Advanced Pay Per Click (PPC) Certification Program

Certification Training
2382 Learners
View Course Now!
40 Chapters +

Landing Page Testing Tutorial

1.2 Introduction

Hi, I'm Brad Geddes, the author of Advanced Google AdWords, the founder of Certified Knowledge and the PPC Faculty Chair from Market Motive. In this lesson, we're going to look at landing page testing and how to test two different pages.

1.3 Where to Send Traffic

So first off you just have to remember what people think when they get to your website. Number one, are they in the correct place, and is this what they expected to see? Now these two items are really set based upon your ad copy and the expectations your ad copy laid out. Next is do they trust your site? Often adding credibility elements can be useful. How long will this take? People value their time tremendously so whenever you affect the big three, time, money or family, you need to be careful what the message is on your landing page. Where should they go next? And then finally what you don't want them thinking, should they use the most used navigational element on the entire web? Click the back button. So we'll look at conventional wisdom. You know, where do you send traffic? Conventional wisdom says the furthest logical point in the buying cycle or a page dedicated to the search query. Now conventional wisdom is right most of the time, hence why it's called conventional wisdom, however, it's not always correct. There are always exceptions. So it can be useful just to do some basic testing for yourself to see if you have other pages that will do better from a conversion standpoint. So we'll look at some search queries and search query intents and look at how to devise just a simple test based upon what someone is looking for.

1.4 Informational Queries

Now informational queries make up by far the majority of queries on the web. If you do a search for candle burning time you often end up in product pages. Well product page or category page can be useful if you really want to monetize something, however that wasn't really the question the searcher asked. The search ask for how long do candles burn for, often, in a case like that, maybe they're hosting a dinner party. They want to know, do they need tea lights or tapers or pillars to last throughout their entire dinner party. So what you can do it take another page, make a simple chart, here is our candles, here's how long they burn for, then click to get to the product information for that particular page. That way you answer the query and give them a chance for next conversion step. So a simple test, take a product page and a comparison page, send traffic to both, measure conversion rates, it might be that your comparison page adds another step to the process, lowers conversion rates. Where product page in some cases might be better because you can monetize that click other times you're going to lose users because you did not respect the buying cycle they need to learn information and enough about something before they can start shopping. So a simple task then, especially when you have things further up in the buying cycles, here's learn-based page, here's a shop-based page, which one brings in more profits?

1.5 Local Business Queries

Use it when you do local business queries, for instance kitchen remodeling. Conventional wisdom says, let's put a picture of one of our kitchens, talk about how good we are, maybe even before and after pictures of kitchens, so forth. The problem with local contractors, and this is especially true of contractors, is that everyone has a horror story they've heard, about how they hired this guy to redo their kitchen. And it's three months later and they still don't have running water. There's a lot of trust elements with contractors. So another page that I've seen work well many times, but by no means always, are about us pages. Here's who we are. We're a member of the Chamber of Commerce, the Better Business Bureau. Here's our customer testimonials. Integrity, innovation, experience, right? Highlight why you're so good. And then from a simple testing standpoint, here's about us page. Here's our remodeling page. What does better over time? Simple test to run.

1.6 Narrow Theme Sites

Sometimes with narrow theme sites, you find odd things happen, where dedicated pages don't always perform best. In fact, we'll look at some stats on this one a little bit later in this module. So if you do a search for Chicago nanny services, conventional wisdom says let's put up a picture of the Chicago skyline, some sample nanny resumes. Talk about how We can get a nanny for you in Chicago. But sometimes, home pages do work well. For instance, you get to a page, then it says, families go here. Choose if you're looking for a nanny, and choose your zip code, or maybe it links to a zip code already. And if you're a nanny, you can sign up, but by the way, we're going to make sure you're tested by the nanny Association, we're going to do a background check, we're going to make sure you're certified. While this page, monetarily is for the families, there is a trust element saying, here's how we get our nannies, we do all this background check stuff first, so it's just another trust element. So in a case like that simple task, there's a home page or even dedicated page to already a zip code or a city name. It says families go here. By the way, nannies, we're going to do all these various items before we put you in our system. Simple test to run, what does better? One that's focused on just the query or one that's focused mostly on the families, however It has trust elements of how these people are hired who take care of your kids. Who are these people coming to your house?

