The Global VC

More Users, Less Spend — 10+ ways to Trim the Fat & Save 5-figures on Adwords (and beyond)

Article image

You need to put your Adwords Display Network Campaigns on a Diet.

Actually, not a diet… just trimming the fat with aggressive liposuction surgery might be better.

Google Display Network (GDN for acronym lovers) is AMAZING.

With over 2 million Display Network sites that reach over 90% of Internet users worldwide (Source: comScore), there are a lot of opportunities to reach customers.

But with that power… there just as many dangers.

In fact, it’s so dangerous that I rarely meet a startup that’s doing everything I’m going to share with you in this guide.

Let me be clear….

The ENEMY in any Performance Marketing Campaign is one and only one:


If you eliminate WASTE — defined as Impressions, Clicks or Any Action that doesn’t perform  the way you want — you’ll only end up with just the things that work.

If nothing is working, then your problem is a bit more serious 🙁

But if you are post-product market fit, and find that GDN still isn’t working for you — then it’s because you are not aggressively cutting the waste or trimming the fat.

How Did I Learn All This?

Over my life as an online publisher and entrepreneur I’ve generated millions (3M+ and counting) of leads with Adwords, and 90% of those have come from the Google Display Network.

Those leads weren’t free.

Not sure if you’ve followed GOOG stock lately but with billions of dollars in revenue they seem to charge for the traffic.

And because I’ve never been “venture backed,” all that ad spend was my own money. I needed to come back quickly with a positive ROI.

I could never afford to spend $5 to make $1 because the 2 little guys in the picture below needed to be fed, sent to school, etc.

Felipe (2) and Pedro (4) are my sons. If I waste money on adwords they get less diapers, toys and chocolate 🙁

So I needed to figure out a way to make the GDN work for me…


  • Because Google Search sometimes can be incredibly competitive and expensive
  • And not only that… I can be limited (more on this on my article about 23+ campaigns that every startup should run)
  • I wasn’t going to allow my company not to use one of internet’s biggest channels

I knew that the way to tame the GDN beast was to Slice and Dice the data patiently and with discipline. Go into the account and do the boring stuff. Exclude placements, raise/lower bids, exclude devices, keywords, etc.

But there was a problem …

Every time I’d go into my Adwords account… the sheer amount of options, tabs, etc. was a bit overwhelming and for someone that like any other entrepreneur was always in a hurry -and kind of ADD- I never got nothing productive done.

So I remember the following story from my childhood. Yeap, I’m the nerd that used to read Greek myths for fun. Here is the full story but the relevant part is this:

Theseus announced to King Minos that he was going to kill the Monster, but Minos knew that even if he did manage to kill the Minotaur, Theseus would never be able to exit the Labyrinth.

Theseus met Princess Ariadne, daughter of King Minos, who fell madly in love with him and decided to help Theseus. She gave him a thread and told him to unravel it as he would penetrate deeper and deeper into the Labyrinth, so that he knows the way out when he kills the monster.

Theseus followed her suggestion and entered the labyrinth with the thread. Theseus managed to kill the Minotaur and save the Athenians, and with Ariadne’s thread he managed to retrace his way out.

Theseus took Princess Ariadne with him and left Crete sailing happily back to Athens.

Here’s how you kill the monster, the minotaur… THE WASTE.

So I built a little “thread” and algorithm that allows me to go in my adwords account (or my employees) do the optimizations, kill the monster, trim the fat… and go back without getting lost.

Here’s how to cut the waste from your PPC campaigns, step by step.

1. Remove Adsense For Mobile Apps

How Does Google Pay for Acquisitions?

They make it a default option on Adwords.

Most people will never take the time to optimize and they’ll trust it to the conversion optimizers. But you know what… when I sell education courses I have enough info to know that most of the traffic from adsense in android comes from games and rarely converts.

Most of the clicks are “accidental clicks anyway” so if you are on a budget I strongly suggest excluding the following placement from your campaigns.

Again, do this only if it makes sense for your market… but in most markets it does.

The guys at “Wordstream” make a good case here.


You can start your campaigns with this setting or give it a go and if then you see that mobile apps are not converting you exclude them after.

2. Exclude Particular Apps

Suppose you still want to appear on mobile apps, well now go ahead and see what apps are converting and what apps are not! Then exclude those.

Look at the last 2 … less than 1% conversion rate after more than 600 clicks!! I know for a fact that this traffic is not going to convert and is giving me expensive signups!

So if you don’t want to exclude Mobile Apps all together at least use this type of exclusion to remove the apps that are causing WASTE in your account.


This one’s not for beginners. For this one, you need data in order to know which apps to exclude. This is one you need to grow into — don’t start here.

