Advanced Pay Per Click (PPC) Certification Program

Certification Training
2370 Learners
View Course Now!
42 Chapters +

Ad Copy Testing Tutorial

5.2 Introduction

Hi, I'm Brad Geddes, the author of Advanced Google AdWords, the founder of Certified Knowledge and the PPC Faculty Chair for Market Motive. In this module, we're going to look at an overview of testing, look at how much data is necessary to run a test, and then finally look at how to do ad copy testing.

5.3 Overview of Testing

So with testing all you're really trying to figure out, is A better than B? Is B better than C? So all we're really trying to do here, if you have two or three or four ideas for ad copies or a few different landing pages, which one helps you reach your advertising goals? Testing's a crucial part of optimization. Without benchmarking and testing your data, you can't know how to improve your campaign's performance. Through the use of testing you can determine the most effective means to reach your advertising goals. Now it's common when you first start testing to have some dramatic results. It's often due to the fact this is the first time you've really thought about testing and credibly thinking about how ads, keywords, offers, and searchers interact with each other to help you reach those goals. Once you start testing for a while, it's more common to see incremental results and not dramatic results. With most paid search campaigns, you can test a lot of data. You can see the average CTR of a headline by individual city. There's a lot of data points you can really test based upon what's the most important metric to you.

5.4 Most Common Split Tests

But what we're really going to focus on, is conversion rates, cost per conversion, and click-through rate. And we'll look at those by ad copy and by landing page. In this module, we'll look at it by ad copy. In our next module, we'll look at it by landing page. But, one of the problems with testing, once you really start getting it's event and do some advance testing is what leads to more profits. A high conversion rate and low click through rate, in other words not a lot of visitors but good conversions from them, or a high click through rate and a low conversion rate. So lots of visitors but not a lot of conversion per each one. So when dealing with mixed metrics such as this we'll eventually get to my favorite metric overall which is profit per impression or profit per click in our advanced testing modules. One of the big questions Is how much data do you need. In the next few minutes we're going to look at some overall rules for how much data you need with testing. However, if you're a statistical wizard, there's another way you can get confidence data for your testing. Excel has a plugin called Data Analysis. Once you add that plugin to Excel, and it's a free plugin from Microsoft, you can then run ANOVA, which is analysis of variance. And you can get in other confidence factors for running the data. If you are an Excel wizard or you need a lot of testing and you want to spend a few hours learning how to run a ANOVA and how to get some of these other confidence factors, that's the best way of getting the data. However, if you're not, and you just want some general rules to look at, we'll look at some of the basic information of how much information you really want.

5.5 How Long Do You Want to Test For?

So the first question's, how long do you want to test for? Well, at least a week. Buying behavior is different on weekdays than weekends. So even if you get 1,000 clicks on an ad copy in a single day. It is still better to run the test longer to accommodate for the fact that weekdays are often more B2B behavior, and weekends are B2C behavior. However, a month is just fine. Often you'll see the beginning of the month and end of the month have lower conversion rates, the middle of the month has higher conversion rates. And again, this depends greatly upon your industry. It is just fine to do monthly testing. Or ideally, three buying cycles with at least one month. An easy way of doing testing, is the beginning of the month, sit down and walk through the process we'll look at, create some ads. The end of the month, go ahead run an analysis of those ad copies and see how they performed over that month of time. If you don't have enough data, it's better to wait a little longer than just to finish running a test to say you've done some testing. What you don't want to do is make decisions on inaccurate data.

5.6 How Much Traffic and Conversion?

Next question is how much traffic do you need? At least 300 clicks an ad 500's better 1,000's definitely ideal so just make sure your confident in the information. Finally how many conversions do you want? At least seven an ad, 15 is definitely better. What you really want to make sure of is that a few people can't make dramatic differences in your click-through rates and conversion rates. So for instance, if you have 100 clicks and one conversion, you have a 1% conversion rate. If the 101st person were to convert, so now you have 101 clicks, and 2 conversions, your conversion rate just went from a 1% to a 1.98%. Essentially, if you did ROI bidding, your bid would double based upon one person. So there's a lot of calculators online that you can see have one conversion off of seven clicks and the other one has five conversions off of 13 clicks. They will tell you you have enough data to make a decision. No you don't, you want to make sure that just a few people can't dramatically change your information.

