Google Knows You More Than You Know Yourself!

Joined
Jul 28, 2017
Messages
25
Likes
60
Degree
0
If you want to see how far technology has come, read up on digital advertising and tracking technology and how the tech giants (Google, Meta, Twitter, to name a few) are collecting huge amounts of data to literally predict the future. Think this is nuts? Read until the end.

I was setting up a campaign for a client a few weeks ago in Google Ads. I used the target ROAS bidding strategy. Don't let the name scare you. I'll explain what it means. Basically, you tell google that for every $1 spent on ads, I want to make x amount in return.

Google then uses audience signals and then taps into the huge amount of data they have to predict user behavior. They know what the user is searching for, and they know what he likes to read online.

They know what he loves to watch online and much more, and they can infer from this data what action a user is likely to take, anyway, back to my client's campaign. It was a shopping campaign. After setting up the campaign. I let it run for a few days (learning phase)

I set up a target ROAS of 300%, meaning for every $1 you spend on ads, you make back $3 in revenue) After a week or so, the campaign stabilized and had an overall ROAS of exactly 300%! Can you see how insane this is?

They show my ads for a subset of users who have a high likelihood of purchasing from the site based on user signals like age, gender, affinities, etc. Like they know for a fact that x user has an x% chance of purchasing x product.

It's just something that is really fascinating and got me thinking.
 
@simplebuilder, do you feel that these companies (Google and Facebook come to mind) where it's like "here, just give us your budget and let us run the ads and optimize everything for you based on your goal" is a better approach for us as advertisers than trying to get into the nitty gritty with a million optimizations of our own, micro-managing negative keywords and all that?

I think the approaches are to either micro-manage crap like creatives and keywords and spend forever doing it, or let them optimize based on co-horts and user data we don't have access to.

My assumption is we'll spend more money letting them do it, and perhaps eek out a lower ROI than if we spent 100 hours a week micro-managing, but we don't have to micro-manage.

What's your thoughts on this (or anyone else with experience)?
 
@simplebuilder, do you feel that these companies (Google and Facebook come to mind) where it's like "here, just give us your budget and let us run the ads and optimize everything for you based on your goal" is a better approach for us as advertisers than trying to get into the nitty gritty with a million optimizations of our own, micro-managing negative keywords and all that?

I think the approaches are to either micro-manage crap like creatives and keywords and spend forever doing it, or let them optimize based on co-horts and user data we don't have access to.

My assumption is we'll spend more money letting them do it, and perhaps eek out a lower ROI than if we spent 100 hours a week micro-managing, but we don't have to micro-manage.

What's your thoughts on this (or anyone else with experience)?
I will answer from the Google Ads side of things, as that is where most of my experience lies.

Both approaches have their own advantages and disadvantages.

I take a hybrid approach to this.

Google is a machine-learning mechanism that needs data to work its magic.

Without conversion data on the account, automation will suffer to produce results. So it is best to get conversions with the manual approach in the beginning and then let Google’s automation take over at some point; this is the best way I have found thus far.

When I first set up a Google Ads account with no conversion history, I take a manual approach and micro-manage every minute detail to get the first few conversions; once that is done, it is better to hand over *some* control to google and let them optimise based on the data that we don’t have access to, but making some changes from time to time such as adding negative keywords is important, else the campaigns will leave money on the table.

By doing this, you get the best of both worlds, where you tailor your campaigns to your specific needs and take advantage of the advanced targeting and optimisation that Google has to offer.
 
Last edited:
Without conversion data on the account, automation will suffer to produce results. So it is best to get conversions with the manual approach in the beginning and then let Google’s automation take over at some point; this is the best way I have found thus far.

I am guessing you haven't tried the PMax Campaigns yet. They are the latest way Google is trying to automate everything from rotating headlines/descriptions to showing your ads on the best places "they think" you will get conversions. So many people jumped on this feature to later realize that Google was literally using their money to test ads on every network they have and even cannabilizing their brand keyword campaigns.

All I am saying is be careful how much control you give them. At the end of the day, they want your $
 
I am guessing you haven't tried the PMax Campaigns yet. They are the latest way Google is trying to automate everything from rotating headlines/descriptions to showing your ads on the best places "they think" you will get conversions. So many people jumped on this feature to later realize that Google was literally using their money to test ads on every network they have and even cannabilizing their brand keyword campaigns.