3. Add Site Exclusions

Do you really want your ads showing up in pages where there are Crime, Death and Tragedy, etc.?

Well… if you sell coffins that might be a good idea, but not for everybody.

Do you really want your ads showing up in parked domains?? Maybe in remarketing campaigns you do, cause you are following the prospect/client around no matter what… but as your first acquisition touch point?

The point is… use Site Category options to exclude types of sites that won’t fit your advertising goals.

This is especially true if you are on tight budget. (Who isn’t?) Do this from the get go.

Lots of people don’t exclude this and they end up spending large portions of their budgets on parked domains 😉

Sexually suggestive? Think about that twice… Look at the guys at Eat24


You can start your campaigns with this setting or give it a go and if then you see that mobile apps are not converting you exclude them after.

4. Create a Black Book or Black Ledger of Placements

This one is one of my favourites.

We analyze data from all of our campaigns in all of our websites and we figure out what sites did not convert for ANY of our websites or products. With that data we create a “Black book or Black ledger” that we exclude from EVERY campaign we run.

I don’t need to spend money twice on torrent, software, ebook site or coupon site that it’s keyword stuffing and SEOing everything just to show adsense ads.

So after you have enough data… go ahead, create your Black Ledger of Placements and add it as site exclusions in all of your campaigns. That way you won’t spend the money twice to find out that there are some pretty shitty internet sites.


You’ll need months of data probably to build a good black ledger… but once you built it I always start my campaigns EXCLUDING those placements that never worked for me. No need to spend money twice to learn my lesson.

But if you want, and you think your campaign is completely different and this time will work on the spammy coupon site… you can exclude it afterwards.

5. Remove Demographics That Don’t Make Sense for Your Product (Age, Sex, Parental Status)

This one seems “so obvious” that I don’t get why more people aren’t doing it. All the more reason why you should implement, now.

There are some websites and products who clearly target a specific demographic. In other words, there is no chance in hell that people NOT in that demo will ever become customers.

For example… if you are selling a product to women, you can “Exclude” people that GOOGLE KNOWS FOR FACT are male.

Yes, sounds extremely simple, but many of your competitors aren’t doing it.

Here’s another example:

If you are selling something that requires a Credit Card Transaction, then you would want to EXCLUDE all leads that don’t have a credit card.

For example, in LatAm, people under the age of 24 typically don’t have a credit cards. Likewise, it’s equally unlikely that a 65+ retired guy will enter his credit card on my website.

Same if you are selling diapers… probably a lot better to exclude people that Google knows ARE NOT A NEW PARENT.

These are all ways that you can use simple demographic segmentation to do nuanced exclusions and save a LOT of money.

If you are less risk averse, then you can leave “unknown” on… but if you are on a tight budget you can exclude those also!

[?] Unknown: This group includes people we haven’t been able to associate with any gender/age/parental status.


You can start your campaigns excluding (or targeting) some specific demographics or see how it goes and exclude them later.

But, if you are selling to specific demographics and your budget is tight, I recommend starting with this settings.

6. Remove Your List of Leads/Customers

This is another “obvious” one that I also don’t get why most people are not doing it.

Let’s say you are “Groupon”… you need to get “Leads” / “Email Addresses” to your list so you can offer them something. Cool.

You set up a Campaign SPECIFICALLY FOR THAT and it’s working.

So, if someone becomes a lead and signs up… shouldn’t you EXCLUDE THEM from seeing ads that send them to SIGN UP AGAIN?!! Of course you should.

Yes, you can set up Remarketing Campaigns to get them to come back, to buy more… but not to SIGN UP… because they already did that.

In another scenario, let’s say you are a cool SaaS company growing at “Slack Rates” (good job!).

Thanks to your excellent acquisition campaigns, people are becoming paying customers left and right.

Shouldn’t you exclude those paying customers from seeing your ads to buy and your Christmas Promo with a discount?!! Of course you should.

You can target them to come back, to use the app… but please (please!!) exclude them from the original campaign as soon as that campaign serves its purpose.

The screenshot below shows you exactly how to do it.


I normally start with this settings. If someone signed up for my “Offers Newsletter” or my “Free Trial” No need to keep showing them ads to do that “again”.

I might show them ads to “come back” or “buy” or something else… but I’ll exclude this guys from the campaigns from the start.

7. Remove Geo Targets

Living in a “Developing Country” (3rd World is a term that our friend Donald Trump would use… ignoring that it’s the “same world” we live in) I can tell you that not all “Markets” or “Countries” are created equal.

I know for a fact that people in Venezuela and Cuba (although things may change from now on for Cuba) will have problems paying for my product.