5.7 Meet All 3 Requirements

Now you do want to meet all three of these requirements. Minimum amount of time, minimum amount of traffic, minimum amount of conversions. Now these are overall rules. If after a month of time you have one ad copy that has 500 clicks, but it has 37 conversions. You have another ad copy that has only 400 clicks, so not quite your minimum number, but one conversion or zero conversions, you can make some assumptions. And that's what using ANOVA and some of the other statistical methods are better for looking at the data. However, if you have enough time, traffic conversions, make sure you have minimum amounts of data and that you trust it, that just a few people aren't making big changes in your conversions.

5.8 Low Volume Accounts

If you've looked at this and you're thinking to yourself, you've got a thousand clicks a month in your account and no ad gets more than 50 clicks. I can't do testing. Yes you can. You can do test with low volume accounts. Now we'll look through all the step by steps to doing this in a future module. But at a high level all you're going to do is write two different tag lines, two calls to action, two different discount codes, and then put those lines in multiple ads across multiple ad groups. Then you can just use a pivot table to then combine the data across the ad groups. And look at how different calls to actions or different taglines are working from a conversion standpoint, cost per conversion, conversion rate, etc. So you can do ad copy testing, regardless if you have a high volume account or low volume account. The only difference in volume is how quickly you can make decisions in your testing. And whether you can do unique ads where you're looking at individual ad copies against each other versus just two taglines or calls to action against each other across multiple ad groups.

5.9 Split Testing & Ad Optimization

There was one important setting in your AdWords account for doing testing. Now by default, and this goes to how Google displays your ad copies. So by default, Google shows optimize for clicks, where they show the ads that will get the best click through rate the most. With testing, often, we're looking beyond click rate. We're looking at profit information. We're looking at conversion rates. So there's another setting called, Rotate. This shows your ads evenly. So you'll have ad copy one, then two, then three, back to one again. Now Google's had a bug for years, in that when you turn on Rotate, your ads will not have identical impressions. You'll see ad copy one has 450 impressions, ad copy two has 525, ad copy three has 500, they're directionally similar, but they won't be absolute. Now Rotate does effect every ad in your campaign. So just note, you're changing a campaign level setting, it does change everything. But when you're testing across all your ad groups, then Rotate's a better setting to use. And you should be testing across your ad groups all the time.

5.10 Ad Copy Testing

So next we're going to look at Ad Copy Testing. Before we look at how to do this we need to look at the ad's writing, the different types of messages, and pick out some different overall ideas for tests. So first remember, an ad copy is the only part of your account a consumer ever sees. Consumer does not know what your geographic settings are, your keywords, your landing page. All someone knows is your ad copy. So your ad copy's goal is to draw attention to itself. Tell someone what their going to find after the click. So set those expectations for landing pages. Now often to increase that CTR and increase that conversion rate, you want to make an emotional connection with individuals. So there is some universal truths about people. Think, what do people want? Everyone wants more time. Everyone wants more money. Everybody wants better health, better appearance, promotions. There are also things that nobody wants. No one wants more work. Most people have enough now. Most people don't want to be ignored, they don't want doubt or criticism. So when you start thinking about what people really want or what we avoid, look at your ads. Are you using adjectives and adverbs, nouns, verbs that help to create that emotional message? Something you really should think about when writing your ads, is there any emotion? Or are they very boring ad copies, because they're just factual statements. They don't really reach out to a user and say, hey. We're going to make you more confident in yourself if you were to buy this product. We're going to take away unacceptable risk by showing you how to make proper trades on a dynamics stock market. So think first about emotional connections. However, there are some things that you can test.