1.7 Ambiguous Queries

Ambiguous queries such as Chicago Mortgage or merchant accounts often these are pages that have some information and some form on them. Now this is where forms, you need to be careful, often with forms questions are asked which a consumer can't answer. So if you get to a page like this, and all of a sudden there's this required question that says, do you want a debt consolidation loan? A home equity loan? A purchase, or refinance, or reverse mortgage? A loan modification? A lot of times, consumers might not know what all these things mean. So all you did was drive them back to a search engine, and they're going to search for the difference between a reverse mortgage and a loan modification loan. You lost the consumer. Where, instead, make a page that's, here's our four types of loans. Or here's the five different options you have. Here's some bullet points about each one. And here's how either you apply for it or contact us to learn more about it. So simple tasks to run. A comparison option page which again puts one more click though in front of someone getting to conversion page and the actual application form. Try two pages, see what does work better, from a profit standpoint.

1.8 Product Queries

And often you do product queries, you end up on pages with thousands of options. It's overwhelming to users. Now the problem with product pages is that they have more options than humans can process in short term memory. People's short term memory has five to seven abilities for processing. That's it. One is wided a search and what things you're keeping in mind. Another one's you've got a meeting in 15 minutes. If you overwhelm someone, all of a sudden, a decision feels bigger than it might really be. And forces them to bookmark a site or put it off until they have more time. Where instead if you put them to a page, here is our three highest rated products. Or here is our highest rated, our current sales product and our best featured product. Narrowed down the options for the user. And at the bottom of the page you can put a link that says, see all of our products. You'll have some people who want to read specs of 57 television sets. Fun, give them a pathway to conversion, but a simple test you can run. Send some traffic to your category page, shows all your TV sets. Try either a sales page or a limited product page for your other ones. Now if you're e-commerce, you're not going to do this for every single possible product query. That's way too much information. You'll do this in one of two ways. One, you'll choose some of your best-selling products overall, the products that you really want to sell and feature. Make some dedicated pages for those. See how it works. And if it does work well, then you can take some of your higher overall products that you want to increase conversion rates for, and make static pages. They might even live outside of your content management system. Another way of doing it, is take potential templates for your site. Choose a half dozen or a dozen different types of products and then do testing at the category template level. And if you find templates that do better, apply that template to the rest of the site. Often with e-commerce sites, it's hard to make dedicated landing pages because you're running off template driven information. So instead of changing your templates console, which sometimes is not possible, make pages outside of the template that could become templates, test out overall layouts. And if it does work, then apply the information back to the template.

1.9 Brand Searches

Even branded search queries. So you do a search for ThinkGeek, often you end up at a home page. And this is generally a waste. Not because you shouldn't buy branded queries, there's a lot of good reasons to buy branded queries. Because if you look at your site, you generally have pages you wish more people saw. So choosing a what's new about us page. So here's our new offerings, here's our new services. Try a page that you wish more people saw on your site. It might even be, here's our newsletter subscription. Take those branded queries, put them in your newsletter page. Now someone just wants your homepage. They can click on your logo. They'll go back to your homepage. But just simple test to run. Here's our branded query, here's our homepage. Here's a, What's New page. Which one does better overall?

1.10 Thank You Pages

Even thank you pages are useful to test post transaction. One, create an account. So, you don't want to force account creation before someone can actually check out on a site. But here's one option, thank you for placing an order, would you like to join our site, create an account, and so forth. Or thank you for contacting us, we will get back to you as soon as possible. While you're waiting, would you like to download a whitepaper, follow us on social media, learn more about our company? Some simple tests to run are, take your thank you pages, separate them in other places to see what gets more long term revenue and engagement from people who did convert. The most expensive thing to do in marketing is get the first consumer. It's cheaper to keep a conversion than find a new one. So make sure your thank you pages are helping to engage people who did convert.

1.11 Test Homepages

Now your homepages get thee most traffic, test out your homepages for layout purposes. See how you can increase conversions, we've had many sites. We've tested, four ,five, six different homepages. Right at the beginning. because your homepage is essentially a big segmentation page. That is so, here's our stuff now get further on our site as quickly as possible. So test those homepages. Now often, you can't just duplicate your homepage four different times. And have four different homepages as a whole. It's only not possible you'll mess up your, your organic rankings, so forth. So a good way to test homepages, take just two or three homepage variation and take your branded traffic and send it to your home pages. Then when you find ones that do better overall, then make them your actual home pages.

