Why Applying tROAS or tCPA Hurt Conversions

Regine: 00:00:11

Hi everyone, Regina here at Starter PPC.

2

: 00:00:13

Today we're talking about the bid

settings called TCPA and TROAS.

3

: 00:00:19

That stands for target cost per

acquisition and target row as if

4

: 00:00:23

anyone has tried to use these settings

and you notice that your impressions

5

: 00:00:28

and your clicks plummet along with

your budget, Being spent 100 percent

6

: 00:00:35

when the T ROAS or T CPA is applied.

7

: 00:00:38

I'm going to explain why that happens.

8

: 00:00:40

But first, let's just look at

an example so everyone knows

9

: 00:00:43

what we're talking about.

10

: 00:00:44

We're all on the same page here.

11

: 00:00:45

So this is one of our clients and

what I'm going to do is head over to

12

: 00:00:50

the change history here on the left.

13

: 00:00:52

And I'm just going to look back in time

through a couple of months because I want

14

: 00:00:57

to find an example of when we applied

the target ROAS or target CPA settings.

15

: 00:01:03

Now for those of you who aren't familiar

with these settings, it's, one of many

16

: 00:01:07

ways of telling the algorithm how to

bid right within the auction what kinds

17

: 00:01:12

of cost per clicks you're willing to

pay and what kind of a return you need.

18

: 00:01:15

So if you set a TROAS setting, you're

saying, Hey, algorithm, I want you to get

19

: 00:01:20

me a ROAS of whatever your setting is.

20

: 00:01:24

Let's say you set it at 200%.

21

: 00:01:26

You're like, I really need this return.

22

: 00:01:29

Where I spend a dollar and

I get 2 in return, right?

23

: 00:01:32

That's a 200 percent return.

24

: 00:01:34

you're kind of automating.

25

: 00:01:36

Whatever the algorithm does and saying

you can bid whatever you want, wherever

26

: 00:01:39

you want, however you want, as long as you

get me an average of 200 percent return.

27

: 00:01:44

that's the intent behind a TROAS setting.

28

: 00:01:47

does it always work out that way?

29

: 00:01:49

here we go.

30

: 00:01:49

Performance Max shopping only.

31

: 00:01:51

that's just a Performance Max campaign

where we don't really give it any assets.

32

: 00:01:55

We only give it the shopping feed.

33

: 00:01:57

So it kind of runs like a

smart shopping campaign.

34

: 00:01:59

So on April 19th, we applied a Target

ROAS setting of a hundred percent.

35

: 00:02:03

And on April 28th, we removed

that Target ROAS setting.

36

: 00:02:07

So let's go and look at this

campaign between these two

37

: 00:02:10

dates, the 19th and the 20th.

38

: 00:02:12

This was April 19th through April 28th,

and let's just compare it with the

39

: 00:02:17

previous period so we can say, we can see

how the numbers changed during that time

40

: 00:02:23

period compared to the previous period.

41

: 00:02:25

Okay, so first thing that we can

see for this PMAX Shopping Only

42

: 00:02:29

campaign, that's the one here on the

top, is cost went down 74%, right?

43

: 00:02:34

So it was spending 600 per

day, now it's spending 166.

44

: 00:02:39

Impressions went down 76%, wow.

45

: 00:02:41

Clicks also went down 75%.

46

: 00:02:44

The average cost per click

actually went up a tiny bit.

47

: 00:02:46

However, click through

rate went up a tiny bit.

48

: 00:02:48

That's good.

49

: 00:02:49

Conversion rate went up a ton, 500

percent increase in conversion rate.

50

: 00:02:54

And best of all, our return,

return on ad spend, right?

51

: 00:02:58

Which we all know.

52

: 00:02:59

ROAS isn't exactly accurate because we're

probably missing a lot of conversions

53

: 00:03:03

in here due to privacy restrictions and

Google's inability to track appropriately.

