Conversion Optimization FAQ

What is Conversion Rate Optimization (CRO)?

“Conversion optimization, or conversion rate optimization, is a process for increasing the percentage of visitors to a website that become customers, or more generally, that take a desired action on a website. This is commonly referred to as CRO.” — Wikipedia.

CRO mainly focuses on improving the performance of the website. Since CRO is a catch-all term, several myths have arisen over the years, such as that CRO is synonymous with A/B testing or that best practices always work... Perhaps the best way to describe CRO is to compare it to the scientific approach. By studying the behavior of a website visitor, CRO experts can identify weak spots in sales funnels and create strategies and hypotheses based on hard and soft data.
‍

How do you calculate conversion rates?

Conversion rates are calculated by simply dividing the number of successful conversions, which depend on the goal of the website, by the number of website visitors or potential conversions.

What is A/B testing?

A/B testing (sometimes called split testing) involves publishing both a control version of a website and a modified version at the same time. The experimental page usually has small (and a few) changes in design elements, such as call-to-action (CTA) buttons, headers, or colors. However, it can also involve testing multiple pages, focusing on a particular element or component, and it’s not always a small change; it can also be a completely new look and feel. A/B testing differs from multivariate testing because it focuses on a single element, while multivariate testing involves testing many different elements simultaneously.
‍

What is multivariate testing?

Multivariate testing "is a technique for testing a hypothesis in which multiple variables have been changed" - Optimizely. It is most effectively used when one wants to determine the best combination of variations by testing all possible variations against each other.

Multivariate testing is most useful when a website contains too many conflicting elements and/or when conversion rates are very low. Despite multivariate testing not being as popular in the CRO world as A/B testing, it can drastically improve conversion rates if applied correctly.

Moreover, multivariate tests require a lot of traffic due to the additional variations. If traffic is low, high conversion rates are needed. A general rule of thumb: if your traffic is less than 100,000 unique visitors/month, you might be better off conducting A/B tests instead of MVT. The only exception is if you have lead generation pages with a high conversion rate (10% to 30% conversion rate).

Is it better to perform A/B or multivariate testing?

If you have enough traffic, use both. They serve different but important purposes. However, A/B testing should generally be your standard.

With A/B testing, you can:

• Test more significant design changes;
• Execute tests in a much shorter time than with MVT;
• Implement and evaluate advanced analytics for each variation (e.g., mouse tracking information, call tracking, analytics integration, etc.);
• Isolate individual elements and interaction effects for learning and building customer theories;
• A/B testing generally yields larger gains (as you often test bigger changes).
• A/B testing tends to provide meaningful results faster. The changes between pages are more drastic, making it easier to see which page is more effective.

Thus, A/B testing harnesses the power of large changes, not just adjusting colors or headings, as is sometimes the case with MVT. Optimizers typically start with A/B testing, as there is potential for more significant gains.

Multivariate tests focus on measuring interaction effects between independent elements to see which combination works best. MVT is great for follow-up optimization of the winner from an A/B test once you have narrowed down the field.

While A/B testing doesn’t tell you anything about the interaction between variables on a single page, MVT does. This can help you in redesigning by showing you where different page elements have the most impact.

How much traffic does a website need for A/B testing?

It’s not necessarily about the amount of traffic that matters, but more about statistical significance and time. For example, you could run a test on a low-traffic website, but it would take months to derive reliable results unless the conversion rate is very high and/or the difference between A & B is substantial.

In general, we recommend getting the user acquisition engine running before you embark on a serious conversion optimization campaign. The reason for this is that websites with low traffic don’t always have enough data points to determine which ideal customer to optimize for.

However, there are plenty of ways to improve conversion on low-traffic websites that are not necessarily tied to analytical data.
‍

How do you know what to A/B test?

There are countless elements you can test on your website. Every element of your website can be tested, from CTA buttons and images to psychological principles. But the best way to find out what to test is to look at the data and ask yourself a few questions:

• Who are the lowest converting visitors?
• Who are the highest?
• Why is this the case?

Then conduct thorough research that includes the following:

• Analytics analysis
• Competitive intelligence
• Formulating a hypothesis
• Sample size calculation
• Iteration (testing)
• Analyzing test results
• Reporting results
• Repeat

From this research, you will compile a list of points and insights. Some points, such as bugs or issues, can be adjusted immediately, while for others, you need to come up with a solution. You can then rank these solutions based on PIE (Potential, Importance, Ease).

• P = What is the potential this change will have to improve your pages?
• I = How important is this change to your page? Will the change affect pages with the highest traffic, or pages that are seen by visitors who arrive on your site through costly pay-per-click ads?
• E = How difficult will it be to implement a test on this page or template?