They’ll sign up for the free trial or the free lessons, but then they won’t pay. So I exclude those type of countries from the start.

Also if you are selling physical goods and you don’t ship to some countries or cities in your country… is better to exclude them!


For the obvious cases (leads in that country can’t pay, or you don’t ship to that place) I start with excluding those geo targets from the get go. Depending on your risk tolerance if you “think” that something is not going to work you might wanna exclude it too.

8. Hours/Days of Week

Not every day or hour performs the same for every product. In my case the performance is quite stable for days of week.

But again, if you don’t ship or sell on some days… you might wanna stop the campaigns those days. You’d be surprise how many people don’t look at this!


Whether you start here or grow into this depends on a few factors.

For the obvious cases (you have a local store and you are not open on Sat, Sun) maybe you want to start here and use time exclusions right away. But, then again, maybe someone goes to your website on Sat, Sun and then they visit your local store on Mon. So you have to try.

9. Devices

Technically some people won’t consider “devices” a dimensions because some of the settings to exclude them are at the bid level … but I don’t care.

Be aware that people in computers, mobile and tablets convert at different rates!

I’m amazed at how many people:

a) still don’t have a mobile responsive/friendly fast website


b) even if they don’t (because they did not have time to build it or whatever)… that they don’t exclude mobile users from their paid campaigns!!

So please… if mobile traffic is not converting for you — for whatever reason — just make sure you exclude it!

In AdWords you can do that that by bidding -100% for mobile.


If your site is not mobile friendly or fast, you should exclude mobile from the get go. If not, try it out… and see how it goes. Instead of killing mobile right away (huge traffic is there) you can also make some effort to optimize it before turning it off.

10. Other Ideas

I don’t want to make this longer than I should. I think you get the point.

There are a couple more dimensions you can look in GDN but I feel these most people do.

Basically you can:

  • Remove Topics that are not Working
  • Remove Specific Placements with bad performance
  • Remove Keywords (or put them as Negative) that bring placements with bad performance, etc.
  • Remove Ads with Bad Performance

But these most people do these. On the other hand… the ones I talked here I see very few startups doing it regularly.

Some Statistical Thoughts and Numbers to Take Into Account When Optimizing

But how many clicks, conversions, etc. do I need to wait BEFORE I exclude something?

Well… I’m glad you asked.

Of course the answer to this falls in the category of “It depends” but let me give you some useful guidelines.

Depends on what? Basically on your conversion rate and the statistical confidence that you want to have.

But let’s think about it in the following way.

If your average conversion rate of a campaign from click to free trial signup is 15%… and if you got 30 clicks and no signups.

To get to 15% conversion rate … the next 5 clicks should be a signup.. Possible, but highly unlikely. So I’d kill that.

I never kill something BEFORE 30 Clicks.

Now if you are optimizing towards something that has a 1% conversion rate on avg. You expect 1 sale every 100 clicks. But you know… that’s the avg. So in that case I normally wait up to 4X to 5X the clicks. So if I had 400/500 clicks and no sales or just 1 or 2 with a negative ROI then I turn that dimension off.

This is not 100% based on math but it’s good common sense.

Applying This To Other Platforms (Facebook, Taboola, Outbrain, Bing…)

Of course this approach to reducing dimensional wastage is not only for adwords. You’ll find that this can work in Facebook, Taboola, Outbrain, Bing, Yahoo and many others.

I just wanted to give you a sense and dive deep in Google Adwords cause I believe GDN is one of the most underutilized traffic source in startups because of these pitfalls.

Here’s a simple Facebook Example… You can go to:

Male and Female


You get the point. With Tools like AdEspresso (a 500 company… I love you guys!) you can zero in into a lot more details.

Note For Startups With Agencies/Freelancers/Outsourcers Running Their Adwords Accounts

I get you are founder with A LOT going on. I still think marketing should be a core skill inside ANY company and outsourcing it I believe it’s a mistake… but I get it… sometimes you do what you can. But… Make sure you delegate things to knowledgeable, hard working people.

If they are not doing these things… Time to have a serious conversation with them 😉

My Last Ask…

If you’ve got any value out of this blogpost… please share it.

On Twitter

Share on Facebook

Share on LinkedIn

Email it to people:

Upvote on GrowthHackers

Upvote on Inbound

You’ll not only save them 1000s of Dollars by doing it and probably open one of Internet’s Biggest Channels To Them (The Google Display Network)… you’ll also make me super happy 😉

So yeah… please share it and feel free to reach out on twitter @juanmartitegui specially if you did some of these analysis on your account and you found some REAL WASTE & SAVING OPPORTUNITIES.

500 Global Team

500 Global Team