The Definitive Guide To Conversion Optimization

The Definitive Guide To Conversion Optimization

Written by Neil Patel & Joseph Putman

Chapter Two

Getting Started

At this point you’re probably excited about conversion rate optimization. You’ve learned that doubling conversion rates will cut your cost per acquisition in half, and you’ve learned how the 2008 Barack Obama presidential campaign used CRO to activate 288,000 more volunteers and generate an additional $60 million in campaign donations.

With numbers like that, you’re likely ready to dive right in and start testing. That’s great, but you have to know where to begin. A lot of people get excited about CRO, and then they run the first test that comes to mind. That’s fine, but it’s not the best way to increase your chances of getting a win.

Why You Should Begin by Gathering Data

One of the most important lessons I learned after spending $252,000 on conversion rate optimization is the importance of gathering both qualitative and quantitative data before beginning to test. If you don’t gather data, then you’re left guessing what to change based on your gut. This isn’t a good idea and can lead to wasted time and money. It’s better to spend a month or so gathering data so you’re testing based on insights you’ve learned, not on your gut feeling. Without taking this step, you’ll end up running a lot of tests that fail.

It’s also important to gather both quantitative and qualitative data and not just to look at numbers. Quantitative data provides a good starting point, but numbers don’t paint the whole picture. For that reason you also need qualitative data which means you need to ask your current and potential customers questions about your product or service. Qualitative answers provide you with more insights about why customers are or are not buying your product and service and most often lead to the biggest breakthroughs.

Use tools like SurveyMonkey to gather insightful qualitative data.

Last but not least, it’s better to have too much data than not enough. It’s easier to identify patterns and find useful information with 100 survey responses than it is with 10. Yes, you can learn something from both sample sizes, but you’ll learn even more with additional information. And if it ends up being too much, you can always ignore superfluous information, but it’s difficult to make a decision based on inadequate data. This is why you should spend at least a month gathering and analyzing data before you even begin testing because you’ll end up saving time and money as a result.

Option One Google Analytics

One of the best places to start is with Google Analytics (GA). GA provides numerical data about how people are using your site, which steps in the funnel they’re dropping off from, and much more.

Google Analytics provides a great starting point for analytical data that will help you learn more about your website and how customers are using it.

To begin, click on “Conversions” and then “Funnel Visualization” to see what the funnel conversion rates are for your site.

If you don’t have this set up, you’ll first need to set up conversion goals, and then you’ll need to set up a conversion funnel. Conversion goals show you what percentage of your site visitors convert, and the conversion funnel shows you how many visitors move on from one step of the process to the next.

If you’re site visitors follow a standard conversion funnel, you can benefit a lot from evaluating this information in Google Analytics. For example, if you run a SaaS website that directs visitors from the homepage to a pricing page to a free trial sign-up page and finally to a confirmation page, then the conversion funnel in Google Analytics will show you what percentage of visitors move from the homepage to the pricing page, what percentage moves from the pricing page to the sign-up page, etc.

Knowing this allows you to measure each step of the funnel to find out where there might be a bottleneck for sign ups. If 50% of visitors move from the homepage to the pricing page, but only 5% move on to the sign-up page, you need to figure out how you can get more people to move on to your sign-up page in order to increase overall conversions. Conversion goals and funnels in Google Analytics helps you to keep track of these valuable stats.

To learn more about conversion goals and funnels and how to set them up, read this post from KISSmetrics titled The Google Analytics Conversion Funnel Survival Guide. You can also watch this Youtube video titled How to Set Up Conversion Funnels in Google Analytics.

Once you have conversion funnels set up and give it some time to gather data, you need to come back and evaluate the results for conversion rate optimization purposes. The key thing to look for is where in your funnel there might be a drop off. Are more people dropping off from step one to step two, or are more people dropping off at the final step? Take some notes about where there might be a bottleneck, and consider what changes you possibly can make to improve conversions.

Now that you’ve identified the drop offs, you can use the information to come up with hypotheses to test each step along the way and also to prioritize the tests you’ll eventually run. It’s still a bit early to start running a test, but go ahead and make a mental note of what you’ve found and begin to consider what changes you might be able to make to improve your conversion funnel. You also want to monitor your overall funnel conversion rate to know how well your site is converting. (And don’t worry, we’ll talk more about how to come up with ideas and how to prioritize in later chapters.)

Another key piece of information we can learn from Google Analytics is load times for your site. To find out what the load times are for your website, click on “Behavior,” “Site Speed,” and then “Overview.” The chart shows the average load times for your site, and you can dig in to find the average and daily load times for individual pages.

Knowing load times for your site is crucial because every second counts. A one second delay in page response can result in a 7% reduction in conversions. If an eCommerce site makes $100,000 per day, a one second page delay could cost $2.5 million in lost sales every year. In addition, 40% of visitors abandon a site that takes more than three seconds to load. (You can learn more about loading times and conversions by reading this nifty infographic: How Loading Time Affects Your Bottom Line.)