All I am saying is be careful how much control you give them. At the end of the day, they want your $
I see where you’re coming from, but PMax is literally the future.

It may not work on every niche and geo, but when it works, it works spectacularly and perform leaps and bounds better than any campaign type.

In my earlier days with Google ads, I used to be very anti-automation and tried to do everything on my own and avoid Performance Max and automated bidding, and was very fearful of letting Google control any aspect of my ad spend.

Then I got a client who has a local business; we tried almost every campaign type Google has to offer: search, display, youtube, local ads, call-only ads, anything.

We weren’t able to go beyond 2x ROAS.

Then we tried Performance Max; performance max needs assets (creatives) + audience signals.

You can choose from the predefined audiences that Google has, but we uploaded the list of previous customers (customer match list) and let the campaign run.

The more data the campaign gets, the better it can target and optimize; in less than two weeks, ROAS jumped to 7x and has been consistent ever since.

I also tried it for various eCommerce niches, and it works spectacularly, outperforming standard shopping campaigns.

In the local business that I mentioned, the time-to-live was short because the account already had conversion history, and we uploaded the list of past customers as an audience signal, so the campaign could go and look for audiences with similar characteristics.

Unlike other campaign types that produce near instantaneous results, PMax takes time to work.

PMax requires time and effort to get working, but it is worth it.

Check this tutorial; it is the best resource on this subject: https://sol8.com/performance-max/

If you’re running Google ads for a local business or eCommerce and not taking advantage of performance max, you’re missing out big time.
 
It may not work on every niche and geo, but when it works, it works spectacularly and perform leaps and bounds better than any campaign type.

Let's agree to disagree here. Without going into too much detail - the fact that they run your ads on the display network using PMax should tell you all you need to know about it.
 
Let's agree to disagree here. Without going into too much detail - the fact that they run your ads on the display network using PMax should tell you all you need to know about it.
Nothing wrong with the display network. I'm managing ads right now for SaaS and we are getting 1500+ paid subscribers on a daily basis solely from standard display campaigns.

But, let's agree to disagree.
 
do you feel that these companies (Google and Facebook come to mind) where it's like "here, just give us your budget and let us run the ads and optimize everything for you based on your goal" is a better approach for us as advertisers than trying to get into the nitty gritty with a million optimizations of our own, micro-managing negative keywords and all that?

My personal experience is that it's usually not worth it to let Google run the show, particularly not for low search volume or smaller budgets.

What I've seen from the auto-bidding is that Google really likes to massively overbid on some users and some queries. Like if you have a small budget of $20 a day for a campaign, Google will just bid $10 on a single user it thinks will convert. This kind of bidding strategy creates a lot of variance, which as I understand it, must be offset by a larger budget.

For my long running affiliate campaigns, with very predictable search volume and user base, it does better, where the ROI was not great in the first place. Then it usually is about equal.

The thing is, that when I talk to Google reps, you definitely get the idea that they want auto-bidding to be the choice for your average small business and in my experience, these are the people who in my opinion are the least likely to benefit.
 
Nothing wrong with the display network. I'm managing ads right now for SaaS and we are getting 1500+ paid subscribers on a daily basis solely from standard display campaigns.

But, let's agree to disagree.
Wait - 1500 PAID subs p/day?

You have to talk more about this.

I want to outsource paid traffic later this year.
 
Why the hate on Pmax?

I've been running paid ads since 1999 ( GoTo ) and Pmax is one of the most profitable campaigns I run. Not for just 1 company, but dozens of clients.

We're talking million dollar budgets per client, running on all types of campaigns that include Pmax and Pmax being just as profitable, if not the most profitable ( sometimes ).

ROAS consistently 8.00+ ( 800%+ ) months on end on it.

If you know how to run paid ads, you can make almost anything work as long as the offer is not a dud.

And, Pmax isn't just display network. It's all their networks which does include display
 
Wait - 1500 PAID subs p/day?

You have to talk more about this.

I want to outsource paid traffic later this year.
The campaigns are pretty simple; these are standard display campaigns.

I installed conversion tracking, and then I got the creatives from the design team and launched the campaigns on the display network with Google audience segments.

After a campaign gets 15 - 30 conversions, I switch to automated bidding.

