Oh, sorry.

221

: 00:10:54

Brand is not existing customers.

222

: 00:10:56

So when I pulled 25, 000 out,

you would think, well, the

223

: 00:11:00

brand traffic's coming in.

224

: 00:11:01

Oh, that branded traffic is going to

be a whole bunch of return customers.

225

: 00:11:03

No, it didn't change.

226

: 00:11:05

So I have slightly better MER.

227

: 00:11:07

My NCAC held, I pulled out fatty, like,

you know, overspend on optimizations.

228

: 00:11:13

And now for this last 30 days, we

are sitting at 359, 000 in conversion

229

: 00:11:18

value for a grand total of 50 K.

230

: 00:11:22

That is a 6X MER.

231

: 00:11:24

My new customers didn't suffer.

232

: 00:11:26

performance max.

233

: 00:11:27

Awesome.

234

: 00:11:28

If you're going to measure in

app, if you're going to say,

235

: 00:11:31

look at my ROAS, if you're going

to say, look at my ROAS, yeah.

236

: 00:11:36

I made my row as worse

and I perform better.

237

: 00:11:40

The client thinks I'm perform better.

238

: 00:11:42

I'm performing better, even

though Google ads looks worse.

239

: 00:11:46

And I think that's the lesson for all

of us here is that if you have a, if

240

: 00:11:52

you have a client that you can speak to

at this elevated level, it's something

241

: 00:11:56

that is, is, is very, very noteworthy.

242

: 00:11:58

Let me just see here real quick.

243

: 00:12:00

Let me pause this real fast.

244

: 00:12:02

Okay, here is actually the email from

the client on Tuesday, July 11th.

245

: 00:12:07

I said that we I think pulling back

from PMAX and pushing a standard

246

: 00:12:10

shopping hub to like year over year,

we're even up 41 percent in revenue.

247

: 00:12:13

So it's completely agree.

248

: 00:12:14

Excited about the results.

249

: 00:12:15

Want to keep strong results on some

lower spend and then go into the fall.

250

: 00:12:19

Like to start back up next week when

you have, you know, blah, blah, blah.

251

: 00:12:21

So we're going to start meeting again

next week, but that was in the past.

252

: 00:12:24

So that's the thing too,

is we actually have.

253

: 00:12:27

A lot less spend, more new cold

traffic, took my CAC down 150 bucks,

254

: 00:12:32

a year over year, I'm 43 percent

higher than we were last year.

255

: 00:12:35

And that's the reason why if I

measure ROAS, or if you look at

256

: 00:12:38

performance max, you're going to fail.

257

: 00:12:41

you'll have high ROAS all

the way until you get fired.

258

: 00:12:45

I'm Jon Moran.

259

: 00:12:45

I hope this is worth worth

some learning for you.

260

: 00:12:47

Bye.