Ad Creation Part 2
Ad Creation Part 2In part 1, I discussed how to create solid Ads that will generate average, to above average CTR (click through rate). Today, I will discuss how to perform split-testing and more Creative Ad Creation techniques that will have you hammering your competitors CTR.
The concept of Split Testing is an easy one. In one particular ad group, you have two different Ads. You run your campaign till about 30-35 clicks occur (which is usually enough time to ensure that you have gathered enough data to be statistically sure which one is better), and see which Ad has the better CTR. Your better Ad becomes the “basis”, and you create a new ad to try and beat your basis. You continue this process for the life of your campaign and you can very frequently double your original CTR, which is the exact goal of split testing.
Split testing is going to help you beat your competitors, but why? Shouldn’t they also be doing split testing? The simple answer is yes, they should be. The reality is that a lot of advertisers simply don’t do split testing or if they are, they get lazy and stop doing it, or do it very infrequently. Let me rephrase a previous statement of mine…”Split testing is going to help you beat your competitors, but only if you do it!”. You are going to need to develop a habit of continually tweaking your ads if you are going to hope to have higher CTR than your competitors.
However, there are many different ways to do split testing and make it more efficient and effective. I’m also going to give you several ideas of things to change that you might not have thought of. Some very small changes can produce large results!
The very first thing that you need to do to start split testing is change your campaign Ad Serving options. First, click on your campaign, then click on “Edit Campaign Settings”. Under “Advanced Options” you will see “Ad Serving” options, where your two options are “Optimize” and “Rotate”. Click on Rotate and hit the “Save changes” button. If your campaign has Optimize selected, then Google will determine which ad is better automatically and show that Ad more frequently. Sometimes it will show it 95%-100%, even though there is only a slight change in the Ads. For me, its too lopsided of a tool. “Rotate” will show your ads mainly evenly, so that you can gather the same amount of data between your two ads.
Also, another important note when split testing. When you are comparing the data between the basis and your test Ad, you need to be looking at the same time period. For example, lets say your basis ad was created in January, and then your test ad was started on February 2nd. Don’t look at the “all time” data for your ads, but rather February 2nd to the present day. Why? You don’t want to compare apples and oranges…you want to compare the ads under the same conditions. Perhaps you were paying more per click in January than in February, so you were getting naturally higher CTR in January, so the basis ad appears better than your test ad if you are doing all time data. So, if your test ad didn’t exist in January, then you can’t compare it with the data that you gathered for the basis ad in January! Test ads where all things are equal, except the differences in the Ads.
Another thing to watch out for. If you are using the content network, you’ll eventually notice that the CTR on the content network is much lower than the search network. Google knows this…they don’t count CTR on the content network towards your rankings and score on the search network, so don’t worry about that. However, if you are trying to split test ads, this will skew your numbers….with so many impressions from the content network, it will make the CTR on both ads pretty similar. For example, maybe without the content network your two ads have a CTR of 2.0 and 2.5 respectively, so one is obviously better, but if the content network is on and added like, 500,000 impressions, it drives the CTR down to .25 and maybe .26 respectively, so its seems that both ads are about the same. So, what you can do is have two adgroups, one where the content network is off, and the other where its on, to help determine which Ad is better. Or, if the #s are still substantially different from each other even with content network #s, then you can split test them with the content network on, as well.
What to Test?
If this is your first campaign, then I would suggest that you simply create 2 Ads, where you first start with two different headlines. Basically, you want to start split-testing the most important parts of the Ad first, since it is the most effective. When you find a good enough headline, then move on to the Ad body (the 2nd and 3rd line descriptions), and then move on to the Ad’s display URL.
Remember, the basic idea is to have two Ads, where there is only one difference between them (for example, two different headlines). After 25-40 clicks you see which Ad performed better, then you know that one headline is better than another. You keep your basis, and then create a new Ad which is the “test” Ad. In another 35 or so clicks you repeat the process.
A different strategy for the Basis
Split testing can take a long time before you start seeing results. If you are only changing one thing at a time, then to get a radically different Ad than your original Ad would take a long time. So, after you get a campaign or two under your belt, I would suggest this alternate way of creating your initial basis, so that you know that you are starting off with the best possible ad before continually refining it.
When you first create your campaign, instead of creating 2 Ads, create 5-6 substantially different Ads. Then wait till you get about 100 or so clicks, then analyze the ads. If there is one Ad that is substantially better than the others, then use that Ad as your basis. If there are 2 or 3 Ads that are performing similarly, then use those Ads for a while longer, and delete the other Ads that didn’t perform as well. Eventually you can be sure that one Ad is performing better than the rest, and use that one as the basis.
If you did narrow it down to 2 or 3 ads, and eventually decided on 1 Ad, then as your next “test” ad, test parts of Ads 2 and 3 together with your basis Ad as a good starting point for your test Ad.
In this way, you can get a great basis Ad that you will be continually working on for the next few months. You’ll be confident that you at least started on the right foot!
Split Testing for the Lazier Advertiser
If you are managing several campaigns (which you should be if you are a click flipper…you need to test 5-6 campaigns to make sure you find the most profitable ones!), then split testing 2 ads per adgroup every 30 or so clicks can be a real time taker.
If you are finding that you just don’t have enough time to do proper split testing, then I would suggest that you split-test 3-4 Ads at a time. If you do this, then you’ll have to gather more data to make it statistically sound, so wait for 60-70 clicks.
Split Testing Ideas
Here is a list of ideas that you can try when you are split testing your ads. Some of them are obvious, and some I have mentioned in Part 1. The important thing is that if you already have a basis and a test, just change one thing between the two ads, so you can know for sure which way is the best way for your particular affiliate product. Remember, different techniques are going to effect different demographics differently. Ads geared towards gamers are going to be a lot different than Ads geared towards wall street brokers!
