I like AdSense for their pure simplicity. Google took the art of monetizing a website and made it a cut and paste activity. If you can get approved for AdSense (and comply with their policies), their program is a great choice.
While it is easy to get started with AdSense, it is very hard to get things right on the first try, particularly if you're new to their program. Most publishers earn less than they could from AdSense - this is particularly true for small sites. Speaking from personal experience, optimizing our page layouts and user experience to increase our AdSense revenue is one of the highest return ($$/hour) areas that we've invested in.
We've done no less than four major waves of optimization on our primary website and averaged a 50% improvement in monthly revenue from each project. These projects also results in a nice bump in user experience (longer sessions) and organic traffic (once Google saw the better customer experience). This article will walk through the highlights of what we learned in this process....
We're going to start by talking about the more common "leaks" in many sites, most of which can be fixed with minor changes in your ad selection and layout. Next, we'll describe the process used for advanced optimization - using data and art to understand how visitors and advertisers really think about your website. Finally, we'll talk about how you can automate this improvement process.
Before we dive into the science of revenue improvement, there are usually some basic opportunities we should address. Many sites have gaps you can fix for an immediate (low risk) gain. A good publisher in the AdSense program should think about three things:
The logic here is simple. Advertising optimization that boosts revenue at the expense of customer experience metrics or policy compliance is short term thinking. Good visitor experience has a strong effect on the long term growth of your traffic and earnings.
Mobile Screen Limits: This is a common one, since the policy was updated recently. Make sure your ads are far enough apart only one mobile banner shows on screen.
Large Units On Mobile: You are allowed to use a 300 x 250 Medium Rectangle on your mobile layout (and you should) but it has to be below the fold (this will trigger emails from Google). Large units such as the 300 x 600 are not allowed on mobile device layouts.
Drop Down Menus: You cannot overlap any piece of a Google AdSense unit. This is a common problem with navigational bars (overlapping top of content ads).
Content Violations: Be careful about content violations from user comments and third party ads. Google will hold you accountable for the issue.
Other Ad Networks: Be wary of policy violations triggered by other ad network units, such as features that will overlap an AdSense unit. I see this often on mobile placements.
Use Recommended Ads: Google identifies five units as recommended sizes; these tend to be more effective (in general) and have higher interest from advertisers (thus, more competition for bids). It's usually simple to swap out off-size ads for popular sizes and you will see an earnings bump.
Responsive Ad Formats: You can resize Google AdSense units using CSS media queries so you're always displaying an appropriate size for the screen. This is a tremendous improvement in customer experience and (frequently) revenue since it allows you to match the ads and layout to the screen.
Best Ad First in HTML: Ensure your best performing AdSense slot is the first one listed in your HTML code; this unit has an advantage in the bidding process which usually results in higher ad rates. Ideally this will be the largest unit on the page and one of the recommended AdSizes.
300x250 Below the Fold on Mobile: For the mobile version of your site, test converting your below-the-fold mobile banners (300x100, 300x50) to a 300x250 unit. Results of this change are frequently very good.
Target 3 - 5 Ads: Google invested a lot of energy setting the right number of ads per page and kept their guidance in this range for a long time. This may be the optimal number of ads for a typical web page (due to things like page speed or customer experience). If you're above or below this number, test closing the gap.
End of Content: With all the concern about visibility, many people ignore the end of the content. This is frequently a goldmine for clicks (since the reader is looking for what to do next).
Bottom of Page: It was apparently stylish to place a 728x90 unit at the very bottom of the page, below the comments and related posts. I'd test moving this unit up closer to the content.
Header Ads: Evaluate moving these units closer to the actual content. You could also convert the slot into a 300x250 Medium Rectangle and drop that into the content area.
Most of this isn't terribly complicated. You can make a significant improvements to a site in under a few hours.Stop for a second and go read your website. Better yet, have a couple of friends pull it up and watch them read the article (uncoached). What do they do? Which parts do they read? What do they skim?
Take the perspective of a new visitor. You show up at a website, clicking on a link served up by Google. The site may or may not have the answer to what you're looking for. You've probably clicked on a couple other results (varies by search type). How quickly do you get an indication that you're in the right place and the site can give you some high quality answers? Or are you stuck waiting for ads to load...
Advanced advertising optimization should always begin as an exercise in user experience optimization. The reason for this is two-fold. First, there's always a certain amount of cruft you discover in the process which can be used to improve the site "for free" (in terms of ad performance). Second, if you do your job correctly, the advertising is likely to be more impactful and you'll need every improvement you can find in UX quality to keep your audience engaged. Your site will also frequently see an improvement in your organic search rankings, increasing your total traffic.
You're going to need data to measure your performance. If you are doing the optimization effort yourself, you can get most of the information you need from Google Analytics and the AdSense performance reporting system. I usually focus on four key metrics:
I've got a couple of standard views I use for landing page improvement. Google Analytics has good reports for landing pages (under behavior) and page activity. One of my favorite views is to take the standing landing page report, filter it down to a specific page, and drill down using the "second page" field. This will show which page was driving engagement with your content. Look at where these links are on the page.
You're looking for topics and hotspots, things your typical user is looking for and places their eyes are likely to hang out while they're reading the site. Once you identify a hot topic, see what you can do to bulk up the links on the page with messages featuring content related to that theme. Similarly, you're going to want to test moving your advertisements closer to the hotspots.
See, the reader isn't looking for an advertisement. They are supposed to be looking at your content, which is why Google referred them to you. So one of several things is happening:
All advertisers aren't created equal. I usually break them out into three tiers:
Managing the balance of this activity across ad placements is critical to success. While we can't usually pick our advertisers, we can position our units to appeal to certain types of bidder and not others. For example, a brand advertiser will be highly focused on placement visibility. A CPC bidder is focused on paying a specific cost per click. We can steer the bidding by positioning our ad units excel in one area or the other.
