Disclosure: I have no affiliation with Adalysis and have not been paid for this review.
One of the longest lasting ideas I had stored in my Evernote idea notebook was to create an ad testing tool. Ad testing is such a simple idea, but a complicated process, and I was baffled that no one has tried to offer a fix for the PPC community. None of the big PPC tool vendors has come close to an effective solution.
However, during Hero Conf last year I saw a presentation from Brad Geddes where he spoke about ad testing (and briefly sprinkled in screenshots from his new tool, Adalysis). I was intrigued about what it could do, so I went in to check it out. At first I was taken back by the somewhat poor design (no offense, Brad), but as I started working with Adalysis, I quickly forgot about the aesthetics.
I have used Adalysis for about six months now, and I have to admit I’m overly excited to share my experience with you today.
How Adalysis Streamlines Your Ad Testing Workflow
Adalysis allows you to:
- Automatically see how and when new ads were introduced
- Data is only compared with time ranges where ads were competing
- Quickly identify ad groups without any ad tests running
- Great workflow for testing ads
Adalysis contains a lot of features, but the best is found in the fundamentals:
Automatically see how and when new ads were introduced: I recently wrote a blog post about how you can use labels to streamline your ad testing process. But if you’re using Adalysis, this is largely unneeded. Adalysis will automatically log when a new ad was enabled and show it right in the interface.
This is something I’ve always been baffled isn’t standard within the Google AdWords platform:
Another very helpful fundamental feature is that when you compare metrics between the ads you’re testing, Adalysis will only showcase the metrics from the time range when both ads were running simultaneously. Gone are the days with switching around the time range whenever you switch ad group. For me, this is a huge time-saver, and it removes a lot of hassles:
You also get a complete visual interface of how your ads are running. In the screenshot above, you can see the various metrics that you can test your ads against, some of them include:
- Click conversion rate
- Conversions per impression
- Cost per converted click (CPA)
- Conversion value / cost (ROAS)
- Conversion value / Impressions
All metrics serve their purpose, and there is something for everyone. I use the conversion value metrics for e-commerce campaigns a lot, while I focus more on reducing/maintaining a steady CPA for lead generation campaigns.
Great Workflow For Testing Ads
Have you ever jumped from ad group to ad group trying to find the ad tests with statistically significant data? If so, then you will be thrilled about what I’m telling you about next:
The screenshot above is a small part of the main Adalysis interface. In this interface, you can see the statistical significance within each main metric for all the ad groups you’re running. You can sort the table to see all the ad groups with statistical significant ad tests that need to be processed.
(Note the screenshot has been slightly edited to fit on this blog. You can see a complete picture here.)
Pro Tip: Adalysis will show you ads with “statistical significance” even though there are only one or two conversions. To better work with Adalysis’ other features, I recommend going to the settings and choosing a minimum of 3-10 conversions before you deem an ad test statistical significant:
Two Distinct Ways to Determine Your Ad Test Winners
After you have sorted all your ad groups by statistical significance, you have two ways to determine the winners:
- Manual one-by-one assessments
- Bulk pausing all losing ads as per your account thresholds
There are advantages and disadvantages of both processes:
Manual One-by-One Assessments
As opposed to the way you jump from ad group to ad group in Google AdWords, once you have sorted your table by statistical significance in Adalysis, you can review these by themselves.
Inside Adalysis, once you have reviewed one ad you can click the next button:
However, instead of seeing the next chronological ad group, you will see the next statistical significant ad group. You only have to see what’s relevant and don’t have to first wonder if this a relevant result.
The advantage of manually processing your ads is that you can get a sense for what type of ads are doing the best. You get to read and understand what is making your winning ads win.
Bulk Pausing All Losing Ads as Per Your Account Thresholds
A very neat feature intended for large accounts is that you can pause all losing ads according to one of the five metrics available:
This is a very speedy process, and I’ve saved plenty of time by doing. You might also notice you can see a breakdown of how much better your winning ads were performing than your losing ads.
By pausing all the losing ads in this particular case, we would have reduced our CPA by about $9 and increased the revenue in those ad groups by $2,100 over the last 30 days.
Compare What Would Happen if You Were to Pause All Ads Due to CTR:
One thing I really appreciate is the ability to showcase why it’s so important to track conversions and revenue in PPC campaigns. Below, you can see the breakdown of what the results would have been in the last 30 days if we had only been running our winning ads:
As opposed to before, where we could only see the data for the specific metric we were looking at, now you can compare what would happen if you paused your ads on the basis of other metrics.
The first time I saw this I immediately sent it to our entire account management team at White Shark Media. I have not in my seven-year career working in PPC seen a better breakdown of why you should be tracking revenue. Imagine this account wouldn’t have had any tracking, and we only focused on increasing our CTR.
Note that these are the same ads we are testing, but we are just grading them on different metrics. The difference between what would happen is astonishing, to say the least.
In the example above, if we had paused all losing ads according to the CTR metric we would have lost an incremental revenue increase of more than $5,000. This is only based on 48 ad groups out of an account with +1,600 ad groups. Imagine the impact the rest of the account would have!
Automatic Ad Testing Now Possible
Admittedly, this was one of the gems of my own description for an ad testing tool: automatic ad testing. The idea was that when an ad has lost a split test it should automatically get replaced by another ad from the queue.
Luckily, Brad Geddes and co. already created this tool, so you don’t have to wait for me to get started.
The advantages are many and you can save yourself a lot of time, but the biggest advantage for me is that I can gather all my creative energy at one time. I often felt held back by the fact that you couldn’t test 10 or 20 ads at the same time. It takes time to get into a creative mindset and once I had the next ad I wanted to test, I would often come up with 5, 10, 20 more ads that I thought would be interesting to test.
Yes, I would of course write these ad test ideas down and have them for next time, but now I can just add them directly to Adalysis and let Adalysis do all the work. It’s highly efficient.
Get Started With Adalysis
It goes without saying that I highly recommend using Adalysis. Both small and large accounts can take great advantage of Adalysis.
I personally believe that ad testing is the cornerstone of any Google AdWords campaign. If you don’t perform continuous ad tests, you are missing out on performance gains and potentially even experiencing a declining performance.
Disclosure: I have no affiliation with Adalysis and have not been paid for this review. Brad and Chad don’t even know it’s coming.
Featured Image: Created by White Shark Media for Search Engine Journal
All screenshots taken January 2015