Loading...
13 April 2015

A/B Testing: A Complete Guide to Improve Conversions (Part-I)

A/B Testing: A Complete Guide to Improve Conversions (Part-I)
images
By Joydeep Bhattacharya

a-b-testing

A/B Testing isn’t a rocket science to understand and implement, yet it is no so common among the digital marketers. Similar to other major subject of Internet marketing like SEO, SMO, PPC Ad campaigns etc, A/B testing also helps in knowing about the insights visitor’s behavior, which is essential for improving the conversion rates of a website, but its benefits are always underestimated.

Similar to its name what it sounds like, A/B testing is performed on two different versions of a website (A and B) on specific set of metrics to define which of these two versions yields better results. Whichever version performs best is selected to be implemented as the final version of the website. If the A/B testing is planned in an appropriate way, it can help in creating buzz about your business in the marketplace.

Here in this article, I am going to present a complete guide about A/B testing that includes it comparison with other scientific optimization techniques, its impact on your website’s conversions and performance and the different ideas for implementing this testing successfully on your website.

Table of content:

Varied Testing Techniques
A/B Testing Vs. Multivariate Testing
A) Measure Website’s Performance

B) Prioritize What to Test

C) Performing A/B Test- Coming Up with Various Ideas

Varied Testing Techniques by Different Levels of Sophistication

A/B testing belongs to a specific category of Scientific Optimization techniques, however, there are two other categories’ Multivariate testing and Experimental design. The basic functions performed in all these three testing techniques are listed as follows:

  • A/B Testing: It involves testing of between two versions of same website to see which version drives better conversions.
  • Multivariate Testing: It involves testing several elements of a website to know which of these elements will help in achieving the overall goal of the website.
  • Experiment Design: It involves the creation of one’s own research method for performing a profound analysis of specific elements.

A/B Testing Vs. Multivariate Testing- Major Differences

While creating an experiment, Google might ask you to select between the A/B testing and Multivariate testing, but the former is always preferred over the later as it testes 2 varied versions of the same website while in Multivariate, you have to just test few important elements of your website to check the overall performance of the website. If you go from the visitor’s point of view, either of these methods would not make any change for them, but for a simple testing that requires fewer combinations with more changes, A/B testing is a good choice to go for.

A/B Testing Vs. Multivariate Testing

Before proceeding with the measures for calculating your website performance while conducting A/B testing, it is vital to know the actual definition of what A/B testing is? According to the standard definition, it is a method for authorizing that all the latest modifications introduced in the webpage to improve its overall conversion. The steps performed during A/B test includes:

  • Measurement of Performance
  • Prioritizing the test opportunities
  • A/B Test
  • Repeat

ANALYZEThis test confirms which of the two variations yields better conversions and this test usually involves lesser variations with extreme changes like change in headlines, background colors, icons, pictures, and navigation panels.

So, after getting familiar with the basics of A/B testing, the next important steps are to learn the key considerations for measuring your website’s performance.

A.Measure Website’s Performance- Key Considerations

Before deciding and implementing changes in a website for enhancing the conversions, the most essential step is to measure the website’s performance to know what are the actual conversions taking place on your website and reasons behind the low conversions.

1.Getting Actionable Data is Important via Google Analytics

Getting Actionable Data is Important via Google Analytics

If you want to receive the actionable data via Google Analytics for your website, it is important to set certain mandatory goals and objectives for your business as well as your website along with the adequate selection of key performance indicators (KPI). These are vital for all digital marketers to ensure that their website is generating the leads and sales as per their defined goals.

  • Set your Business and Website’s Objectives

Without specific goals and objectives of your business, you cannot decide or evaluate the overall revenue in the right way. For example, if you have an online store for selling garments, then your goals should be to increase the sales of garments through your website and once, this is set, you will be able to focus on the performance of your website to check if it is able to generate the sales as per your expectations or not. For increasing the conversions and sticking with your established business goals, you should have to:

  • Add clear and high quality images of garments on your product pages.
  • Put efforts for increasing the click through rates on your web pages
  • Reduce the shopping cat abandonment rate

Once, your goals have been set up, you have to select the define the Key Performance Indicators (KPI) of your website, then only you can perform the A/B testing in an appropriate way.

  • Select the Key Performance Indicators (KPI)

KPIs are metrics that are used for evaluating the current performance of your website and to know if it is performing in accordance with the defined goals or not. All metrics cannot be defined as KPIs, but a metric will be called as KPI only if it measures something related to your business’ objectives.

Suppose, your garments online store sells 60 garments in the last month and for defining KPIs for your online store, you need to define target for them. Like if you defined target for the KPIs as 200 garments per month, then you have a proper framework to know if your website is selling as per the defined goals or not.

Key Performance Indicators

Note: Typically, KPIs includes the combination of spreadsheets, charts, reports, and dashboards to display the information needed for a business success. Then this information is analyzed to report the results and take action if the results are not as per the established goals of the business.

2.Track the Behavior of Visitors

Once, you have collected the required data for performing the A/B testing on your site, you must track down the behavior of visitors on your website to know the user engagement rate on your site along with the duration of their visitation.