5.11 Test Pricing

One of them is prices. Prices are really good to test in price conscious industries. Highly competitive industries. One that is large price ranges for products. One comparison shopping queries. Now when these three ad copies, we look at the prices. There's already a winner without even having to this task the majority of the time. The top ad copy creates selection from 89 to 199. The second one creates selection from 89. The third one from 89 and up. What's important with numbers is that people focus on the number most beneficial to them. When you're buying a product, the most beneficial number is the lowest price point. So, top ad copy, from $89 to $199, means it's a $199 product. The second one. It's from 89. It doesn't make inference that there's more expensive sales compared to that third one from 89 and up. And this example, generally ad copy two will do better before for you even test. Now, if you look at an auction, and there's two completely different prices on the page, you can dictate how consumers feel about this price point by the way you write this ad. So right now we have $119 and $199. These prices are so different it's hard for consumers to know what the actual price point is. So if you walked into this auction, and you say it's $189. So the price has become $119, $189, $199. Well, $189 and $199 validate each other. $119 looks like some cheap knock off. $189 is a little bit cheaper, often it'll get the better CTR. Now, if instead you said it's a $129 product, so the ads are $119, $129, $199. Well now $199s are super expensive, who are they? $119's a little cheaper than $129, but those two numbers validate each other, $119 ends up with the better CTR. So if you'd had left your price at $129 instead of $189 you'd have lost $60 dollars a conversion. So it's not that you want to be the cheapest on the page. What you want to do is validate the most expensive ad on the page. Or something more in line with how you want the pricing to look. Now, the issue with Google is, every time you hit edit on an ad copy, and change anything, a price, a call to action, a source code off of a destination URL, anything about the ad, you change it and hit save. In Google's mind you did not edit an ad copy. You deleted one and created two. When you create a new ad, it may go live on Google right away. It will not go live on the search partners or display network until it's approved. So if you change your ads every couple of days, your ads will never get very good exposure. If you do want to play with numbers, but don't want to edit your ads all the time, use discounts or discount ranges instead. Now, with these two ads, there is already a winner the majority of the time. The first ad copy is 25 to 60% off. The second ad is 25% off. 60% is much higher than 25, the first ad will probably do better. Now, it might be you only have one item at 60% off, everything else is 25. It doesn't matter. The first one will generally get the better CTR because it's more beneficial to the customer. Now discounts often increase conversion rates. You're selling stuff at a discount, consumers like discounts, fantastic. However, just because they increase conversion rates doesn't mean you should use a discount. There's a little more math you should be doing with discounts to make sure it's beneficial for you to sell at a percentage off as opposed to just selling at regular price.

5.12 Discounts Chart

So what we want to do is just build a simple chart that shows as our discount goes up, our conversion rate goes up, of course our sales go up, but does our profit also increase? So in this example, if we offer no discount, we make $900 profit. If we offer a 10% discount, our conversation rate doesn't change that much. So all that happens is our profit goes down. Often with discounts your average order value also goes up. So this is where you need to not look at just how much did you make on a sale, and your conversion rates and see if your profit's higher. But you also need to take into an account average order value. In this example, if we look at offer B we offer 10% off on that one. Our commercial rate goes up just a little bit, so our profit's higher than A, we'll make 818 versus that 750, still below our baseline of what we would make without any discount. Now, by the time we get to 20% off there's enough increase in conversion rate that we get enough more sales. We are making more profit than our baseline. So with discounts just do a simple chart like this. Look to see how your spends change, your conversion rate changes, if your average order value also changes, and then look at your profit metrics. There are times that adding a discount, much more profitable, even at 10% or 5% off. There are other times that you're better off not offering a discount at all. So just do a little more math when offering discounts.