1.12 Testing Category & Conversion Pages

And you can't make assumptions on things as well. That's why testing is so important. We have a membership site and we had a page where it was just a simple, the default form, make a membership. And we made just a slightly nicer variation of this particular pitch. Now, what was surprising is that when we were dealing with developers, they liked the first version more. They had a higher conversion rate. We're dealing with product individuals or small businesses. They like the second version more. Hence why you have to leave your assumptions outside of your testing. Assumptions will give you hypothesis for testing but then you need to look at the data to determine what really does best.

1.13 Segmentation Pages

Now, big mistake a lot of advertisers make is they'll have two different intents of users. There's the same word, they have two products, or they have two landing pages for them. And so, what they do is they make two ad copies, one with one intent, one with the other intent, and just hope the right person sees the correct ad. That's not usually a good way of doing it. Instead make a segmentation page. Are you a business traveler go here? Leisure traveler go here. Business travelers want to know such things as do you have free wifi in the hotel?, do you have an executive lounge? Do airport shuttles. Families want to know that you have cribs to rent, you have a swimming pool, are you close to the amusement park? Now, these two users just search for Chicago hotel, Los Angeles hotel. They don't search for Los Angeles hotel for business travelers. So, when you have either ambiguous intents in the query itself, your product has two or more uses for it, don't just make two ads with different landing pages for the same keyword. Make a segmentation page. Now, be careful with segmentation pages. Because you do add more clicks before conversions happen. So, this is one at the New England Journal of Medicine, you got. So, their page you saw physician resident students. You clicked on it you then saw someone with the same demographic as you were. Fantastic. That's a nice segmentation. Then, you clicked again to get to a free sign-up. Free is often considered of low value. And many times, a test like this might actually not do well. Because free should be simple in many cases. So, then you see a nice pathway of pages. Every one does increase the number of clicks before conversion. So, in that case, what I'm going to test here's just a landing page before you sign up right here on the page. Here is another one that uses a pathway. What does better? The question is how much navigation should you have. So, this is a test eHarmony ran, and there's very little difference between these two pages. The biggest difference is the Why eHarmony, What to Expect, Real Couples, this could be, kind of like, about us. And what you'll often find is that based upon the amount of information somebody already knows about you? The less additional About Us stuff you need to add, the less they know the more About Us you need to add. So, if you're well known in this city right now, you might choose a version without all additional links because it gives people just more options to get lost in your content. Where someone doesn't know that much about you, then the question is do we need to add more about us information so they can determine if we're the best company for them? So, with testing, you might even look and say here's our test for cities where we have good brand recognition. Now, let's maybe even run the same test. But let's look at the data from cities where we don't have good brand recognition, and segment the data based upon how much users know, then measure our test results. Unfortunately, a lot of ad copies say, use XYZ for 10% off. And then you get to the landing page, you don't see it, maybe you walk around the site a little bit, you find that it convert, you don't remember it and what happens is consumers just don't convert without the code. They go back and look again for it. If they can't find it, you often annoy the customer, the searcher, and they don't convert at all. Now discount, as we talked about with ad copy testing, often increase conversions to always increase revenue. So, a simple test, one landing page shows discount codes on it, one landing page does not. One landing page is just full price. Now, measure conversion rate changes in your profit over time based upon adding a discount. Does it increase conversion rates enough to offset the fact you don't make as much money each conversion?, or. Does it increase conversion rate so much it's worthwhile selling at a discount? So, does it when you add discounts, do it extra level of math to see the difference and the simple way is two landing pages. One focused on discounts, one focused on just convincing someone to pay full price.