Assign a score from 1 to 10 for PIE for each proposed solution and use the average of these three scores to prioritize your list of points. Start with the highest scores.

Ensure that you have a hypothesis for each solution. A good hypothesis is framed using the following structure: “Because we’ve seen ‘ABC’ in the data for ‘THIS’ group, we expect that ‘ADJUSTMENT’ will result in ‘XYZ’ improvement. We will measure this through ‘KPI’.”

How do you measure the actual conversion rate?

First of all, there are different types of conversions. There are hard and soft conversions, such as superficial conversions like shares and followers on social media, or hard conversions like transactions that lead to revenue. Tracking different types of conversions will help you understand where customers are in the sales funnel.

Additionally, there is a difference between your average conversion rate and your actual conversion. Simply dividing conversions by traffic is not truly indicative of the actual conversion. We are not saying that it’s not worth calculating, but it doesn’t paint the entire picture.

Some suggest calculating only traffic generated from places where your target audience is located, eliminating further sources. However, we offer a slightly more detailed recommendation. For example, we split conversion by the source of traffic, location, operating system, device, age, gender, new or returning visitors, etc. Niche markets require knowledge about niche conversion rates.

What is the ROI of Conversion Rate Optimization?

Investing in CRO delivers significant returns. If you currently drive enough traffic to your website but are not meeting your goals with traffic alone, it’s time to consider optimizing your website (your conversion engine).

As there is a clear trend in online advertising, thinking about these topics is becoming increasingly relevant. More money is being spent on ads, yet fewer people are seeing those ads. It’s like throwing money away. Total madness.

Perhaps one of the simplest and clearest visualizations of ROI comes from Unbounce:

While the graph demonstrates hypothetical numbers, it still illustrates similar trends that we also observe with our clients.

‍

How can heat/scroll maps contribute to conversion optimization?

Heat/scroll maps can show you, in a very simplified way, how users behave on your website. They are extremely useful for finding answers to questions such as:

• Which sections does the user ignore?
• Are distracting elements taking attention away from the main CTA?
• Where are users trying to click, but the site isn’t working?

In many ways, popular websites are built similarly, with certain features placed in comparable locations. When you use heat/scroll maps like Crazyegg or Hotjar, you can deduce where users are typically inclined to click first.

But beyond simplicity, heat/scroll maps are inexpensive. Both large and small companies have something to gain from using these tools.

How long should you run an A/B test?

The easy answer: Until you have a statistically significant result.
The more complex answer: Well, that’s a bit more complicated.

For a fantastic answer, we recommend reading Ronny Kohavi's response on Quora.com. The duration of an A/B test should be based on more values than just the p-value. Sample size can be important, but if it doesn’t reflect your total user group, your test results may be inaccurate and misleading.

What are the best available A/B and multivariate testing tools?

We are big fans of Google Optimize for A/B testing and personalization, but there are CRO tools available for every budget and need. Feel free to check out this short guide on CRO tools that can help you make a choice.

How can live chat help improve conversion?

Studies show that live chat is a great option to enhance customer trust, reduce impatience, and increase conversions.

Analysis shows that:

• The same analysis shows that these numbers are even higher for people who shop online at least once a week.
  
• 90% of customers find live chat helpful
• 63% are more likely to return to a website that offers live chat
• 62% indicated they were more likely to repurchase from that website
• 38% of all respondents stated that they made a purchase due to the chat session itself

What are the best case studies for conversion optimization?

The best available case studies are those you can use to draw real conclusions that are directly actionable for your business. A good blog post from DigitalMarketer.com outlines their process for determining the value/validity of a case study. As readers, you should ask yourself:

• Is the sample size available?
• What is the uplift percentage?
• Are the figures for the raw conversions published?
• Is the conversion metric specified?
• Is the confidence ratio published?
• Is there a testing procedure? What is it?
• Is the conclusion justified by data?
• What was the test timeline?

Without these very specific evaluation metrics, the conclusions that the case study seems to draw may be “too good to be true.” And often, that is the case...

In addition to the questions above, readers should also keep an eye out for studies that seem counterintuitive. Sometimes small text changes or removing certain elements can significantly improve conversions.

Is personalization a result of CRO?

Personalization is likely a consequence of the internet in general, and not necessarily tied to conversion optimization. Many people expect it. In fact,, 77% of digital natives expect a personalized experience.