If your load times are slow, you can take steps to decrease page load speeds in order to improve conversions. These steps include decompressing images, switching to a better hosting company, optimizing your CSS, and more.

There’s additional data you can use from GA that provides insights about your conversion rates and how you can improve your website, but we’ll use conversion funnels and page load speeds as a starting point and move on to cover how to gather qualitative data.

Option Two Customer Surveys

The next step in gathering data is to survey your customers. By surveying customers who successfully made a purchase or signed up for your service, you’ll learn what motivated them to buy and you’ll gain a wealth of data about how to convince more customers to sign up or make a purchase.

But before we get into the details, here are a few tips for conducting successful surveys:

  • Don’t ask too many questions at once.
    Customers tend to get annoyed by longer surveys, and everyone’s time is valuable. If you present customers with a shorter survey, there’s a greater likelihood that people will fill it out, which means you’ll receive more responses to evaluate. Our recommendation is to limit your survey to 5 or 10 questions and to use multiple surveys if you need to ask more questions.

  • Offer a prize as an incentive.
    As we’ve already mentioned, everyone’s time is valuable, including your customers. By offering a prize as an incentive, you’ll increase the number of people who take the time to fill out your survey. This can be a $50 gift card to Starbucks, or a free iPad. Either one of those options is sure to get more people to fill out your survey.

  • Ask open ended questions.
    Multiple choice questions are great for some surveys, but when it comes to CRO, you want to hear subjective responses written in the customers’ own words. If you ask too many multiple choice type of questions, you’ll miss out on insightful customer answers that don’t match up with the multiple choice options you supply.

As far as a surveying service, you can use whichever one you prefer. Lots of people like SurveyMonkey so you can’t go wrong using them. Other options include Wufoo and Google Forms.

Once you’ve picked a survey tool, you need to decide which questions to ask. The goal is to ask questions that will provide insights about what convinced customers to make a purchase and what hurdles stood in their way and almost deterred them from making a purchase.

Good questions include (these questions are variations that borrow heavily from questions recommended by CRO experts Peep Laja and Joanna Wiebe):

  1. How would you describe [product/service name] to a colleague or friend?
  2. What other options did you consider before choosing [product/service name]?
  3. Why did you decide to go with [product/service name]?
  4. What almost prevented you from signing up?
  5. What questions did you have about [product/service]?
  6. What ultimately convinced you to sign up?
  7. How could we do a better job of persuading your friends or colleagues to choose [product/service name]?
  8. How would you persuade more people to choose [product/service name]?
  9. What are you hoping to accomplish with [product/service name]?
  10. When did you realize you needed a product like ours? What was going on in your world that caused you to come looking for [product/service name]?
  11. What problem would you say [product/service name] lessens for you?
  12. What two adjectives/words would you use to describe our product/service?

Each of these questions needs to be tweaked based on your product, service, or niche, but each of them provide open-ended responses that will teach you more about your customers and help you to get a better idea of what they value about your service, what features matter to them about your product, the words they use to describe what you do, etc. It’s ok to ask some multiple choice types of questions that make sense for your business, but you’ll learn the most from open-ended responses. (We’ll talk more about how to evaluate the responses in the next chapter.)

This is an example of a CRO survey created as a Google Form with open-ended questions.

Option Three On-Site Surveys

The next type of survey you can consider are on-site surveys. Options for this include Qualaroo and Google Feedback Surveys for Website Owners.

On-site surveys allow you to ask questions to visitors who are currently on your site which provides you the opportunity to survey visitors who eventually become customers and ones who visit your site but may not end up making a purchase. This second type of visitor is exactly the one you want to target with on-site surveys. Why? Because the goal is to find out what’s preventing visitors from becoming customers. You want to figure out what it is that’s preventing them from taking whatever step you want them to take on your site. Is it the price? Are they confused? Is the website not working? These are all questions you want to get answers for with on-site surveys.

Here are some great questions you can ask to learn more about your site visitors:

  1. Is there anything you can’t find on this page?
  2. Is there anything confusing about this page?
  3. Do you have any questions at this point?
  4. What’s your biggest concern about purchasing [insert product name here]?
  5. What’s the number one reason that’s stopping you from making a purchase?
  6. What else would you like to see on this page?
  7. What can we help you find?
  8. Why didn’t you complete your purchase today?
  9. What could we have done to convince you to complete the purchase?
  10. What’s the biggest problem we can help you solve?
  11. What are you looking for in your ideal solution?
  12. What else can we place on this page to convince you to buy?

Again, the purpose of on-site surveys is to find out what’s preventing visitors from becoming customers. The answers will let you know if they’d like to try a free trial first, if your prices are too high, or if something else is the problem. We’ll talk more in the next chapter about how to evaluate the results.

Option Four Usability Tests

Usability tests are another great way to learn more about your visitors and to find out what’s preventing them from placing an order on your site.