5.13 Calls to Action

Now, when your focus is conversions, your call to action is really important. What do you want someone to do after the click? You put it in your ad copy, you echo it on your landing page, it often increases conversion rates. because if someone doesn't want to do that action, they're not going to click on your ad. However, most calls to action are boring. Call for an appointment today, contact us 24/7 for your appointment, schedule a visit. Well a boring one, subscribe to our newsletter, versus sign up for powerful marketing tips, receive your free secrets today. Now you're giving someone an incentive beyond just sign up for something, to why they should sign up with a benefit message. So if you have a call to action, see if you can combine it with a benefit message when writing your ad copies. Now the reason we do a search is we want to learn the answer to a question, which is a piece of information. So just looking to see, what does better, an informational ad versus a heavy call to action ad? Easy test, notice top ad copy, Bluetooth Headset Sale. Easily discover what headsets are compatible with your phone. If someone has done any amount of research about Bluetooth headsets, they know there's not a compatibility issue. This site will not do very well. However, if they're at the very beginning part of the buying cycle and they just search for Bluetooth headsets or how to buy a Bluetooth headset, it's a decent ad copy. Later on in the buying cycle, it's a poor ad copy. How do you buy exercise equipments? Usually people go buy it month one, month two they might use it, month three it's in the garage, month four it's for sale on eBay. So Justin Abbott says, learn two facts that will keep you using our treadmills. Break the cycle of selling this stuff off, and here's why you're going to keep using it after month three when it's normally in the garage. Third ad copy, chicago Real Estate. Moving to Chicago? Easily compare neighborhoods. Now in many cities, real estate agents cannot legally compare neighborhoods. If you were an MLS database lookup site, and you weren't an agent, you can compare neighborhoods, if you have access to information others don't. That's great benefits to put in ad copy. Now the overall rule of local is, the more someone know about an area, the more specific your ad should be. The less they know, the more reassuring your ad should be, you can help them out. So this top ad copy, Wicker Park Real Estate, Northern Chicago Specialists, Find Your Dream Condo. Now if someone has just searched for Chicago real estate, or how to move to Chicago, it's a bad ad copy. It calls out Wicker Park, which is a neighborhood. It calls out it's northern Chicago, not a good ad copy. If the search query was Rogers Park Real Estate, Wicker Park Real Estate, Gold Coast Real Estate, which are all neighborhoods of Chicago. Now it is a good ad copy, it's meeting someone further in the buying cycle, because they know enough to know what Wicker Park is. They're just moving there, it's a poor ad copy. So the Second one, Downtown Chicago Hotel Located in the Shopping District, Easy access to All of Downtown! That's a good non-local ad copy. Non-locals want to know, are you by shopping, the amusement park, Navy Pier? Locals don't want to know that. Shopping District in Chicago's huge. They want to know Michigan Avenue or Rush Street or the third ad copy, Chicago Bluetooth Headset. Touch the headsets before you buy, Located in the Water Tower. Now, unless you're a Chicago resident or visited the city several times, you have no idea what the Water Tower is. It's a landmark in the city. Locals know where it is. So, now it's a good ad copy. Tell someone exactly where you're located. Bad ad copy for a non-local. So when writing local ads also think not just how much do they know. But where are they in the buying cycle, because that does change how much they know as well. And the way to look at where they are on the buying cycle is by their keyword intent. What query did they do?

5.14 Questions

Why does someone search? To learn the answer to a question. When you're ad copy echoes their question, people are like, oh wow, that looks great, you understand what I'm looking for. Now you don't need to add the entire statements to your ad copy. The act of adding a question mark, puts the understood, are you, in front of the line. So are you visiting Chicago? Are you looking for a new home? Do you need a new headset? So an easy test is one's a question ad, one's an informational ad, completely different types of ads. See what one does better for you. Display URLs, just as important to test as well. Display URL, click rate is a component of quality score. It's not something you can focus on too much, but is a component of quality score. But it tells someone where they're going after the click. People like to look at display URLs. If you show someone a display URL of Electronics.com, what does that mean? Nothing. Electronics.com/BluetoothHeadset, now they know the product they're going to on the site. Now, often display URLs include product names or service names, sometimes geographies, you can also make other synergies with the ad as well. The top ad copy. Need a new headset? Compare all Bluetooth Headsets, Electronics.com/Compare. Now you're showing it's a comparison page, or slash sales. So good things to do are not just try a product folder or a service type, but also other ad copy synergies. Now the way you write your URL is important as well. Now Google normalizes all the letters in URL to lower case. So this does not work on Google at present, it does work on many other page search engines. Also very useful if you are advertising in print or on television, just uppercase the first letter of each word. Because if you don't, and your URL has multiple interpretations, you're leaving that up to the user. The fact of upper casing the words, of upper casing the first letter of each words removes the interpretation from the user. Are you a sex change site or a commenting exchange site for blogs? Big difference, same letters.

5.15 Testimonials

Testimonials and reviews, fantastic. Because a testimonial and a review helps shift the consumer's mental state from surfing the web in an isolated state to realizing they're becoming part of a community or buying a product that others have already enjoyed. So users prefer Jawbone three to one, read customer reviews. Read thoughtful headset reviews left by expert and Bluetooth users. See what others have found is the best Bluetooth headsets. Now normally you cannot use absolutes in ad copy. You can't say best, number one, unless you have a third-party verification. If you've won a Best of the Forbes Award, you're number one on JD Powers rated, you may have access to words in your ad copy others don't. Now in cases like that, the verification has to be on the landing page. However, it's a good idea to put your JD Powers number one rated on your landing page, it's a good seal to increase credibility, and it lets you use an absolute in ad copy.