- Use different Headlines (look at part 1 for ideas on different headlines).
- Use different descriptions (line 2 and 3). Now, this is a bigger issue, since you can have a lot more variety in your descriptions. I personally like to have Action Verbs be at the beginning of both lines, since it would draw the attention better. I also like having my first action verb be closely related to the adgroup, so that Google will bold it frequently for me (since hopefully it is in a users search). I have also heard, and agree with, that ad descriptions need to have a “sing-songy” element to them. They need to flow off the tongue well. This way people will have an easier time reading them and they will more frequently read the entire description. Remember, your first line DEFINITELY needs to be more important than the 3rd line since people will read the 2nd line much more frequently than the 3rd. You’ll notice that if you study writing Copy yourself (sales pages, etc), that they frequently talk about rhythm and rhyme. Say your descriptions to yourself outloud,…if it doesn’t flow well, then rewrite it!
- Benefits, Features. I talked about this one in Part 1. You can try different benefits, adding features, only listing benefits, re phrasing benefits, re ordering benefits, etc. There are many ways to say the same thing, and for sure one way is going to be more eye catching and pleasing to the ear than another!
- Use different Urls. Once you have a good Headline and Description, then start testing your Url. When you first create your Ads, I would first have the www. in your Url, with the first letter of each word in your Url Capitalized. If you are using SubDomains and SubFolders, you get to experiment with changing the titles of your SubDomains and SubFolders, for further customization. Just remember to change them on your webhosting account too so that they still resolve to your landing page! (also change the destination url as well…if you have your keywords in a spreadsheet program like Excel it makes it easy to change destination urls quickly).
- Add/Change/Remove punctuation. Try adding commas, question marks, exclamation points, periods, etc in your Ad text. You’ll be surprised at how such small changes can make a difference.
- Use different Capitalization. Think it won’t make a difference? It will! Try capitalizing all words, capitalizing everything but conjunctions (and, with, to, etc), and only capitalizing the most important words.
There are a lot of different things that you can do with your Ads, and so testing them all will take a split testing habit and persistence.
More Creative Ad Creation Techniques
At first, you might think that you have to create a unique and creative Ad that will blow users socks off. This is not true! Creative Ads are volatile and risky. Some will be complete failures and others colossal successes. That is why you need to make sure that you have a solid foundation of sound Ads already under your belt, so that these “normal” Ads will compensate you while you test more creative Ads. You don’t want your first $50 in advertising blown because your creative ad is a dud!
These next ad types take benefit of the natural tendencies that we have as humans. They relay to our fears, our desire to be included, etc.
Include Benefits in your Headline
The easiest type of creative Ad simply has benefits included in the headline as well. We already know that benefits are a fantastic way to describe our product, and telling benefits are usually better than telling features, in my opinion. So, why limit ourselves to only describing benefits on the 2nd and 3rd line?
Continuing on our “Dog Training” example from part 1, headlines with benefits in them could include ->
- No More Barking Dogs
- Well Behaved Dogs
- Own a Dog that Does Tricks
- He’ll Never Chase a Car
You get the idea. Simply create a headline that describes a benefit. You make sure that you use the technique when there is a benefit that far outweighs other benefits, or is stressed more in the affiliate product that you are selling, or that this benefit can’t be found in other products.
Using Fear in your Ad
With this group of ads, you actually aren’t trying to scare your users like a horror movie, but to shock them almost. These Ads have headlines that almost always need further explanation in your text. Lets look at a few examples ->
You’ll Never Train a Dog Unless you use this Professional Ebook from the Humane Society. Is Dog Training a Scam? No It Isn’t! You just need to follow our tips and secrets. A lot of these types of ads are going to be duds, but some will work fantastically! They are really hit and miss because you are losing a lot of benefits from standard ads, such as relevancy (you aren’t using as many keywords), and having your words be bolded by Google (unless you can manage to have an fear ad that has many keywords in the explanation).
However, they can work great! Just make sure you have a steady “basis” ad that nets good CTR so you can maintain your campaign without using too much advertising dollars on these risky ads.
I think we all know what this is, however if you don’t, this is where you tell your kids not to eat their vegetables and you hope they disobey you. There is also a natural human tendency to disobey people that works better in some markets than others. Lets look at some examples ->
Don’t Train Your Dog without our patented “Dog Training Secrets Revealed” newsletter! Quit Searching for Trainingyou can already find all of the tricks to Train your Dog Here! These types of Ads have the same advantages and disadvantages of the fear type Ads. You’ll lose valuable space in your descriptions because you’ll have to explain your headline. But there is never any harm in testing!
The Lemmings Approach
This is a great technique where you still get to list many benefits and features in your description lines, since it is geared as a change only to the Headline. These types of headlines answer the question, “If everybody is jumping off a bridge, would you?” Humans want to feel included, and it dissuades fears if they know that other people are doing the same thing. You can change your headlines based on this fact.
- How Thousands have Trained their Dog
- 21,983 Dogs Are Now Trained
- Hundreds Have Used These Secrets
As you can see, these Ad headlines rely on the lemmings approach to marketing. Just don’t exaggerate on the numbers too much. For instance, if I’m selling an ebook on training hamsters to jump through hoops of fire, no one is going to believe that millions have used the ebook to do such a thing.
I hope that you have enjoyed this discussion on creating winning Ads for your campaigns. Ads are a very important part of your Adwords campaign, and lots of time should be dedicated to them. Here is to your success!.