54

: 00:03:07

However the ROAS that we can see is

still relevant in comparison to the

55

: 00:03:11

ROAS before in the time period before

we can see that this went up quite

56

: 00:03:16

a bit, actually almost 50 percent

increase from 72 percent to 106%.

57

: 00:03:21

And what do we tell it to

get with the TROAS setting?

58

: 00:03:23

100, right?

59

: 00:03:24

So we were getting 72 in this campaign

and we said, we'd like you to get us 100.

60

: 00:03:28

you might be looking at this going,

well, that's not a very high ROAS,

61

: 00:03:31

but again Is shrinking every day.

62

: 00:03:33

You guys, this is a video for

another time, but with privacy laws

63

: 00:03:37

and Google's inability to track

appropriately, if you're seeing a

64

: 00:03:40

hundred percent inside of your Google

ads campaign, check your, check your

65

: 00:03:44

total return for the total business.

66

: 00:03:46

I'd be willing to bet you that

it's a 200 or a 300 percent.

67

: 00:03:49

That's how bad ROAS is right now.

68

: 00:03:52

It just looks worse than it is.

69

: 00:03:53

here's what happened.

70

: 00:03:54

On April let's see, April 19th through

the 28th, we applied a tROAS setting.

71

: 00:03:59

We told it we wanted 100.

72

: 00:04:00

Google's algorithm

said, Great, no problem.

73

: 00:04:02

I can get you 100.

74

: 00:04:03

But what was the cost of that?

75

: 00:04:05

It got way less conversions, you

guys, actually, I don't even have

76

: 00:04:09

the conversions column pulled

up, but if I pulled it up, I

77

: 00:04:12

can tell you with certainty,

actually, let's go and pull it up.

78

: 00:04:14

I'm going to click on columns, modify

columns, search for conversions.

79

: 00:04:19

Conversions.

80

: 00:04:20

I like conversions by time.

81

: 00:04:22

Conversions went up.

82

: 00:04:23

Interesting.

83

: 00:04:24

it spent less.

84

: 00:04:25

Impressions went down a lot.

85

: 00:04:27

Clicks went down a lot, but

somehow conversions went up.

86

: 00:04:31

That's unusual.

87

: 00:04:32

Usually the conversions will go down

and you'll just get less conversions.

88

: 00:04:37

I wonder what would happen if

we looked at conversion value.

89

: 00:04:40

A lot of businesses, especially

those that are just starting

90

: 00:04:42

out have limited budgets.

91

: 00:04:44

And so because of this, they're turned

away by most ad agencies because most ad

92

: 00:04:47

agencies have minimum budget thresholds

that they're willing to work with.

93

: 00:04:51

So what happens is the business owners

end up learning Google ads themselves.

94

: 00:04:54

Okay.

95

: 00:04:54

Conversion value went down.

96

: 00:04:56

So that's really the

only thing that mattered.

97

: 00:04:57

Wow.

98

: 00:04:57

Conversion value went down 62%.

99

: 00:04:59

So it looks like it's continuing

to get conversions, but it's just

100

: 00:05:02

going after people that don't

have a high average order value.

101

: 00:05:06

I think there's an average order

value column that we could pull up.

102

: 00:05:09

Of AOV, which is probably a custom column

I just pulled it up, but I guess I didn't.

103

: 00:05:14

Let me try again.

104

: 00:05:14

Average order value.

105

: 00:05:16

Oh, there we go.

106

: 00:05:16

Now it's pulled up.

107

: 00:05:17

Average order value went down 76%.

108

: 00:05:20

So in this case, the algorithm

said, sure thing, I can get

109

: 00:05:23

you that 100 percent return.

110

: 00:05:25

I can get your ROAS to go up.

111

: 00:05:28

But in doing that, you're

going to lose conversion value.

112

: 00:05:32

In fact, I'm going to make that happen

by just going after the low hanging

113

: 00:05:35

fruit of people that have a lower average

order value because I know that the

114

: 00:05:39

conversion rate on those is much higher.