5.16 Negative Information

There's a reason when you turn on the news or pick up the newspaper, the headlines are negative. Negative information stops us in our tracks, we look. We don't want to fall victim to something that's commonly known, so be careful of Bluetooth. You think it's safe? Headsets waste money. Don't buy non-compatible headsets. Or don't buy a headset without reading our secret shoppers guide. These are often affiliate ads. They're not always, but often they are. Now, negative ads often increase click-through rates. Because they stop people, they want to learn more. They do not always increase conversion rates, so even a negative ad versus a testimonial ad, good test to run. Now before you ever run a negative ad, have a corporate branding strategy in mind. If you want to be the family friendly amusement park, don't run negative ads. Everybody knows someone who won't vote for a politician because of their negative campaign. These ads will be seen, they will affect your corporate brain perception. So, if you use a third party to run your ads, just have an overall policy. Don't run negative ads. We want to be the friendly family brand. Or we don't care if it makes revenue, great warn them. So just have something in mind for how you feel at negative ads as opposed to your brand perception.

5.17 Dynamic Keyword Insertion

Another simple test to run, dynamic keyword insertion versus your static ad, or dynamic keyword insertion and the headline versus dynamic keyword insertion in description line two. Simple test to run. So a text ad testing, you're really starting first with just completely different types of ideas for your ads.

5.18 Rich Media Ad Testing

But it's also important to test rich media ads as well. Now, we talked a lot about image ad best practices in our Rich Media Module. So I'm not going to rehash the same ads over and over again. Just let's remember the most important component is to really look at, with image ads. First one's a call to action. When you see image ad without a call to action, it looks like a picture. What ad looks better, this one, the bookdesign something, if you can read what that particular word is saying, or Share your book with millions of readers, Learn How? Right, simple ad, we don't read text in ad copies. In fact, if you can read all the text in that ad copy, I'm impressed with your eyesight. It's hard to see all the text there. Both these ads are for trucks, just different dealerships. One, See More Photos, that's not even an ad that says go buy this truck. See More Photos, now, that's an interesting call to action, a test. Take a free test drive versus see more photos, could be the same image ad, just two different calls to actions. Commercial versus free, the first ad, Up to 70% OFF Top Brands. Of course there's again, not a call to action. Sierra Trading Post should add one. But it's showing you this is going to be something that is more based around getting your money. The second ad, Get Your Free Investor Kit. Now, this is useful for display is trying free, free trials, free downloads. Because often on display, someone is not in the buying cycle yet. Therefore, by including something that's free, you introduce them to your buying cycle. But if your goal is sales, feel free to be commercial. Just make sure your ad is easy to see and understand. The first ad from Chromebook, Nothing but the web. That's a Chromebook statement because there's no local programs, Buy Now. Or AMD Hex Core Configurator, $689 or up, so it does show a higher price point. No call to action, and it's an ad copy that has a huge contrast on the landing page. So, in this example, if the second ad, the AMD Hex Core, added a good call to action which they don't have now, then you make one ad that has a very clean layout and background. Another ad with a very busy background that would stand out on a lot of clean publisher pages because actually it's address shown, publisher pages. See what does better, the very clean ad copy, or does this blend in too much with clean publisher pages, or the one that has a big background contrast?

5.19 Everything Done Correctly

And when I show this ad previously, and I'll show it again because it's such a great ad. Number one, i've made this ad about four times larger than normal, you can still read the text. There's very little text but it scales well because it's already in big letters, so, even as you make it larger, you can still read it. The benefit message is a call to action, start living pain free, the, display your ail, the first letter of each word is uppercased, Laser Spine Institute. Good use of colors, you can tell it's being worked on. So, this is a good ad copy, but even in this If this were your ad changing the different benefit messages one starting living pain free. Another one could be learn about new surgery techniques, two different ad copies one's more learning, one's more pain free but very similar calls to action just test it out. The benefit message is a green button, what happens if you made it red and matched the color of the spine? That way it sort of flows together, so even if you have good ads that work, changing smaller things over time, are still good things to test to see which small elements still makes big differences.

5.20 Testing

And the easy way to get started with image ad testing is use Google's Display Ad Builder. Choose two or three templates, write your ad copies, do your initial tests with their templates. Then, see what benefit messages, overall templates, color schemes, are working best for you. Then, take that information and hand it to your designer or hire a designer to make your better image ads based upon the free templates you learned from. With video ads, you're often not going to create two, three, four videos and upload them. What you're going to do with video ads is start by making one video and use different opening images. So with a video ad, they don't auto play, someone has to click on the video to make it start playing. So just take two different images, just like you would an image ad and then, see what gets a better play rate initially.