This is evident from major companies like Netflix and Amazon, which consciously prioritize the implementation of personalization. “We know that if we don’t capture the user’s attention within 90 seconds, they are likely to lose interest and move on to another activity,” said Nick Nelson, global manager for creative services at Netflix, according to an article from Fastcompany.com.

Another trend in personalization is the rise of AI. While some may not be very impressed by things like Amazon's Echo, and others delve into the limits of complete personalization on the internet, it is a fact that AI and personalization are not going away anytime soon. It will be interesting to see how integrated our lives will become with personalization software and hardware.

What is the biggest challenge of implementing CRO?

The biggest challenge for successfully implementing CRO is the lack of a solid strategy for the subject. Without a well-thought-out process, companies essentially halve the effectiveness of their CRO efforts. he fact that CRO does not directly relate to customer acquisition often causes it to be sidelined. And most companies don’t even have a budget allocated for CRO! This means that CRO is usually a secondary concern, while it should be in the spotlight for any serious marketing team.

Other issues include:

• Budget constraints: It is quite expensive to set up properly and to hire a full internal conversion optimization team—let alone the hiring process.
• Lack of intelligence: This isn’t about being smart, but more about common sense. Figuring out what exactly to test and further developments requires a deep dive into analytics but is intertwined with all business processes. Understanding the CRO process involves more than just looking at case studies and picking random elements to test.
• HiPPo’s: Because the field is relatively new, not all marketers understand the true value of testing. Sometimes, HiPPo’s (Highest Paid Person’s Opinion) can hinder progress.
  ‍
  

How does website speed affect conversion?

Load time is clearly a crucial part of the user experience, and it has a dramatic impact on conversions. According to surveys conducted by Akamai and Gomez.com, approximately half of internet users expect a website to load within 2 seconds. The same group of people is also likely to leave the website if it doesn’t load within 3 seconds.

A delay of 1 second in page load time results in:

• 11% fewer page views
• 16% decrease in customer satisfaction
• 7% loss in conversions
  ‍
  

What is the average conversion rate?

If you had to put a number on it, the average conversion rate is approximately 2.35%. That said, asking what the average conversion rate is can be quite ambiguous. This ambiguity arises from the fact that you can break down conversion rates by industry, niche, demographics, and so on. Additionally, conversion rates can vary based on the device and KPI.

The MarketingSherpa Ecommerce Benchmark Study addresses the issue of varying conversion rates across e-commerce industries by segmenting them based on product type. After surveying 2,885 marketers, they created a chart of 25 retail categories, displaying the average conversion rate for each:

They found that companies selling multiple types of products, like Amazon or Zappos, tend to have a higher average conversion rate. Companies selling only one type of product have an average conversion rate of 16.3%, while those selling multiple products have an average conversion rate of 17.2%.

Wolfgang Digital also conducted an extensive e-commerce study, looking at conversion rates by device. Their data showed that while mobile and tablet accounted for nearly half of their website traffic (43%), they contributed only a quarter of the revenue (26%). This is particularly highlighted when looking at average conversion rates by device.

Email marketing is a completely different story. The conversion rate for email marketing generally ranges from 2% en 12%. This wide range is due to the varying factors that depend on the industry and whether someone has signed up for the email list.

MailChimp has compiled a spreadsheet that breaks down open rates and click-through rates (CTRs) across 46 industries. They provide monthly updates on their email benchmark statistics, which serve as a valuable source of comparison when trying to determine if your email statistics are in good shape.

Let’s be honest, cold emailing can be tricky. Even the most recognized media outlets, such as FastCompany.com, struggle with cold email campaigns. Fast Company conducted an experiment where they sent 1,000 emails to busy executives to see how many would respond. “Of the 1,000 emails, 293 bounced. Of the remaining 707, 45.5% were opened… a great open rate for such a busy group.” Even with an impressive open rate, only 12 out of the 707 people responded. That’s 1.7%.

Now let’s dive into ad conversions. WordStream compared Facebook ads with those on Twitter by looking at four categories: network reach, ad performance, mobile ad performance, and ad format. According to their data, engagement rates on Twitter can be as high as 1-3%, which is much higher than Facebook's 0.119% CTR. The difference primarily lies in the fact that Twitter ads appear in the stream, while Facebook ads are shown on the side.

This significant difference may seem critical when deciding between using Twitter ads and Facebook ads, but looking at the average CPM reveals another side of the story. Twitter’s average CPM is $3.50 compared to Facebook’s $0.59.

The average conversion across all industries for Google Ads is 2.70% on the search network and 0.89% on the display network. The average conversion rate for Google Ads varies based on industry, just like website conversions and email marketing.