Optimize your Websites for more leads with Kentico CMS 6.0

If we would like to optimize our website, we can optimize it for search (Search Engine Optimization), for an on-line advertisement (e.g. PPC optimization), to convert our visitors to potential leads, and more and more. In this article, we will focus on the online conversions, i.e. on “How to optimize a website to get more sales leads”.

Before we start the optimization, we have to decide how we will identify that our optimization is going well or not. The measure often used for it’s called a Conversion. A conversion occurs when a prospective customer takes the marketer's intended action. The conversion action can be making an online purchase, submitting a form to request additional information, subscribing to a newsletter and others. This part is very important and can rapidly affect how we will compare the effectiveness of the optimization.

After we have identified the right conversions, we can compare effectiveness of our optimization based on achieved Conversion rate, which is the percentage of visitors who take the conversion action. So, if we have identified wrong conversions, we can select the wrong website optimization based on the conversion rate!

What are the main steps for preparing and running the website optimization that we can follow?

The base of website optimization is optimization of a web page. The most common is optimization of the Landing page (LPO). Current website optimization knows two types of optimization: A/B testing and MVT (Multivariate testing).  

A/B Testing

A/B split testing is a method of marketing testing that involves two different versions of a web page to see which of them is more effective. In a basic A/B test, a random visitor is shown either the “A” web page, which is generally the current page, or the “B” web page, which is the new testing page.

The reaction of the visitor is measured by the conversion. Based on the conversions, we can determine which page, page “A” or page “B,” will become the new page for the website.

And when to use the A/B testing?

E.g. we would like to change the design and layout of the home page and we want to know which design is better – the new or the old one?

Kentico CMS version 6.0 brings A/B testing support.  And the process of preparing and running any A/B test in Kentico CMS is straightforward.

Figure 1 - A/B test process

 

Prepare pages

A/B testing in Kentico CMS allows us to test not only TWO pages, but we can test as many pages as we want!  We can test any existing pages in the content tree or create a new page variant as a copy of an existing one.

 

Identify conversions

We extended the number of conversions that can be tracked in Kentico CMS 6.0. We will be able to use the following conversions:

-          Page view -> in case we would like to track how many visitors visited a particular page

-          Newsletter subscription

-          User registration

-          Ordering of a product

-          BizForm filling

-          Submitting a Poll

-          Custom activity -> if we will need any special conversion, we can track it through a custom code

Using A/B tests in Kentico CMS 6.0, we can track more than ONE conversion through a test. We can also run these tests inside any Campaign.

 

Setup an A/B test

Having pages and conversions prepared, we can setup an A/B test. The setup is variable so we can set:

-          Date-time period within which the test will be running

-          No. of conversions after which the test will finish

-          or we can just manually start/stop the test

 

Run and analyze the test

Depending on the settings, the test will be running automatically or we can run it manually.  If you want, you can run MORE THAN ONE A/B test in a time.

We can start monitoring the progress of the test via the Web Analytics and Reporting modules. The Web Analytics module will be very improved in version 6.0, including:

-          Using MS Charts, which supports modern graphs, interactive graphs etc.

-          Better look and feel of the Web Analytics UI,  similar to Google WebAnalytics.

-          Web Analytics will contain e.g. management of Conversions, Campaigns, A/B tests, MVTs and others.

-          We can also use a Dashboard, a brand new feature in Kentico CMS 6.0.

Figure 2 - Preliminary Web Analytics UI

 

Select the best variant

When the test is finished, we can decide which variant is the best. In Kentico CMS 6.0, we can use A/B test reports which help us better understand A/B tests and we can understand e.g. conversions per each test, conversion per each page, effectiveness of any page, etc.

Having the A/B test winner, we can replace the original page (page which was tested) with the new version very easily. 

 

A/B test scenarios

In Kentico CMS 6.0, we can create several A/B testing scenarios. From simple A/B split tests of two pages to very advanced and complex multi-page A/B tests, where one page could participate in more A/B tests.

Look at the following schemas:

Figure 3 - Simple A/B split test

 

Figure 4 – More complex A/B test. ONE page participates in two different A/B tests.

MVT (MultiVariate Testing)

Multivariate testing is similar to A/B testing, but it’s focused on simultaneous testing of multiple changes of ONE page. Using MVT, we are not preparing and testing several pages, but we are testing different variants of particular parts of a page.

E.g. there are two parts of our page, which we would like to test:

1)      User Registration Button – we would like to discover which color to use so that a visitor will register. (we are thinking about 4 colors)

2)      Newsletter subscription – what’s the best position on the page (we are thinking about 3 positions)

If we wanted to cover all possible combinations in an A/B test, we would have to prepare 12 pages. In MVT, we are using the same page; we just define and test variants of each part

We can use Kentico CMS 6.0 for MVTs and if we look at the following schema, we can see that the process of preparing and running an MVT is similar to A/B testing.

Figure 5 - MVT process

 

Prepare variants

If we want to start an MVT, we have to identify which parts will be tested. And then prepare variants for them.

In Kentico CMS 6.0, we can test different variants of:

-          Webpart zone -> we can create different versions of the whole zone, e.g. if we would like to find the right order of webparts inside the zone

-          Webpart -> e.g. we would like to test a different “TOP Product”

-          Editor Widget zone -> content editor can prepare variants of the widget zone

-          Editor Widget -> content editor can prepare variants for the widget

 

Identify conversions

This is the same as for A/B testing. We can use the same set of conversions.

 

MVT test setup

Similar to A/B testing where we can setup up how long will be the test running.

As there could be “tons” of possible variant combinations, you can also set a list of combinations that will be/not be tested during the test.

 

Run and analyze the test, Select the best variant

This is very similar to A/B testing, but we will use prepared MVT reports.

A/B testing vs MVT

There is one most common asked question:” How does an MVT differ from an A/B test?”

A/B split testing offers the following advantages:

1)      A/B testing is quick and easy to set up. We can think of a new page headline over lunch, implement a test, and see results within a few hours or days.

2)      A/B testing offers unambiguous results. If we are only testing a single page with two variants, the results are typically available quickly.

3)      If we are optimizing pages of a new website, we may want to test an existing page against a page containing all of the "best practices" we have discovered for designing landing pages, etc.

Multivariable testing should be used in different circumstances:

1)      It takes some time to set up a multivariable test. We have to design and implement all of the individual page variations at once.

2)      However, if we have a number of variations and combinations to test, multivariable testing can save us much time and aggravation in the long run.

If the site is small in terms of pages and visitors, A/B Testing will work best because it is more manageable to evaluate and interpret changes in behaviors. Also, if we would like to test basic changes on the site, such as a video on a landing page versus a photo on a landing page, A/B Testing will be able to show the results that are easy to understand and analyze.

Multivariate testing is best used for bigger sites with large amounts of traffic.

Conclusion

A/B testing and MVT are modern marketing/business website optimization techniques which can help you to increase the number of potential sales leads.

We can use any of the existing 3^rd party tools like Google Webite Optimizer (http://www.google.com/websiteoptimizer), Omniture (http://www.omniture.com) and others.

 

Or wait until Q2/Q3, when Kentico CMS 6.0 will be released!  A/B testing and MVTs will be a part of the Advanced package …

 

Stay tuned as we will provide you with more information about A/B testing and MVT as soon as we are finished with the development of these features!