5.21 Commercial vs. Non-Commercial

So again, we looked at this with video ads. Just remember, very noncommercial images. It's a simple story. Good opening image. Highly commercial ones, you might try them out. Usually has a lower play rate. Then, click to play the revealing 1 minute film. So usually video ads opening images that are stories do better than more commercial ones. Now what is useful is the way you name your rich media ads. If you name them, the size name and the theme of the ad copy, those two naming conventions. You can easily run pivot tables and look at the data across all the ads at once. So to see whether there's different sizes that are working. To see if it's different themes that are working. If you use Google's display ad builder you'll often end up with six different ad copies. Well really, you're probably not looking at the individual size so much as the theme or template of the ad itself. So if you make two different display ad builder ads, each with six ad copies, just name the themes independently, then you can see how the theme itself is doing regardless of the published size. There are so many things you can test. You can test shipping methods. This is often useful near holiday seasons. Christmas time we can get your product there by December 24th. Now be careful, it's common for them to say get your product there by December 24th. But what if you're flying out on the 22nd? So even telling someone we can get your product there by the 22nd means it's there before they fly. If they're flying on the 22nd instead of the 24th, then that wasn't useful. So it's another a good one to test even at shopping season. We didn't even talk about benefits, value propositions, products, and features. So if you are testing a lot ads in the learn phase. Do a lot of combo testing of features and benefits. If you're testing on the shopping phase, try different features. If you have guarantees. You have a 30 day guaranteed refund policy. Try that out in ad cop. You often see symbols in ads. If you have a registered symbol trademark copyright. Try them out in ads with ad copy testing. You get to walk away from the numbers. You get to use the other side of your brain for awhile. Be creative.

5.22 Testing Test Ads

Go to your ad group. Write two ads. Hit Save and wait to get enough data. If you're testing image or video ads, go to your Ad Group, upload your Image Ads. Wait. If you're using Display Ad Builder, what you'll see is, the actual size, and in fact, there are six sizes here. So Display Ad Builder looks a little different. You know how it looks, but you can always click and preview the ad copy. So collect enough information. You run an Ad Copy report. Click on the Customize Columns button. Choose your major metrics conversion, conversion rate, cost per conversion, and so forth. And then you'll get the Ad Copy report. So then you can look to see how is everything doing. What is the best ad for you?

5.23 Split Testing Ad Copy - Measuring Results

Now here's some actual data from a test. And what you see when you look at this data is there's no automatic best. There's best based upon what you're trying to accomplish. Now the reason Ad Copy 1 has so much more data is it's the baseline test. So it has a lot more base data. But if you want the highest click-through rate, your Ad Copy 6 is your best ad. If you want the highest conversion rate, that's a different ad copy. If you want the lowest cost per conversion, that's a different ad copy. There is no best. There's best for what you want to accomplish. That's what's really important to look at. If you look at these ads, this is an actual test, the cost per conversions range from $9.19 to $20.38. That's a really big difference. The conversion rates range from 5.69 to 13.22. So often you first start testing, you'll see some dramatic differences. Now in this particular example, let's say conversion rate was our number one item. We have a good lifetime visitor value program. We want the most new customers possible. So we keep Ad Copy 7 which is our best conversion rate. Now we don't throw out the rest of this data. We make a new test based upon this information. So if we have a good lifetime visitor value program a nice retention method, we want the most conversions possible. But then we want to increase our CTR, that's quite another important item to us. So Ad Copy 7 has a 4.96% CTR. The next best is 3.1. So it has much, much better click through rate. So let's take an element of Ad Copy 6. Let's take an element of Ad Copy 7. Let's combine them. Let's see if we can keep Ad Copy 7 conversion rate while increasing the CTR from whatever we use in Ad Copy 6. Make a new ad. Delete the losers now. And now run another test. So you're not going to just throw away data. Look at the outliers of information when you learn stuff. In this particular instance Ad Copy 6 has a good click to rate but terrible conversion rate. So that's where we might also look at why was it such a bad conversion rate, and make sure if that's some statement that we were using, it's not on our landing page. Because if it lowered conversion that much and it's on our landing page, we might want to remove it from our landing page too. So again, don't just look at your ad copy information from an ad copy standpoint. These are messages for consumers. That's what's important to look at. Now in this case was a fact that the ad copies promised something free. The landing page, step one was free, step two is paid. So it was an improper expectation from the consumer. In this example, Ad Copy 8 told consumers it was both a free and a paid option. Hence what consumers realized, they were going to pay money to click on it, lower CTR, better conversion rates. So ads do set serious expectations for landing pages. So keep your best. Learn from outliers. Make a new task. Hit save, wait again, and then you're continuously doing ad count testing.