1.14 Video Testing

Now video is often really good to test as well. So we put on a lot of seminars, and we just had a simple page that says here's our ticket information, get grab your tickets, here's where the place is located, so forth. And this page had a 6.7% conversion rate, so pretty good for a high priced product. Then we did some video testing. Now one of the biggest things you need to test with video, is do you autoplay it? Do you auto play it muted? Do you not autoplay it at all? So our control 6.7% conversion rate, when we autoplayed the video. And it, people's computers started speaking our conversion rates dropped quite a bit, because this was a B2B offering. If someone's sitting in a cubicle in an office, and they get to a site, and their computer starts speaking, they click the back button, and look around to see who noticed. Now often in a B2C environment, it can be useful to autoplay. Another option's to autoplay it, but make it muted. That way it draws attention to itself, but someone has to unmute it before they can actually play it. So in this particular case, our autoplay conversion rates dropped. Autoplay muted, eh, a little better, not a lot. We didn't autoplay it at all, our conversions are much higher. Though, that's not always the case. So, a simple test is add video, and video testimonials are fantastic. Add some video, then do one no video. One page with a video that doesn't use autoplay. One that does autoplay, see what does better. Easy test to determine. Just make sure if you add video, that people play it. This is something Toyota ran, and they used their own video player. And it was really hard to tell it was actually a video you could play, and everyone skipped over it. So that's why, even if you're not technical. Use YouTube or Vimeo, one of the other places. Let's just embed the video. Everyone's use to those by now. If you want to truly control the landing page environment then go ahead and use your own video system if you want to. Now also make sure as you're testing pages and creating content, you don't let developers write it. And that makes sense, but the majority of errors on forms are not written by marketing. They're written by developers. Just go to your form. Hit the the submit button. Don't fill out anything, see what happens. Often if you get that pop-up box that says, oh everything with an asterisk's required. Then you have to hunt around the form to try to find all the asterisks. It becomes difficult to see that. Or do you just highlight the errors and say here's the differences? So making forms easy to navigate, very useful. But also a good test to run is, here's a form with very little information. Here's a form with a lot of information, what does better overall? Now what's going to happen is your form with little information will get a whole lot more conversions, but less qualified conversions. So it's not just conversion rate in some cases, you want to follow through all the way through your CRM system, to see what hatches the best sales. Now another option, is on page one, make it really simple form, name, email, maybe phone number. Page two, ask for all that additional information, your website, your industry, additional contact information, so forth. And then sales can order the queue based upon the amount of data they did receive. It's a slow day, they're going to take the information that is very little qualified leads. It was just name and email. On days when they have lots of good forms to look at, then they're going to take the more qualified ones, where someone's took the time to fill out everything. But sales order their cue based on the amount of data they do have. But simple test, do we get more leads or less leads, but more qualified leads? Now I want to stress probably the most, is that you don't have to have gorgeous designed pages. If you're a big company and you have the budget, fantastic. Make some really pretty pages. If you're a smaller company, you might not have the budget to really make really fantastic dedicated pages. That's okay. What's important with pages, is easy to understand, easy to comprehend information. These are some pages that are externally used to run, and there were tens of thousands of these. Now you look at this page, and it's not something that's necessarily a fantastic looking page. Right, it's easy to understand information. These pages have fantastic call through rates, because you can tell what you're looking for. So don't clutter your design, make it simple information. That's what's so much more important, is easy to comprehend information. So if you have a lot, don't make pages really busy. Make it easy to see what you actually do.

1.15 Two Ways to Test Pages

Someone testing multiple landing pages. There's two easy ways to do it. You can just duplicate your ads in an ad group. So the ad texts are the same, and then change the destination URL to each landing page that you want to test. And we're going to walk though how to do this. If you're using an ad server, then look at how your ad server sends traffic to various pages, you can use Analytics Experiments. There's many third party tools, or even just to rotate which visitors go to which landing pages. So all your landing pages receive traffic, and then you can see how each one does independently. There's a simple way to do this. Write an ad, send it to page one. Duplicate the ad copy, send it to page two. That's really simple with the ad group [INAUDIBLE]. Then wait, collect the data. Finally, you can run an ad copy or a destination URL report. Either one you can run. Now, when you do ad copy testing or at landing page testing, any testing at all, never combine search and display. You can segment out the data to make sure you're looking at only search or only display information, and in adware, this is called network. So don't combine them. Now, the data we're about to look at in a second, I can't show the actual advertiser for NEA reasons, but I can show similar pages. The query itself was Chicago Naming Services, and one page was the home page, said nannies and baby sitters, here's your family search. Put in your zip code, go. And of course has the background information about nannies apply here, but we're going to do all this background check. The other one was just a page dedicated to the query itself. Pictures of the Chicago skyline, some resumes then sign up. Now we look at all the data, search and content combined. ] What we see is that the cost per conversion and the sales rate or conversion rate, was better for the page dedicated to the query, that city landing page, it did better. But this is why you never combine the data together, segmentation is important. If you look at the search information, the home page is a $13 cost per conversion while the city landing page is $19.99. Cost per sale, or cost per conversion. The conversion rate, homepage 16%, the city landing page 9.4. Homepage is better, in fact this company has done a significant amount of tests. Their homepage is better for search. As they send tens of thousands of search queries to their homepage, defies conventional wisdom. However, display is different, for display their homepage is a 6.5% conversion rate. City landing page is 11%, in fact this is a company who can't get enough display conversions because their display conversion is nine dollars. They're search is 13, so some companies do better with display. And this happens to be for a revenue event, it's not for anything free, people are paying money on theses conversions. So make sure you segment that out. Now think back to how people go through this process of buying it. If someone searching, they've thought about who's this person coming into my house to take care of my kids, sometimes many family again. They've really though this more, who are these individuals, maybe compared some sites up for it. Now someone's on display. Often you're introducing them to your buying cycle. You may introduce them and get conversion that same moment. That happens a lot too. But they haven't critically thought nearly as much about, who are these people, how do they do these checks? How do I do these background checks? And so in that case that city page might be better than that home page where search, someone's thought more about it, they want more information beforehand. So, with search and display, test your pages independently.