I spend a few minutes every once in a while removing bad placements and working on optimising the copy and creatives.

On some campaigns, we are even paying for conversions only (this is an option on the display network), meaning we don’t pay google if we don’t get paid subscribers.

I believe what is making the difference is the geo we’re targeting.

Google recently added the option to target Sudan in Google ads, this was previously not possible due to sanctions, so they are not a lot of advertisers locally, and the average internet user in Sudan doesn’t get a lot of ads.

CPC is anywhere from $0.01 - $0.04
CTR is 1% - 5% (the industry average is 0.57% on the display network)
The monthly subscription to the SaaS is approximately 6$/month
The cost per new subscriber from Google ads is $0.30 - $1

We are promoting multiple SaaS products currently, all of which have a broad appeal. One of them is a gaming service; another one is an audiobook service. We are currently running ads for five SaaS products, all through the display network.

If this was a very niche product, I would’ve concentrated on search campaigns and used display campaigns only for remarketing, but as I mentioned, this is somewhat different because of the broad appeal of the products.
 
Last edited:
The campaigns are pretty simple; these are standard display campaigns.

I installed conversion tracking, and then I got the creatives from the design team and launched the campaigns on the display network with Google audience segments.

After a campaign gets 15 - 30 conversions, I switch to automated bidding.

I spend a few minutes every once in a while removing bad placements and working on optimising the copy and creatives.

On some campaigns, we are even paying for conversions only (this is an option on the display network), meaning we don’t pay google if we don’t get paid subscribers.

I believe what is making the difference is the geo we’re targeting.

Google recently added the option to target Sudan in Google ads, this was previously not possible due to sanctions, so they are not a lot of advertisers locally, and the average internet user in Sudan doesn’t get a lot of ads.

CPC is anywhere from $0.01 - $0.04
CTR is 1% - 5% (the industry average is 0.57% on the display network)
The monthly subscription to the SaaS is approximately 6$/month
The cost per new subscriber from Google ads is $0.30 - $1

We are promoting multiple SaaS products currently, all of which have a broad appeal. One of them is a gaming service; another one is an audiobook service. We are currently running ads for five SaaS products, all through the display network.

If this was a very niche product, I would’ve concentrated on search campaigns and used display campaigns only for remarketing, but as I mentioned, this is somewhat different because of the broad appeal of the products.
Excellent response.
Now for the big question - do you think you can swim in the big leagues? Ie. US/UK/CA
 
@simplebuilder, do you feel that these companies (Google and Facebook come to mind) where it's like "here, just give us your budget and let us run the ads and optimize everything for you based on your goal" is a better approach for us as advertisers than trying to get into the nitty gritty with a million optimizations of our own, micro-managing negative keywords and all that?

I think the approaches are to either micro-manage crap like creatives and keywords and spend forever doing it, or let them optimize based on co-horts and user data we don't have access to.

My assumption is we'll spend more money letting them do it, and perhaps eek out a lower ROI than if we spent 100 hours a week micro-managing, but we don't have to micro-manage.

What's your thoughts on this (or anyone else with experience)?

I worked at a very large agency in different functions (SEA is one of them). Before smart bidding we did manage everything on the most granular level. This worked as we got much better results than any other competitor.

When smart bidding came around, eventually the company jumped on it. Even the best SEA specialists were not able to outperform the algorithm (which makes sense to be honest). Even when you would not be counting the hours you are saving by using smart bidding. The key thing is to feed the algorithm as much as relevant data as possible. The company also developed some bidding algorithms on its own, but they also did bite the dust when competing with G (which also makes sense).

I cannot add to much to the PMAX discussion here as I left around the time it was introduced.
 
When I first set up a Google Ads account with no conversion history, I take a manual approach and micro-manage every minute detail to get the first few conversions; once that is done, it is better to hand over *some* control to google and let them optimise based on the data that we don’t have access to, but making some changes from time to time such as adding negative keywords is important, else the campaigns will leave money on the table.
We are doing the same. First week we cut out non performing keywords or obvious low quality keywords and add some negative ones.

I see where you’re coming from, but PMax is literally the future.
Yes. We just put in as many creative assets as possible and let it run with a high conversion goal.

After 2 weeks we usually have enough conversion data (sales in our case) and start lowering the target CPA to increase our ROI. Depending on the product we have 50-100% ROI. The latest product has 60% ROI after 7 days with minimal time investment on our side.

