5 Ad Optimization Experiments that You Should Try

There is no magic when it comes to optimizing your ads. If you get lucky, you’ll immediately meet the best setup or strategy that works perfectly for your site — eureka! Otherwise, you will have to persevere while performing trial & error.

Here are some of the ad optimization experiments worth a shot:

1) Adjusting the CPM floors

Most ad networks will try their best to pay you at or above the CPM floor you set. If they can’t buy it, the impression will be passed back to the next best performing network or your passback tag. In short, if you set the CPM floor too high, you risk getting a lower fill rate with high CPM performance. If you set it too low, you will get better fill but CPM performance may or may not perform well — this is something you need to find out. Experiment!

Related Read: Apply Ad Exchange’s Recommended Minimum CPM

2) Passback Waterfall

You may have 1-3 or even 4 ad networks set up in the passback chain. Just keep in mind that while impressions are transferred from one network to the next, you could be losing some of them — we’re not in a perfect world. This is where impression auditing should come in. Contact the respective ad representative if you see large discrepancies. If they can’t resolve it, best to ditch them in the passback chain.

Related Read: DFP Order Report Check-In: 6 Reasons You Could See Impression Discrepancies

When setting up the waterfall, start from the best performing ad network in terms of CPM and put the 100% fill network as the endpoint regardless of its CPM performance. This strategy requires critical thinking. Analyze your results and revise setup if needed, then observe again.

Related Read: How to Setup Backup Ads for Google AdSense or Ad Exchange

3) Ad Targeting

Experiment varying tier setup. This is most efficient when using an ad server. If you use an analytics tool, study your visitors’ profile – demographics, interests, geo, behavior, technology, etc. Be clever with how you target or serve your ads. Experiment with frequency cap, countries, devices, even the type of internet connection being used, and the list goes on.
Related Read: How to Track Demographic Data on Google Analytics for Better Ad Targeting

4) Text Ad Colors

Do you know that you can play around text ad colors? Yes, why not try to experiment with the color schemes and see which best suits your site? Some publishers choose the Google default color so it stands out. Others use colors that blend with their site so users are encouraged to click, thinking it’s still part of the site content. Go try!

5) Ad Sizes & Placement

Wider ad sizes (e.g. 160×600, 300×250, 728×90) are easier for users to read and interact with. If positioned properly, these ad sizes can make you a lot of money. They also tend to outperform their counterparts — for instance, if you are using a 468×60 ad unit, try replacing it by 728×90 Leaderboard, if running 120×600, replace it with 160×600 Wide Skyscraper.

Related Read: Banner Layout Cheat Sheet: Top Banner Placements that Work!

Advertisers prefer larger ad units when designing their ads. Wider ad units are more likely to attract premium advertisers, thus making them perform well.
If you want expert ad optimization strategies implemented on your site, contact MonetizeMore and we’ll up your ad revenue in no time!

Related Reads:

Kean Graham

CEO and Founder at MonetizeMore

Kean is the resident expert in Ad Optimization covering areas like AdSense Optimization, DFP Management, and third-party ad network partnerships. Kean believes in the supremacy of direct publisher deals and holistic optimization as keys to effective and consistent ad revenue increases.

Get our latest ad optimization tips delivered to your inbox

Fill out my online form.


  1. Mazino

    Thanks for this information. I too is thinking of putting of all these to practice. I read somewhere that 2 adunit peforms better that 3..how true is that?



  1. How to Implement Retargeting in DFP - MonetizeMore - […] 5 Ad Optimization Experiments that You Should Try […]

Submit a Comment

Your email address will not be published. Required fields are marked *