115

: 00:05:42

Conversion rate is 500 percent higher.

116

: 00:05:44

That's crazy.

117

: 00:05:45

So why does this happen?

118

: 00:05:46

Let's talk about this a little bit.

119

: 00:05:48

the way the algorithm kind of

thinks is, is in terms of like

120

: 00:05:51

pockets of the market, right?

121

: 00:05:53

So it might go, well, I

have this pocket over here.

122

: 00:05:55

Who's interested in

this type of a product.

123

: 00:05:57

And they have this pocket over here who

actually went to the website and proved

124

: 00:06:00

that they're interested in this particular

product, it might have a 200 different

125

: 00:06:04

little pockets that it's looking at.

126

: 00:06:06

And within those pockets,

it kind of knows.

127

: 00:06:08

These people average this AOV,

this kind of conversion rate.

128

: 00:06:12

I can bid pay this much cost per click.

129

: 00:06:14

So if you tell it I really

need to hit my goals.

130

: 00:06:17

I really need to pay my bills.

131

: 00:06:18

I would like to get a higher return.

132

: 00:06:20

It's going to say, okay,

I know what I'll do.

133

: 00:06:23

I, I have 200 pockets of the

market, but only a couple of

134

: 00:06:27

them over here at the top.

135

: 00:06:28

Those are the ones that are actually

getting a 100 percent return.

136

: 00:06:31

So I'm just going to get rid of the rest.

137

: 00:06:33

I'm not going to bid on those anymore.

138

: 00:06:35

And I'm just going to go after the,

the most profitable in terms of return.

139

: 00:06:39

And it'll get you what you want, right?

140

: 00:06:42

But it might not spend the daily budget.

141

: 00:06:44

In this case, the cost went

down a whole whopping 75%.

142

: 00:06:48

Unfortunately, this t res setting,

which is so enticing, and it's the

143

: 00:06:52

same with TCPA, you want to use it,

you want to apply it, because you know

144

: 00:06:57

that you have to hit that return, and

you have to get the algorithm to get

145

: 00:07:01

That profitability up for every sale.

146

: 00:07:04

Unfortunately, it's just

hard to scale that way.

147

: 00:07:07

The algorithm now is kind

of focused on goal hitting.

148

: 00:07:10

It's not really just it's, you know, it's

not going after all the other pockets

149

: 00:07:15

of the market that it kind of discarded

because right now it's just looking at the

150

: 00:07:19

tippy top, most profitable ones for you.

151

: 00:07:23

And so businesses come to us often saying,

I don't know why, but I can't scale every

152

: 00:07:27

time I add budget, doesn't spend, or every

time I add budget, my ROAS goes down.

153

: 00:07:32

That's because that TROAS setting

is not a setting for scaling.

154

: 00:07:36

It's good to use temporarily.

155

: 00:07:38

But you have to allow the algorithm

to go through what we like to

156

: 00:07:41

call mini learning phases, right?

157

: 00:07:43

Where you don't have that

setting turned on and you just

158

: 00:07:46

let it you let it do its thing.

159

: 00:07:48

You let it observe new

pockets of the market.

160

: 00:07:51

And in fact, you have to push it through

to tough times too, where during those

161

: 00:07:54

times when the T row has isn't applied,

you add a tiny bit more budget and

162

: 00:07:58

you let it learn a whole new pocket of

the market where you might discover.

163

: 00:08:03

That, oh, this is actually the

pocket of the market that's gonna

164

: 00:08:06

help us do our next scale period.

165

: 00:08:10

be wary of that tROAS setting,

only use it sparingly.

166

: 00:08:12

We actually have a maximum time limit

that we're willing to apply it each month.

167

: 00:08:17

So that's how we've kind of like contained

the temptation to use tROAS over here.

168

: 00:08:22

We have a lot of processes

in place like that.

169

: 00:08:24

best of luck to you.

170

: 00:08:25

If you need help scaling,

we're here for you.