There is one project running with 3-4k ad spend per day completely on autopilot for one year now.

My personal experience is that it's usually not worth it to let Google run the show, particularly not for low search volume or smaller budgets.
We usually start with 300 USD a day budget and have solid performance. CPCs in the first few days are usually high but drop by at least 50% after a few days. But yeah, it always depends on the vertical you're pushing. Finance, law or health has always a lot of crazy high CPC keywords.

If you know how to run paid ads, you can make almost anything work as long as the offer is not a dud.
Correct.
We focus more on bringing in new products and optimizing landing pages. Ads are pretty much auto pilot by now.

yIJZ46G.jpeg


This also applies to Facebook ads, but Googles optimization is better and more stable in the long run. For FB we only target geo, gender and age bracket and optimize for purchases. I remember the good old days picking interest targeting when it was the shit. Damn I feel old now.
 
Wait - 1500 PAID subs p/day?

Don't believe everything you read man, especially from someone just starting out. They always tend to over exaggerate and I have enough experience to call BS on this. It's easy to disprove these stats. Let's do some math:

- Average conversion rate for a SAAS company is usually 3% - 5%. Let's just assume he is getting 20%. This is someone buying, because according to him, he is getting 1500 paid subs a day, which is 45000 paid subs/month.

If you worked with a SAAS client before, you would already know this is ridiculous. This is not Chat GPT (and that shit is free lol).

Now, let's assume the CPC (cost per click) is $2, which it never is. It's usually way higher.

To get 1500 paid subs a day, he would have to get 7500 clicks per day (since we are using 20% conversion rate), which means he would have to spend $15000k/day. Which means, he would have to spend $450000k/month. And remember, this is only a standard display campaign according to @simplebuilder, which he only optimizes a few minutes here and there. This is not the entire Google Ads account.

I spend a few minutes every once in a while removing bad placements and working on optimising the copy and creatives.

If you are not familiar with display network or never ran a display network campaigns, you are literally talking about 100+ million sites that your ads might display on, so good luck removing the ones that don't work. You actually have to wait to get enough data for each placements to remove them anyway because you wouldn't know a placement sucks, unless you get enough impressions and low conversions to remove it. Otherwise, you might be prematurely blocking placements that might work.
This means you might get 2 impressions for 1 site and 100 impressions from another site. I think you get the point. It's a ton of work to get display network campaigns profitable.

Now, let's say the SAAS is charging $10/month for its subscription. This means:

1500 paid subscribers * $10 = $15000 per day in profit, which is $300k in profit per month in sales.

Now $300k in profit MINUS $450k = negative $150k! This means you are losing 150k/month from one single display network campaign.

This already sounds ridiculous and I exaggerated the crap out of the numbers. This also assumes that there is no processing fee (3%), other marketing costs, server upkeep, etc..

First, would you as a SAAS owner, hire a PPC guy with little experience to manage 450k of spend per month? From my experience - hell no. This is not a one man job and a smart SAAS owner would know that.

Second, if you are a SAAS owner - you don't have $450k to spend every month. Unless your dad owns a bank. I don't know.

Thirdly:
- All the numbers I calculated are exaggerated
- CPC fluctuate a lot, especially on display network on 100+ million placements
- Display network is notorious for being shit at conversions, that's why it's mostly used for branding and/or retargeting.

@simplebuilder I wish you all the best. That being said, please don't provide unrealistic numbers and assume that everyone here will eat it up, because it's an SEO forum. Some of us have been here for a long time and have plenty of experience in many IM areas.

This is like the new trend happening via cold email. They promise you the world. "How would you like 40 new leads per month? We did this for.." It's so unrealistic for a reason. To get responses, not results.
 
Don't believe everything you read man, especially from someone just starting out. They always tend to over exaggerate and I have enough experience to call BS on this. It's easy to disprove these stats. Let's do some math:

- Average conversion rate for a SAAS company is usually 3% - 5%. Let's just assume he is getting 20%. This is someone buying, because according to him, he is getting 1500 paid subs a day, which is 45000 paid subs/month.

If you worked with a SAAS client before, you would already know this is ridiculous. This is not Chat GPT (and that shit is free lol).

Now, let's assume the CPC (cost per click) is $2, which it never is. It's usually way higher.

To get 1500 paid subs a day, he would have to get 7500 clicks per day (since we are using 20% conversion rate), which means he would have to spend $15000k/day. Which means, he would have to spend $450000k/month. And remember, this is only a standard display campaign according to @simplebuilder, which he only optimizes a few minutes here and there. This is not the entire Google Ads account.



If you are not familiar with display network or never ran a display network campaigns, you are literally talking about 100+ million sites that your ads might display on, so good luck removing the ones that don't work. You actually have to wait to get enough data for each placements to remove them anyway because you wouldn't know a placement sucks, unless you get enough impressions and low conversions to remove it. Otherwise, you might be prematurely blocking placements that might work.
This means you might get 2 impressions for 1 site and 100 impressions from another site. I think you get the point. It's a ton of work to get display network campaigns profitable.

Now, let's say the SAAS is charging $10/month for its subscription. This means:

1500 paid subscribers * $10 = $15000 per day in profit, which is $300k in profit per month in sales.

Now $300k in profit MINUS $450k = negative $150k! This means you are losing 150k/month from one single display network campaign.

This already sounds ridiculous and I exaggerated the crap out of the numbers. This also assumes that there is no processing fee (3%), other marketing costs, server upkeep, etc..

First, would you as a SAAS owner, hire a PPC guy with little experience to manage 450k of spend per month? From my experience - hell no. This is not a one man job and a smart SAAS owner would know that.

Second, if you are a SAAS owner - you don't have $450k to spend every month. Unless your dad owns a bank. I don't know.

Thirdly:
- All the numbers I calculated are exaggerated
- CPC fluctuate a lot, especially on display network on 100+ million placements
- Display network is notorious for being shit at conversions, that's why it's mostly used for branding and/or retargeting.

@simplebuilder I wish you all the best. That being said, please don't provide unrealistic numbers and assume that everyone here will eat it up, because it's an SEO forum. Some of us have been here for a long time and have plenty of experience in many IM areas.

This is like the new trend happening via cold email. They promise you the world. "How would you like 40 new leads per month? We did this for.." It's so unrealistic for a reason. To get responses, not results.

It’s totally normal to be sceptical, but no. These results are real.

Here’s a screenshot of one of the campaigns (I’m running campaigns for five different SaaS products for this client):

subs.png


In the campaign above, we got 926 paid subscribers for $746, the cost per subscriber is $0.80, we are only paying for conversions (paid subscribers) in this campaign, and this product costs $6/month per subscriber.

Avg CPC is $0.02, and the conversion rate is 3.12%.

Regarding the optimisation of the campaigns, yes, the ads appear on tens of thousands of websites, but most of those have 1 - 10 impressions per site.

I usually remove a placement that either has:
  • An unreasonable cost/conv.
  • A large ad spend with no conversions.

Now, let's say the SAAS is charging $10/month for its subscription. This means:

1500 paid subscribers * $10 = $15000 per day in profit, which is $300k in profit per month in sales.

Now $300k in profit MINUS $450k = negative $150k! This means you are losing 150k/month from one single display network campaign.

This calculation is flawed; unless the product is really terrible, you’re not going to have a 100% churn rate in most cases.

This is not a one-man job

I am the only PPC specialist working on this account.

A Google Ads specialist at Solutions 8 (a really good agency) manages 20 accounts at their full capacity.

Second, if you are a SAAS owner - you don't have $450k to spend every month.

Some advertisers spend millions on paid ads per month and get far more than 1500 paid subscribers per day.

CPC fluctuate a lot, especially on display network on 100+ million placements

If you’re using a manual bidding strategy, you can set a max CPC bid limit; if you’re using an automated bidding strategy (like target CPA), you can set how much you want each conversion to cost. And if your target CPA is reasonable, you will reach it in time.

Display network is notorious for being shit at conversions. That's why it's mostly used for branding and/or retargeting

There is no one-size-fits-all here; I mostly use display campaigns for remarketing, yes. But in this instance, it is working flawlessly for driving conversions; we tried search campaigns, and the cost/conversion was twice as much as it is from display campaigns.

This is like the new trend happening via cold email. They promise you the world. "How would you like 40 new leads per month? We did this for.." It's so unrealistic for a reason

I’m not an expert in marketing B2B services with cold email, but what you stated above is simply not true.

I had a cold email project (B2C - A Legal grey area, I know) and was getting 10+ sales on Clickbank on a daily basis solely from cold email.

Cold email works for B2B as well (100% legal). People are building $50k+/month digital marketing agencies, with cold email being their sole lead-generation mechanism.

it's an SEO forum. Some of us have been here for a long time and have plenty of experience in many IM areas.

I’m aware of that, and that’s why I’m here, to learn from others and share my knowledge and experiences with others along the way.



Getting profitable campaigns for SaaS on Google is super easy.

The first Google ads campaign I ran was for an affiliate product (Clickbank).

The CPCs were high, and the margins were thin.

That’s because, with affiliate campaigns, anyone can get an affiliate link, build a one-page website, and run ads.

The barrier to entry is super low.

Now, with SaaS. Not everyone has a good SaaS product.

CPCs are reasonable, and getting a profitable campaign is very easy because the barrier to entry isn’t low.

Here is a screenshot of a campaign that I ran for a SaaS product targeting tier-one countries (USA, Canada, etc.):

Time-series-2020-06-14-2022-08-03.png



We got 102 conversions for $281.
The cost/conversion is $2.76.

The conversions, in this case, are for the free plan. This is their first paid campaign.

But according to their historical data, 80% of free subscribers end up purchasing the premium plan ($9), and the LTV of a paid user is $260.

In this instance, the campaigns were profitable from the get-go.

Even in hyper-competitive SaaS niches (like CRM), the LTV:CAC ratio from Google ads is reasonable.
 
Last edited:
Glad you cleared that up, but you had said you were getting 1500 paid subs a day man. In your screenshots - the first one - its for a 7 day period and you got a total of 926 conversions. That's 926/7 = 132 conversions/day.

Avg CPC is $0.02, and the conversion rate is 3.12%.

The conversion rate is on point, but it's crazy you are paying 0.02 in the US/Canada. That low cost traffic is never high quality so you can see why it's hard for me to believe those numbers.

A Google Ads specialist at Solutions 8 (a really good agency) manages 20 accounts at their full capacity.

I am well aware that PPC specialists in agencies are over worked. That's how it is these days. This doesn't mean they are doing a good job. It just means they got a lot of work to do and probably don't get enough time to get the best results.

I had a cold email project (B2C - A Legal grey area, I know) and was getting 10+ sales on Clickbank on a daily basis solely from cold email.

Good for you but from the sound it you stopped doing that, which means it's not worth it. I have been down that route before and understand 100% why its not sustainable or worth it in the long run. You are peddling someone else's product and other probably doing the same - plus Google Ads restricts sending traffic directly to "tunnel pages" so tracking can be a pain.

Cold email works for B2B as well (100% legal). People are building $50k+/month digital marketing agencies, with cold email being their sole lead-generation mechanism.

Lol. That's what you "heard". I have been in this game for 10+ years and I can guarantee you that cold email is not as good as everyone brings it up to be. You might get sales, but at the end of the day - it will never make a primary source of income for any agency. You can claim that some agencies make 50k+/month, all day, but in reality there is much more to it than that.

Anyway, good luck on your journey. I wish you all the best. Keep up the good work and I hope you hit the 1M in 3 years. Sending positive vibes to you!
 
you had said you were getting 1500 paid subs a day man.

Yes, we are running campaigns for five different SaaS products; 70% of the subscribers are for one “winning” product, which isn’t the one in the screenshot above.

it's crazy you are paying 0.02 in the US/Canada.

The clicks in the first screenshot are from Sudan; in tier-one countries, I’m getting much higher CPCs.

Good for you but from the sound of it you stopped doing that, which means it's not worth it.

Yeah, affiliate marketing with cold email was a low ROI project for me in terms of hours invested/day. I also outlined the other reasons why I stopped doing it in my thread in the laboratory.

I can guarantee you that cold email is not as good as everyone brings it up to be.

I really can’t see why it wouldn’t work; as I mentioned, most of my experience with cold email was with affiliate marketing but will try it out soon for selling B2B services and report my findings in my thread in the laboratory.

I wish you all the best. Keep up the good work and I hope you hit the 1M in 3 years. Sending positive vibes to you!

Thank you very much, man. I wish you success in your business ventures as well.
 
Back