With a usability test, you provide a series of steps for a user to complete. Then, you find someone to complete the steps, ask them to talk about what they’re experiencing as they use your site, and record them while they’re talking and completing the steps you’ve provided. The video of the session then provides you with information about what was confusing on your site, what the user liked, and what they had questions about.

If you don’t conduct user tests, you can miss really obvious ways to optimize your site. For example, visitors may have trouble with the check-out process because they can’t figure out how to enter their phone number properly. Your IT team may not have given the phone number entry form much thought, but it’s become one of the biggest frustrations for users. Without conducting a user test, you’ll have difficulty discovering these kinds of easy to solve problems.

So how do you carry one out? There are two options. The first is to conduct the user test and to recruit participants yourself. You can set up your computer with a microphone and a screen recording software, visit a local coffee shop, and recruit participants for your test. That’s one way.

Another way is to use a site like makes it easy to conduct user tests by automating the recording process and providing scenario templates you can use for your type of business. The tests cost $49 each, but since you only need around 5 tests at a time to get good results, the cost isn’t exorbitant for the level of service provided.

Regardless of which option you choose, the purpose of user testing is to find out how visitors are using your site, what they think about it, which questions they have, and what they have a problem with. It’s the best way to get into their to heads and really understand what’s going on when they use your site.

So who can you test your site on? Here are some options:

  • You are a prime candidate to use the site and to carry out a user test.
    Often marketers and owners create something and then leave it for other people to use, but to put it in industry terms, you need to “eat your own dog food.” In other words, you need to use what you’ve created so you can understand your site in a more intimate way and so you become familiar with each step of the process.

  • Your customers are another testing option
    since they’re obviously great candidates to use your site and to let you know what they think about it. When possible, it’s best to find people within your customer group to test your site.

  • Anyone else can be a good candidate for testing.
    Even though they may not be within your target customer group, there’s a good chance they can point out glitches or confusing parts of your site since most people are familiar with websites and know what’s confusing and what’s not.

Depending on who’s available at the time you conduct your test can determine who you test. You can begin by asking employees in your office to test the site, move on to ask patrons at a coffee shop to give it a go, and finish by using a site like to test people that are within your customer demographic.

To carry out a test, you first want to come up with a series of tasks for your test participants to carry out. You can start by asking users to follow a path that your customers normally take to get to your website. For example, you could ask them to visit Google to search for your product and to talk about what they see and what impression they have about the search results. You could next ask them to click on your result or visit your website, look at it for five seconds, and then tell you what their first impression is of the site. After that, you could ask them to attempt to order a certain product and then watch them carry out the test to find out if they’re are any hurdles along the way.

This is an example of a user testing template provided as a starting point from

What Does This Look Like In Practice?

Recently, Switch Video was looking to improve conversion rates on their homepage. They weren’t sure what to do, but they wanted to generate more leads through the form located on the right side of their homepage. The form is used to generate leads from companies that are interested in creating an explainer video.

To begin the process, Switch decided to survey their customers to learn more about them.

They asked questions like:

  1. On a scale of 1 to 10, how likely are you to recommend us?
  2. If you were to recommend us, how would you describe us to a colleague or friend?
  3. Were there any questions you needed answers to but couldn’t find on our website?

Not surprisingly, the third question provided the most helpful results. A large percentage of respondents mentioned that the biggest question they had was about finding out how much a video would cost. Switch didn’t include pricing information on their site, so an overwhelming number of responses mentioned that the biggest question they had and couldn’t find an answer for was related to how much a video would cost.

Based on these results, Switch decided to run a test. Their original CTA button copy said, “Get a Free Consultation.” They decided to test new copy that said, “Get a Quote.” So they set up a test, ran it, and evaluated the results.

After testing, they found that the new version increased full form completions by 221%. That’s right. Changing the copy on a CTA button increased lead-generation form completion by 221%.

But before you go out and make this same change to your site, you have to remember that this result was specific to Switch’s business and was based on information that came in through insights from customer surveys. Switch Video didn’t just guess what to test and then get amazing results. They followed a proven conversion rate optimization process in order to generate measurable and lasting results.

In order to replicate this kind of improvement, be sure to continue reading this guide. You’ll learn how to evaluate survey results to generate insights that lead to winning tests, and you’ll learn how to carry out statistically relevant tests that produce accurate results.

Chapter Two Summary

  • Before you begin testing, you should start by gathering data through Google Analytics, customer surveys, on-site surveys, and user tests to learn more about your visitors and gain insights that will indicate what should be tested based on your customers’ experiences. This will save both time and money even though it requires a little more effort on the front end.
  • Google Analytics is a great place to start in order to get qualitative data that will help you find out what bottlenecks you have on your site as a way to increase conversion rates.
  • Surveying customers is the next step you should consider in order to find out what convinced them to make a purchase, what nearly prevented them, and what words they use to describe your product or service.
  • On-site surveys are another great way to learn more about your customers, identify hurdles, and learn the language they use to talk about what you’re selling.
  • User tests are a necessary step to learn more about your customers and find glaring mistakes on your site that may be significantly affecting your conversion rates.

Four ways to share