Visitor Behavior

Evaluating the visitor’s behavior helps in devising and finalizing the right strategies for executing the A/B testing perfectly on the website because anything that affects the behavior of visitors on your website needs to be A/B tested that includes:

Headlines

  • Sub Headlines
  • Links
  • Images
  • Paragraph text
  • Call to Action text
  • Testimonials
  • Awards and badges

3.Check Your Customer’s Feedback for Understanding What Could Bring Profits

Listen and analyze to the feedback of customers given on your website to know what measures taken during A/B testing could bring significant profits. Also carefully evaluate the latest trends in your customer’s feedback to understand the major or key factors affecting the overall conversions on your website. If something striking pops in your mind, thing profoundly and perform the A/B testing by changing that factor to see if positive results are afoot.

customer feedback

Besides collecting feedback of customers, it is essential to segment your target audiences to get the key insights into what valuable information could you receive that you can use for bringing profits. The segmentation of target audiences can be done in three categories:

segmentation

  • Segmentation by sources: Check for the different sources from where visitors landed on your site. Check which of these sources are redirecting loyal customers and what can be done to minimize the bounce of other sources.
  • Segmentation by Outcome: Check out and segment your audiences depending on the types of products they purchase since, all visitors have different requirements when they visit website.
  • Segmentation by Behavior: You can also segment the visitors depending on their behavior that includes the number of times they visited your site in a month and their purpose of visiting your site like are they visiting for comparison of prices or are they from different countries or regions?

Once you have analyzed the profitable segments of your targeted audiences, you can devise your A/B testing plan accordingly.

B.Prioritize What to Test

Once, you have chosen the metrics for checking the performance of your website by split testing or A/B testing, the next important step is to set priorities on what to test first, since a proper execution of any process needs a reliable start but try to A/B test those opportunities first that could promise positive outcomes.

1.Set Priority for Pages Not Performing Well

Set priority for the pages that are not performing as per the defined goals. This data can be received through Google Analytics that keeps track of the problematic pages and the pages that has high bounce rate. Suppose, you have a high shopping cart abandonment rate, then you cannot solve this problem merely by changing the shopping cart views, but you also have to check for your product and category pages for fixing this issue.

2.Set Priority for Top Exit Page

These are those pages that you visitors last see before leaving your website. These pages are labeled as “% Exit” pages in Google Analytics and it will display the percentage of visitors that leave your website after looking at the specific page at the last time. These pages help in recognizing the main area of concerns.

3.Set Priority Depending on Value and Cost

You should also set priority of testing high values ideas first that has low cost investment. For example, you can test the deviations occurring during checkout process as it shows the maximum abandonment rate than any other steps.

4.Set Priority of the Pages of Greater Importance

Apart from setting priorities for the worst performing pages, it is also necessary to set priority to test those pages first that have highest volume of traffic because these tests finish quicker and also help in fastening of your optimization process.

C.Performing A/B Test- Coming Up with Various Ideas

When it comes to perform A/B testing, you can test any element of your website to check if it’s performing as per the defined goals or not. You can A/B test by changing any of the given elements such as:

A/B Tesing Ideas

  • Website Navigation

This is a very major area of a website that could be improved with A/B test. Most of the sites have two sets of navigation menus- primary and secondary. The primary includes the main categories of products, while the secondary includes the FAQs, About Us and Contact Us etc. While primary navigation is integrated well into the website, you can test by changing the secondary navigation for known the important and popular options on the top of your web pages.

  • Call-to-Action Buttons

You can perform split testing on call-to-action button by changing their colors, size, and wording.

  • Landing Pages

The A/B test can be performed by creating variations of the landing pages by changing their design, background, colors, etc. to evaluate which of the variation yields better conversion.

  • Headlines

It is vital to do A/B test on your headlines because headlines are responsible for increasing the user engagement and conversions on your website. If not impressive, they might not convert the potential traffic to sales, which is the core aim of all businesses.

  • Images

It is common human nature that people are visually attracted more towards the graphics.  Hence, A/B testing on images is must for understanding if the mere modifications on the images can help in generating leads as per the established goals.

  • Social Proofs

Once again customer feedback and their logos have greater impact on the mind of your upcoming visitors because they impact a sale and conversion significantly than any other elements on your site. If they are not impressive, it might drive traffic away from your website. Therefore, make sure to A/B test on social proofs.

  • Pricing & Content length on the page

Ask yourself what things on your site motivate your visitors to click on your products and services? Is this an advertisement or an offer that is catching the attention of your target audiences and inspiring them to make a purchase? Or the content length on your site is impacting negatively on the purchasing or clicking decisions of your clients? So, testing pricing and content length is valuable to check which of the given matches up your expectations.

With this, you might be thinking if there are certain tools for doing A/B testing accurately? How you can do A/B testing with Google Analytics? Does it have any impact over the SEO of your site? Read our next article to stay tuned with the information on the above queries.

Next Part

Recent Article

You May Also Like

White hat vs Black hat
images
By Amit Kothiyal
21 July 2012