5.24 Use Analytics for Engagement Metrics

Now let's say though that our goal wasn't just clean metrics, it wasn't conversion rate and click through rate. We want people to be on our site for a long period of time, our goal is we're launching new product. And we really want is engagement with the site, in this case, this is Google Analytics, so the ads are in there. But, analytics programs you can read your ad copies in your analytics when you combine them properly and then look at your ad content and then tie them on site. You're going to pick the ads with the higher time on site, then delete your losers. So even if your goals are engagement, pages per visit, time on site, you can get that information from analytics to determine your winners. If your goal is people watching your video and you really want to see who's watching most of our video because again we're launching product. We want people to see this video, they're not going to buy necessarily in our store in our site. If you're a cell phone provider, your Sprint or T-Mobile or Verizon, you want to see the ads. They're going to walk into your store to finally convert in many cases, so even then we want to know is, who's playing the video for at least 50% or 75%? What's a better opening image for that play length in the dimensions tab in AdWords? Under the free clicks report, you can see your video play rates by ad copies as well. So this is what I want to stress, this is the most important thing to take away, ad copy testing's easy. You write a couple of ads in your account, that's not too hard, five, ten minutes per ad group. You hit save, you wait, a wait is pretty simple, then you go and run a report. That takes about one minute, you look at the data, that takes a couple more minutes. You save your winners, you look at the outlier, you make a new ad, you delete the rest, hit save, repeat, it's simple. When you start seeing dramatic results, you're going to be very happy with that copy testing. Now if you're not the creative person though, you're you think to yourself, well, I can run ANOVA and I can do this crazy Excel math, but I don't really, I'm not a good writer. In a case like that, get someone else involved in your company, do a company contest. Whoever writes the best ad copy for us, best, of course, being based upon your goals, whoever writes the best ad copy gets next Friday afternoon off. As we all know, no one works Friday afternoons anyway, so all of a sudden, you get more people thinking about your company without really losing time with your workers. I've seen examples where companies win contests and they found some new hire who wasn't even in the marketing department was a great writer overall and did fantastic at copywriting. They moved them to the marketing department, so if you're not the creative one, get someone else to write ads for you. But you need to do ad copy testing, it, it is essential.

5.25 Recap

So just to recap, start first by thinking about the messages you want to test. Start by first, this is an informational ad, this is a question ad, this is DKI, this is not DK. Start with completely different messages. Then choose a couple of those ideas. Now, the number of ideas depends on your traffic. If you're doing 10,000 clicks a month, you can test five, six, seven ideas. If you're doing 1,000 clicks a month, only do two, maybe three ideas. Write some ads in your ad group. When you first start, choose completely different messages. Once you do a few rounds, you might want to do some that are more incremental changes, where you're only changing one aspect of ad copy. But please go ahead and continue doing completely different types of ads overall, as you're doing testing, as things change. Wait until you have enough data, just make sure you are confident in the information. Run an ad copy report. Chose the winner, but chose the winner based upon your goals. Delete your users. Write new ad. Write a new ad or ads based upon what you've learned from the outliers. Then collect the data, run the report, choose the winner, look at the outliers, repeat. Simple to do ad copy testing. When you're constantly running ads and doing testing, your advertising just gets better and better over time. As you look at your goals, see what message helps you meet those goals and then put those offers in front of consumers. So they see your ads, they can tell you can help them out, you will answer their questions. They click on them, your ads have set that proper expectation for the landing page. Consumers know what to expect when they get to your landing page and that just makes your conversions get better and better over time.

  • Disclaimer
  • PMP, PMI, PMBOK, CAPM, PgMP, PfMP, ACP, PBA, RMP, SP, and OPM3 are registered marks of the Project Management Institute, Inc.

Request more information

For individuals
For business
Name*
Email*
Phone Number*
Your Message (Optional)
We are looking into your query.
Our consultants will get in touch with you soon.

A Simplilearn representative will get back to you in one business day.

First Name*
Last Name*
Email*
Phone Number*
Company*
Job Title*