1.16 Don't Design by Committee

This is something good to remember for pages, don't design by committee. When you sit down and start designing by committee, you end up with mash-up pages that have no clear direction, because you can't satisfy every VP and every CEO and every marketing manager in your company. The people who should be designing your pages, and telling you what's best, is your customer. They will tell you based upon their credit card usage, their phone calls, their form fills. What page they like best. So as long as you don't make all these assumptions, let the data tell you which one's best, that's when testing become much better. So testing different types of pages? Really simple. Find two different pages you want or make two different layouts. Send some traffic to one, some traffic to a second one, wait, that's pretty easy. And then look at the data, take your winner. You're obviously going to save it. Look at your loser information. See if you can learn about why someone did not convert. Was it a different message you tried? That message doesn't help conversions, good to know. Was it a different type of layout? Well, avoid that one in the future. So learn from failed data just as much as positive data.

1.16 Don't Design by Committee

This is something good to remember for pages, don't design by committee. When you sit down and start designing by committee, you end up with mash-up pages that have no clear direction, because you can't satisfy every VP and every CEO and every marketing manager in your company. The people who should be designing your pages, and telling you what's best, is your customer. They will tell you based upon their credit card usage, their phone calls, their form fills. What page they like best. So as long as you don't make all these assumptions, let the data tell you which one's best, that's when testing become much better. So testing different types of pages? Really simple. Find two different pages you want or make two different layouts. Send some traffic to one, some traffic to a second one, wait, that's pretty easy. And then look at the data, take your winner. You're obviously going to save it. Look at your loser information. See if you can learn about why someone did not convert. Was it a different message you tried? That message doesn't help conversions, good to know. Was it a different type of layout? Well, avoid that one in the future. So learn from failed data just as much as positive data.

1.17 Recap

Just to recap, first off, choose two different pages, A, B testing, a good way to start. You might do multivariate testing later. That's fine, once you've really gotten the high level stuff looked at. But choosing totally different pages, whether it's different messages, different types of forms a discount page versus a full sales page is a good place to start. And then just do that A/B test. So just create two identical ads in an ad group. Send one ad copy to page one, duplicate it, send the second one to page two. Very simple to do, then you want to ad copy report or destination report, either one gives you the same data. Then always segment content versus search, can't stress that enough. Always segment based upon traffic source. A conversion is made up of the traffic source, the offer and the landing page. If any one of those change you need to look at the data independently. Then choose your winner, examine your failed test, delete the loser, pick the same or another ad group to test again and repeat the process. So testing essential, if you ever look at a page search account, and it has several thousand clicks but no conversions, it's almost always the landing page. The worst keyword will occasionally convert, the worst landing page will not. So, page search gets you traffic, it gets you qualified traffic. It gets you qualified traffic that you can use the ads to set expectations for what they find on the landing page. What is the traffic source? The website itself then, has to finish converting the traffic's source in to a customer. And the best way to keep increasing those conversion rates over time is with landing page testing.

Find our Advanced Pay Per Click (PPC) Certification Program Online Classroom training classes in top cities:


Name Date Place
Advanced Pay Per Click (PPC) Certification Program 25 Aug -9 Sep 2018, Weekend batch Your City View Details
Advanced Pay Per Click (PPC) Certification Program 29 Sep -14 Oct 2018, Weekend batch Your City View Details
  • Disclaimer
  • PMP, PMI, PMBOK, CAPM, PgMP, PfMP, ACP, PBA, RMP, SP, and OPM3 are registered marks of the Project Management Institute, Inc.

We use cookies on this site for functional and analytical purposes. By using the site, you agree to be cookied and to our Terms of Use. Find out more

Request more information

For individuals
For business
Name*
Email*
Phone Number*
Your Message (Optional)

By proceeding, you agree to our Terms of Use and Privacy Policy

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*

By proceeding, you agree to our Terms of Use and Privacy Policy