Incidentally, the list of what an advertiser tends to complain about is driven by the above logic. Brand display buyers are very focused on the quality of the environment around the banner and get agitated by obscuring ads and other visual impairments. CPC buyers are worried about invalid and fraudulent clicks. Performance marketing advertisers are insensitive to the impression and click quality - the visitor converts or doesn't.
You can also use this logic to solve problems and reclaim lost space. Got an area that gets a lot of attention but also a lot of bad clicks? Put some CPA driven ads there, particularly ones with a nice long cookie window where you can get credit for future conversions.
Pull up your AdSense account and look at the performance report by ad unit. Try to look at this through the eyes of a buyer. Top brand advertisers are going to want to use premium placements (336x280, 300x600, 728x90) in areas with high visibility. You may have lattitude on click through rates. So there is a market for high visibility units with a slightly lower CTR. The AdSense bidding system will favor the click-hounds for placements with high CTR's regardless of their visibility. This audience is probably working with more common formats (300x250, 728x90). So that's the logical buyer for your end of article placements.
You can also do some basic window dressing of your inventory by allocating different types of placements to different ad units, AdSense channels, or even an entirely different ad network. The key here is to try to split out the different types of space. Position your high visibility spots to appeal to CPM buyers. Position the high CTR hotspots to look good to CPC bidders. And use performance marketing offers and native advertising units to reclaim space and attention (low CTR, bottom of the page) which would otherwise be lost. For example, we use Adversal and several affiliate networks for advertising placements that don't fit neatly within Google's terms and conditions. Or for placements which would interfere with the desired statistics we want our AdSense account to present to likely bidders.
After completing steps one and two, my next step is to sit down with my notes on the (suspected) visitor flow and a blank piece of paper. My intent is to sketch out a new test concept (on plain white paper) that improves the customer experience and allows us to integrate our advertising units more closely into that experience. Actually, I usually sketch three layouts - one for desktop, one for mobile, and one for tablet. There's one final piece of data I want to use in this process....
Google AdSense allows you to set up separate advertising units for each position on your page. We highly recommend you do this. This allows you to understand precisely how much each unit is contributing to your overall results. They have a performance report by ad unit. Key metrics on this report:
Big picture, you want to understand which advertising units are delivering a meaningful contribution to total page revenue and how they are accomplishing this. On the same note, you want to make sure the ad units are accomplishing this by delivering the expected results. Above the fold ads need to have good visibility. Below the fold ads need to optimize for clicks. And be ready to drop non-Google units in the areas that you're concerned have either invalid clicks or low attention (low visibility, low effective click-through-rates).
Now that we understand how we're going to measure performance, lets talk about the process of improvement. We drafted up a full guide to the common AdSense units, which walks through the strengths and weaknesses of the most popular advertising formats. Don't worry about the details at this point (bookmark this page as reference) - here's the reader's digest version. There are five recommended advertising units:
I've run many AdSense optimization efforts and the best opening moves always comes down to getting the basics right. Start making a list of things you want to test in a new design. A couple of good places to start.
If your site gets a lot of mobile and tablet traffic, you should repeat this assessment for different screen sizes. The standard AdSense reporting gives you results by "Platform", including click rates, earnings, and % visible - this can be invaluable in optimizing a size. If you spot one area or ad unit underperforming another, take a look at it on that browser size. A good way to see this is to use the Chrome browser's "More Tools" => "Developer Tools" option to see how your screen renders at different browser sizes. This was significant for one of my recent projects - while the site was performing well on desktop and table, mobile was poor (bad visibility). We wound up redesigning the page so it cleanly displayed all three ads in the mobile view. This translated into a big boost in CPM's for mobile traffic (~30% increase\).
Advanced optimization gets into page layout, ad sizes, and color themes. Effective placement and color schemes comes down to balancing small differences in contradictory goals. To be effective, the Ads must be similar to the content on the page (otherwise, ad blindness kicks in). But at the same time, ads need to be sufficiently different from the adjacent content to casually attract the reader's attention. Sound confusing enough? The process is a subtle balancing act. The best approach here is develop a couple of different concepts and run many, many tests (Google A/B testing to the rescue). While this can be very rewarding, expect to spend a decent amount of time setting up and tweaking tests.
It's a good idea to let an AdSense test run for few weeks before measuring it, especially if there has been a significant shift in click-through rates or view time. This is due to two factors. First, repeat visitors to your site will develop a routine usage pattern and learn to ignore ads in the regular places; moving the ads around will often produce a bump in the first 24 - 48 hours as your regular users notice them again and start to adjust their patterns. Second, AdSense has a safeguard to prevent publishers from profiting from bad clicks known as "smart pricing". If you have a sudden change in site behavior (more clicks, different page views, etc.) your value per click will drop significantly. If the change was legitimate and your visitors continue to perform well for the advertisers, your cost per click will revert back within a couple of weeks. But give the test a little time so you can work through the initial response from AdSense.
There's a delicate balancing act between ad placement and user experience. Placing too many ads will turn off your users and cause them to leave the site early (less pageviews and thus, fewer ad impressions). It's also good to remember that user experience plays into SEO rankings; so improving metrics such as bounce rates, pages per session, and average pageview time can result in higher traffic. This is where using the "experiments" function in Google Analytics helps. You can test multiple versions of the page and see which one delivers the best total revenue and user metrics.
The following are helpful in getting your optimization program moving in the right direction.
This optimization process has also been automated - a startup called Ezoic built a platform that automatically tests a large set of page layouts and ad placements to identify which ones increase advertising revenue. Some potential points of interest about Ezoic: