1. Home
  2. Docs
  3. Using etracker
  4. etracker Optimiser
  5. Testing & Targeting
  6. A/B testing & smart messages

A/B testing & smart messages

When creating tests and smart messages, an assistant is there to help you query all of the necessary data and save it to the corresponding project.

During the runtime of a test, etracker informs you under etracker OptimiserTesting & TargetingProject overview & creation of how long the test will run roughly and ends the test when a significant amount of data has been reached. A test runs for at least seven days. For each variant, there must be at least seven conversions and 200 visitors need to be tracked. The runtime is longer the closer the conversion rates of the page variants are to each other because greater selectivity and thus more traffic is then required to reach a significant result.

To execute a simple A/B test, your website should have at least a few thousand visitors per month.

Project overview

You can find the test overview under etracker OptimiserTesting & TargetingProject overview & creation. It shows all of your projects and their status. The latest projects are at the top. You will find the following information on your projects:

  • The “Project name” column shows the URL of the original page as well as the name you have chosen for the project when you created it. It is linked with the detail page of the project.
  • The column “Visitor conv.” shows how many visitors your project had and how many conversions there were.
  • The “Conv. rate” column first shows the conversion rate of the original page. When a test has determined a winner, then this value will be shown.
  • The condition of the project is shown in the “Project result” column.
    • If the test has not provided any result yet and needs to continue to run, then the approximate remaining time will be shown here.
    • If there is a variant which has a higher conversion rate than the original page, then it will show how much greater the improvement was.
    • If the test has shown that no variant performed better than the original, then this will also be shown.
  • The “REC” icon shows if you have started or paused the project. When it is paused, then all visitors will be shown the original page.

To create visual A/B and Split URL Tests as well as on-site personalisation, click on the button Create project or Create additional project. To stop or delete projects, click on the Select link in the last column. There, you can also navigate to the Project details page

Creating projects

To find ideas for a test, you first need to uncover weak points in your website using Web Analysis. Find out, for example, which elements of the page influence conversion the most. In the test, you can vary the elements which increase conversions. With the analysis, the Visitor behaviour report under Testing & Targeting can be helpful. Here, compile hypotheses like, for example, “We have a high bounce rate on the homepage because customers don’t understand the product features” or “We have a high bounce rate because the customers don’t trust our marketing claims”. Proceed individually as follows:

  1. Think about whether you can prove or disprove the hypothesis by varying page elements.
  2. Then plan for which concrete changes you want to test for, write headlines and copy, search for pictures, design changes in the layout to the page, come up with changes to the functionality.
  3. Using the planned variations, decide if you would like to implement a visual A/B test or a split URL test (see below).
  4. Define key figures which track the test success. Along with the conversion rates, clicks or sales can be used here, for example.

The two test types differ in the following points:

  • Visual A/B tests can be created with the Visual Editor. This loads the original page and enables the selection and changing of elements without any programming. When exporting the page to the visitors, the changes to the HTML code of the page will be ‘injected’. This is also referred to as ‘DOM Manipulation’.
  • Split URL tests require some preliminary work. You (or your webmaster) can programme different variants, for example, by copying and changing the original page. The original page and the variants need to be entered as stand-alone URLs. The pages are either separate HTML files on a web server or contain a dynamic template which can be controlled by parameters in the URL in such a way that it provides different variants.

The following variations can be executed with both test types:

  • Changes to headlines and copy
  • Alternative pictures
  • Variations of affiliate advertising (strictly speaking: just variations to picture and text)
  • Switching stylesheets in order to test out different layouts
  • Switching the position of page elements

Make your decision according to the following criteria:

CircumstancesDecision
Very fast and uncomplicated changes to the original without programming (e.g. integrating only alternative texts or pictures).Visual A/B test
You have little technical expertise.Visual A/B test
You have technical expertise in CSS and jQuery.Visual A/B test
for creating variants with extensive layout changes
You cannot open the original page in the Visual Editor (e.g. because the page contains JavaScript or Flash which is not compatible with the Visual Editor).Split URL test
Your server must send out alternative HTML code for a variant. Split URL test
You wish to change a function (e.g. navigation).Split URL test
You wish to branch into an alternative multi-page function, e.g. an alternative checkout process.Split URL test
You wish to perform a test with Google AdSense.Split URL test, as the Visual Editor cannot be used here.
You wish to test a variation of advertising.Split URL test or visual A/B test
You wish to change the position of page elements.Split URL test or visual A/B test

If the Visitor behaviour report under Testing & Targeting shows, for example, a high bounce rate, you can also create a smart message project independently of any A/B tests by creating a smart message which goes only to visitors who are likely to bounce (e.g. a reference to a Help page or a product video of a product detail page or a specific offer). When creating a smart message, you should consider the following beforehand:

  • Determine on which pages it makes sense to place the triggers ‘Exit Intent’, ‘Greeter’, ‘Attention Grabber’, ‘Always on’, ‘Scroll Position’ or ‘JavaScript Event’.
    • Exit Intent: The visitor wants to leave the site.
    • Greeter: The visitor just arrived at the page.
    • Attention Grabber: The visitor is inactive.
    • Always on: The visitor’s behaviour is irrelevant. The smart message is always shown.
    • Scroll Position: The visitor scrolls down to a specific position.
    • JavaScript Event: The visitor triggers a specific event on the page.
  • Think about what you can offer visitors in order to keep them on the page.

You can also analyse different variants of a smart message in an A/B test project.

Creating visual A/B test or smart message

You can use the etracker Optimiser’s Visual Editor in simple A/B tests and for smart messages in order to vary, among other things, texts and pictures of your page without any programming. To create a smart message or a visual A/B test, proceed as follows:

  1. Select etracker OptimiserTesting & TargetingProject overview & creation.
  2. Click on Create project.
  3. Click on Create smart message or Create visual A/B test. In the following steps, you need to select a template for the general design and layout for a smart message. This is completed by clicking on Enter URL.
  4. Enter the URL of the original page and click on Visual Editor. A dialog window with help information is shown. In compatibility mode, your page will be listed on our server, processed and then opened in the Editor. Use this mode if you have not yet already entered the etracker code into your page. In default mode (compatibility mode switched off), your page will be opened directly in the editor. So that this is possible, the etracker code needs to have been entered into the page. If it is not possible to load this page, then please switch to compatibility mode.
  5. Click on Edit page now. Using the Visual Editor you can then select and vary individual elements on a page without any programming (see below). Each original page and all variants will be shown in separate tabs in the Visual Editor. Alongside the tab for the original page, a tab with the name Variant 1 also shows. Clicking on the Tab name shows the selected variant in the processing window. If a page has password-protection, requires a full basket or can only be loaded on certain computers (perhaps because it is part of a protected company network), then it will not be possible to open it directly in the Editor. You first need to open an openly accessible page on your website and switch from Edit mode to Navigate mode in the Visual Editor in order to access the page you wish to edit. Here, you can place items into the basket in order to, for example, access the checkout process. You can then switch back to Edit mode and create different page variants.

    Note: Some websites cannot be loaded and edited with the Visual Editor if they are heavily dependent on JavaScript or Flash. This can be recognised by these effects:
    • The page does not appear, that is, you see a loading symbol which does not disappear.
    • The page can only be partially seen after loading.
    • You cannot click on or edit the elements of the page. In these cases, however, you can always perform a split-URL test.
  6. Once you have edited the variant(s), click on Define segmentation in the Visual Editor.
  7. Decide if you wish to segment and, if you want, use the Pen symbol to define a segmentation rule before selecting Complete segmentation. Example: You want to present a special offer to users who access your website on a tablet and have already purchased from you once before. Here, you need to set up a page variant. For segmentation, select Segment all variants and create a segmentation rule with the conditions ‘Buyer type: Just one purchase’ and ‘Device type: Tablet’, where all conditions need to be met. An overview of the segmentation rules for smart messaging & testing can be found at here.
  8. Provide a project name, enter a URL for the page(s) to be sent out and click on Define conversion targets. Enter the URL of the original page or a part of it. If there is a match between the page a user opens and the chain of symbols you entered here, then those variants of the page will be sent out once the project has started.
  9. If possible, define only one conversion target and click on Save and Create project. The project detail page opens.
  10. Check the test or the smart message and click on Start project.
  11. Check in the test analysis if visitors and conversions should be counted or not.
Switching between the Visual Editor and different end devices

To customise a variant specifically for display on desktop computers, smartphones or on tablets with the Visual Editor, you have the option to switch between different end devices in it. At the moment, the following end devices are available:

  • Apple Iphone 5 (320×568)
  • Apple Iphone 6 (375×667)
  • Samsung Galaxy SIII (360×640)
  • Samsung Galaxy Note 3 (360×640)
  • Apple Ipad (1024×768)
  • Google Nexus 7 (960×600)
  • Desktop

Here, select an end device in the dropdown menu under Open page with end device. The display of the website in the Visual Editor will then be shown accordingly.

Editing variants in the Visual Editor

In the Visual Editor, if you move your mouse over a page which shows a variant, the Editor will then identify the elements of the page and highlight them. Here, a Page element can be, for example, a block with text or a picture, or also a grouping, like a teaser made of a headline, copy text and image. This depends on how the page is structured in the HTML code. A highlighted page element can be edited, hidden, removed or moved in the Visual Editor after it has been clicked on. Advanced users can also edit a variant by inserting additional code (CSS or JavaScript) into the page (see below). Note: Please note that the changes will only be saved to the database at the end of the creation process after clicking on Save and Create project. Creating, renaming and deleting variants in the Visual Editor

  • To create a new variant, click on the +-tab.
  • To rename or delete a variant, move the mouse to the arrow beside the name of the tab and click on Rename variant or Delete variant.
Creating, renaming and deleting variants in the Visual Editor
  • To create a new variant, click on the +-tab.
  • To rename or delete a variant, move the mouse to the arrow beside the name of the tab and click on Rename variant or Delete variant.
Editing page elements in the Visual Editor
  1. Select an element for editing and click on it. A small menu will open with options for changing the element.
    • To increase the size of the editing area, select Select parent element. Example: You have selected a list element, but wish to edit the entire list.
    • To change text content, select Edit text.
    • To change pictures, select Edit image URL.
    • To change formatting, select Edit HTML or Edit CSS.
  2. Click Save.

    Example: Changing an image in the HTML code
  3. Make the new image available on your webserver.
  4. Click on the current image in the Visual Editor and select Edit HTML.
  5. Enter the name of the new image into the HTML code (after ‘src=’).
  6. Click Save. The changed picture will appear in the website view.
Hiding or removing page elements in the Visual Editor

In the Visual Editor, you can make selected elements invisible or remove them. When hiding them, a transparent placeholder will remain where the element was. If you remove an element, other elements will move up to where it was (if the programming of the HTML page specifies that).

  1. Click on an element you would like to hide (remove). Another small menu will open with options for changing the element.
  2. Click on Hide element (Remove element).
Moving page elements in the Visual Editor
  1. Click on an element and select Move element..
  2. Drag the element to the desired position. Please note that the other page elements do not change their position.
  3. To cancel the movement, click on the red X. If you are happy with the new positioning, click on the green arrow to confirm the action.
Inserting CSS code into a variant

If you have expertise in developing cascading stylesheets, then you can manipulate the layout of the variants by adding CSS code. To do so, proceed as follows:

  1. Move the mouse onto the arrow in the tab of the corresponding variant so that the menu opens out.
  2. Select My CSS.
  3. Insert corresponding CSS classes into the dialog. You can overwrite existing classes.
  4. Click Save. The changes will be shown immediately in the variant.
Inserting JavaScript code into a variant

Along with CSS, you can also insert JavaScript into the variant. This way, you can also use the jQuery library and reconstruct the variant in whatever way you like. etracker Optimiser supplies its own jQuery library. Documentation can be found, for example, at https://api.jquery.com/. To insert, proceed in the same way as you did when inserting CSS code:

  1. Move the mouse to the arrow in the tab of the corresponding variant so that the menu opens out.
  2. Select My JavaScript.
  3. Enter jQuery programming code into the dialog.
  4. Click Save. The changes will be shown immediately in the variant.

Creating split URL test

If you wish to perform a split URL test, then you will need to create variants beforehand and make them available on a webserver.

Tip: If the variants differ in just a few respects, it may be sufficient to enter some program logic into the original page and use URL parameters to control which variants will be shown, e.g.:

  • For each variant, you can affix a different URL to the original page, as follows www.originalurl.php?version=1 or www.originalurl.php?_v=green etc.
  • Add JavaScript code to the original page, evaluate the URL parameter and change the page accordingly.
  • Enter server-side code (e.g. Java or PHP), evaluate the URL parameter and change the page accordingly.
  • Use relative URLs to define the original page and the variants for the test (see the example for optimising templates below).

To create a split URL test in etracker Optimiser, proceed as follows: Select etracker Optimiser Testing & Targeting Project overview & creation.

  1. Click on Create project.
  2. Click on Create Split-URL-Test.
  3. Enter the URLs of the original page and your variant(s). You can enter as many variants as you like. For better comprehension, you should also enter a name for each variant. The variant names also appear in the reports.
  4. If you have defined all variants for the test, click on Define segmentation.
  5. Decide whether or not you want segmentation and define a segmentation rule, if necessary, using the Pencil symbol before selecting Project name, URL and Test period. Example: You decide against segmentation and click on Without segmentation. Then each of your created variants will be sent out to every visitor provided the original page is not randomly chosen.
  6. Give the test a project name, enter a URL for the page(s) to be tested and click on Define conversion targets. Enter the URL of the original page or a part of it. If there is a match between the page a user opens and the chain of symbols you entered here, then the test will be activated and the variants of the page sent out.
  7. If possible, only define one conversion target and click on Save and create project. The project detail page opens.
  8. Check the test and click on Start project.
  9. Now check in the test analysis if visitors and conversions should be counted or not.

Optimising templates in the Split URL test Sometimes a page (and its variant) does not have a defined URL. This can happen if you wish to optimise a page type (a template) for an individual project page in a shop system. Here, you can use the original page itself as a variant in etracker Optimiser and give it a URL parameter. If you wish to adjust the code of your template accordingly, the template evaluates the URL parameter in order to show the variant.

Example:Original page: For the original page, enter the URL of any template page you wish to optimise. In a shop system, you could, for example, use one of the two following URLs for product pages:

  • http://www.abcshop.com/Jersey-skirt/
  • http://www.abcshop.com/Men-blouson-black/

etracker Optimiser uses the URL entered to show a preview of the page and its variants. However, only the placement of the tracking code in the template determines which page will really be used as the original page to be tested.

Variant(s): Instead of entering a URL, enter something according to the following formula for the variant:

?<parameter>=<value>

Here, the question mark is a placeholder for the opened original page. Enter something that makes sense for <parameter> and <value> so that you are transferring a specific parameter:

?_v=1

When etracker Optimiser sends out a variant, then it is called as follows:

http://www.abcshop.com/Men-blouson-black/?_v=1

Note: In the same way, you can also optimise dynamic results pages.

Defining URLs for the project pages

When creating and editing projects, you need to define a URL in the field after ‘URL for the page(s)’, which describes the pages you wish to send out. If the URL matches a page called by a visitor, the test or the smart message will be executed, that is, conversions will be counted.

Note: If you have selected Call target page during the last created/edited item as the conversion target, then an additional input field will be shown in which you can enter the URL of the target page. If this matches the page a user opens, then this will be evaluated as a conversion.

If you enter (everything after the “?”) for the URL parameters, then the test or smart message will only be executed if a visitor calls the page with the parameters.

Example: Entry without URL parameter

http://www.yoursite.de

The project then runs on the following pages:

  • http://www.yoursite.de
  • http://www.yoursite.de/
  • http://www.yoursite.de?campaign_id=123456

In this case, the project does not run on the following pages:

  • http://www.yoursite.de/agb/
  • https://www.yoursite.de

Example: Entry with URL parameters and * as placeholder

http*//www.yoursite.de/product.php?id=* ein.

The project then runs on the following pages:

  • http://www.yoursite.de/product.php?id=10
  • https://www.yoursite.de/product.php?id=10
  • http://www.yoursite.de/product.php?id=

In this case, the project does not run on the following pages:

  • http://www.yoursite.de
  • https://www.yoursite.de/product.php

So you can copy and use the entire URL from the address line of your browser.

Caution: Sometimes there are components in the URL, like Session IDs, which are different from one visitor to the next. If you copy this into the input field, then there won’t be a match with the entered URL for any other visitors.

Example: Entering a URL (copied from the address line)

http://www.yoursite.de/products/?sessionid=2490435sd

In this case, the project does not run on the following pages:

  • http://www.yoursite.de/products/?sessionid=d4jlaw098
  • http://www.yoursite.de/products/

For this reason, it is better to enter just one part of the URL and use ‘*’ as a placeholder. So in this example, you should enter the following URL:

http://www.yoursite.de/products

Caution If you run multiple projects simultaneously, it is particularly important to not leave the partial URL of the page too short and thus possibly ambiguous. Otherwise, it can happen that the page which a visitor opens matches multiple projects.

Example: Entries for Test 1 and Test 2:

http://www.yoursite.de/products/shirts/*
 
*www.yoursite.de/products/*

When the visitor opens the page http://www.meineseite.de/products/shirts/?id=10, this will match both tests. Which test will then be run cannot be foreseen.

Tip: You can specify the page on which a project should run with a regular expression. Enter an exclamation mark as the first character into the input field and then the regular expression; like this:

!/(www.yoursite.de$)|(yoursite.com$)/i

This would then execute the project for the following pages:

  • http://www.yoursite.de
  • https://www.yoursite.com

There are cases when you cannot define the page on which a project should run by entering the URL or part of it into the etracker Optimiser. This can be the case in the following situations:

  • You want to optimise a page in a purchasing or registration process for which all steps of the process have the same URL. This happens when the parameters between the individual steps are transferred during the technical implementation using either session variables or POST.
  • You have a website for which all pages have different URLs and you cannot discern the page type from the URL. This could be the case, for example, with a WordPress blog in which you have both ‘Posts’ and ‘Pages’.

Here, etracker Optimiser offers an alternative to identify the pages of a project: You can set a JavaScript variable ‘et_pagename’ and enter its content in the URL for the page(s).

Tip: The entered URL must include the URL of the original page. Furthermore, you may not use the same URL in two different projects. With the help of the diagnostics function you can check if a project can run according to these criteria. Here, select a project from the project overview. You can then select Diagnostics under Edit project.

Defining number of project participants

When editing a test, except for the test name and the URL for the page(s) you can enter a percentage under Number of project participants. This percentage indicates the size of the sample for the A/B test, that is, the percentage of all visitors to the original page who should take part in the test. If you enter a value, like, 40%, then 40% of visitors will see the original or a variant which is shown at random. The remaining 60% of visitors will only see the original. The value is set to 100% by default. For smart messages, you can reduce the share of project participants here. If, for example, you only wish to send out an ‘Always on’ message to every second visitor at the start and have not defined any segmentation, set the value to 50% here.

Defining conversion targets

When creating or editing a project, you can define one or more conversion targets. A conversion target is the desired action of a visitor. This can, for example, be the following:

  • Completed purchase
  • Registration
  • Click on an advert
  • Click on a product video which you offered in a smart message.
  • Completion of a step in a multi-step registration form
  • Engagement with your website, that is, any interaction that is not immediately leaving the site after opening it (“bouncing”)

To track conversion targets, the following visitor actions can be recorded.

  • Clicking on a link (additional conclusions can be made from the link URL, for example, if it is affiliate advertising)
  • Sending a form
  • Display of a page (e.g. a thank you or confirmation page during a purchase process)

Which visitor actions are tracked by etracker Optimiser as conversions can be selected when creating or editing a project. You can define one or more conversion targets when you create a project. After clicking on Save and Create project, the Project detail page will open. There you can start the project once you have checked it. If a project participant performs a corresponding action, a conversion for the corresponding page will be counted.

Note: Please note that in a project a conversion only counts once for each participant, even with multiple page views, regardless of how often the page is viewed or how often they perform actions there. Which of the selected actions they have carried out is not of importance either. For this reason, you should only select one conversion target. The result is more meaningful in this instance.

Engagement as a conversion target

Engagement means that the visitor interacted with the page either by clicking on a link on the page or by sending a form. Engagement is often counter to the bounce rate. The bounce rate refers to the share of visitors who only open one page and then leave the website. This happens, for example, when the page was found on Google but does not meet the expectations of the visitor. If the engagement of your website is low (and the bounce rate high), then your chances of reaching your actual targets like sales or registrations will be low. For this reason, it can be worth it to optimise engagement and to do so by using this shortcut to improve the meeting of actual targets.

Tip: Use engagement as a conversion target if you only have a low number of visitors or a low conversion rate for your actual targets (e.g. sales). A test to optimise the actual targets would run for a very long time. Because the conversion rate for sales is often just a few percent or even less than one percent when engagement can easily be 40% or more. A test with engagement as the conversion target will lead to results much more quickly. However to get a valid result, your actual target (e.g. sales) should relate directly to engagement. There are cases where engagement increases significantly, but the sales drop.

View of a target link as conversion target

If a visitor to the website clicks on the link whose URL you have entered here, this will be counted as a conversion. You can use this, for example, to track how linking to an action page is accepted by the visitors. You can use the ‘*’ character as a placeholder.

Self-defined (JavaScript) target as conversion target

If you cannot track a visitor action with the above-mentioned options, then you will need to use a Self-defined target. This can be necessary, for example, in the following cases:

  • You would like to count the clicks on a specific link as a conversion.
  • You would like to track the click on an affiliate ad and you have hidden the link.
  • You would like to send off a form.

If you would like to track a link to the basket, you would proceed as follows (for example):

1. To define a JavaScript target, edit the HTML code of your page in such a way that the following JavaScript function is called for the visitor action you request: _bt.trackCustomGoal(eventname) The ‘eventname’ should make it clear what should be clicked on so that conversion is tracked.

<a href="basket.php" onMouseDown="_bt.trackCustomGoal('BasketLink');">To basket</a>

Caution: EIt is important that the _bt.trackCustomGoal() function is opened with the JavaScript event ‘OnMouseDown’ and not with ‘OnClick’. This is because with some browsers, the latter can open the linked page before etracker Optimiser had the chance to count the conversion.

2. In the Project Assistant, select the self-defined JavaScript target as the conversion target and enter the event name you used in the HTML code (in the example: BasketLink).

Calling a target page as a conversion target

You can easily track different targets by defining as a target the opening of a specific page by the visitor. This way, purchase or registration processes almost always have a confirmation page where the completed action can be determined. To define a page like this as a test’s conversion target, select Target page view as the conversion target. In the dialog, an additional input field will be shown into which you must enter the URL of the target page. You can use the * character as a placeholder.

Note: If you have selected View target page as your conversion target, then an additional input field will be shown when you create your next test where you will need to enter the URL of the target page (or part of it). If this matches the page a user opens, then this will be evaluated as a conversion.

Conversion targets in the ordering process

In eCommerce, it can be helpful to define interim targets in the ordering process. In a test of etracker Optimiser, you can track the targets Product seen, Product in basket and Order as conversions. ‘Orders’ will be tracked by etracker parameters and the eCommerce API, while ‘Product seen’ and ‘Product in basket’ will only be tracked by the eCommerce API.

Reaction to smart messaging

When a visitor clicks on a button or link in the smart message which is predefined as a Call-to-action, this reaction will be counted as a conversion.

Note: etracker recommends tracking this conversion target for smart messages as this way the monetary effect of the smart message is measurable, however you can also select different conversion targets.

Checking project (project details page)

As long as you don’t start the project, no variants will be sent out. As such, you can check the project in peace and then activate it. Once the project is created, etracker Optimiser will show the Project detail page. Here, you can see the progression of the conversion rate over time in a graphic, which you have chosen in the Display column. The conversion rate of the original page will only be shown for tests and is highlighted in black. In addition, the project detail page shows the individual values for each page of the selected project in a table:

  • Participants (overall/converted) Shows how many visitors took part in the project and how many of them reached a conversion target. The values will be shown for all pages in real-time. In smart message projects, only those participants will be shown who have visited page variants.
  • Conversion rate Shows the conversion rates of variants and the original page.
  • Project result Here, the result of the project is shown for each page in the tests. If the conversion rate of a page is not yet tracked precisely enough (that is, if the result is not statistically significant), then the accuracy of the result will be shown. Here, 50% accuracy is the lowest possible value. Accuracy of 95% is desirable. An accuracy of 95% and more will show if the variant converts better or worse than the original page. If a variant converts better than the original page, then the improvement of the conversion will also be shown.

Proceed as follows to check/activate the project:

  1. Open a new tab in the browser and navigate to the target page. In your browser, open the HTML source text and check if the tracking code is contained in it. Go back to the project detail page in etracker Optimiser and open the preview for as many of the variants you have created as possible. Here, click on the Preview icon to the left along with the variant name. Then you can check if the variant was displayed correctly..
  2. If all checks have passed successfully, you can activate the project. To do so, select Start project on the project detail page. The project is now live.

After activation, check that the participants and conversions are being counted:

  1. Open the project detail page and note down the numbers of the participants shown (total and converted). Both of them are probably 0. Open the original page. During a random test, etracker Optimiser shows either this page or a variant of it. With a smart message, you need to imitate the trigger behaviour in order to show the smart message. Variants and smart messages will not be shown to you if you exclude yourself from being shown them using an IP block.
  2. Carry out a conversion as a test. Then load the project detail page again via etracker OptimiserProject overview & creationClick on project name. The participants (total and converted) must now be shown for the page you viewed.

Editing or deleting projects

To edit an existing project, do the following:

  1. Select etracker Optimiser > Project overview & creation.
  2. In the project overview, click on the project. The project detail page opens.
  3. Move the mouse to the top right onto Edit project and select the desired editing step.

Note: After any major change to the pages, you should restart the project. Then all counted participants to date will be discarded. To restart a project in the Project overview table, click on the Select link in the line with the project name and click on Start project. To delete a project in the Project overview table, click on the Select link in the line with the project name and click on Delete project.

Pausing project

To pause a project, select etracker OptimiserTesting & TargetingProject overview & creation and click on the REC icon in the line with the project name.

Evaluating projects

Once you have documented all hypotheses and their associated tests, supplement them with the current project result and record if your hypothesis or hypotheses could be confirmed or disproved. The more insight into the behaviour of your visitors and customers you have, the better you’ll be able to predict their behaviour and optimise your website. Even negative project results provide findings which can later lead to improvements.

The report under Smart messages & tests Visitor behaviour now has special importance. You should view it before setting up a project in order to define the target course. Here, for example, you can see which pages require optimisation, which visitor groups have high bounce rates and how certain visitors behave during the purchase process.

The reports under etracker OptimiserTesting & Targeting Smart message analysis show you which smart messages are successful.

In the Traffic sources report, you can check from where visitors to the individual variants are coming so as to evaluate whether or not the random sampling could have influenced the result.

The Visitor profile report reveals further information about the visitors. In both reports, you can check the variants respective to sales. The sales in the reports are always visible, even if you have selected “Engagement”, for example, as the conversion target. This way, you can always keep an eye on whether or not sales have also developed positively. In the same way, the reports under etracker Optimiser Testing & TargetingTest analysis show you which tests are successful.

Note: Please note that the number of attributes and key figures in the reports depends on your etracker edition.

Attributes

AttributeDescriptionSample values/Comments
TriggersVisitor behaviour which triggers a smart message.‘Exit intent’ (the visitor wants to leave the website)
‘Always on’ (the smart message is always triggered regardless of the visitor's behaviour)
Frequency of visitsCategory for the number of previous visits of an individual visitor. ‘Low’ (two visits)
'High' (5 or more visits)

Average order valueAmount of average order values in €.‘0-10’
‘More than 4000’

Device typeType of device which accessed the website.'Mobile Phone'
OriginOrigin domain if there is a referrer.‘www.etracker.com’
‘Google’ (or other search engine)


CampaignCampaign name which serves to differentiate between different campaign actions and is necessary in order to compare different campaigns. ‘Summer2014’
If you do not transfer any individual parameters, then the standard value here will be ‘organic’.
Purchaser typeCategorisation according to the ABC analysis.‘No purchase’
‘Just one purchase’
‘A’ (very important customer)

KeywordEntered search term or phrase in search engine marketing.‘Holiday home Istria’
CustomerCategorisation by customer and non-customer. ‘Yes’
‘No’

MediumType/Channel of a contact.'SEO', 'SEA', 'Social Media', 'E-Mail', 'Display', 'Affiliate'
Newsletter subscriberCategorisation by subscriber and non-subscriber to a newsletter.‘Yes’
‘No’

Project nameName you gave when creating the project.‘Test Istria’
‘SM Istria Promotion’

SegmentName of the segmentation rule according to which the smart message or test was sent out. Was given when creating the project.‘Order value rule’
Page nameName you gave to the page using the JavaScript variable et_pagename (from the tracking code) when creating the project. ‘HolidayhomeDetailsiteVillaX’
VariantName you gave the page variant when creating the project. With the variant, you can, for example, determine which version of a page increases the conversion rate most effectively (e.g. ‘A’ and ‘B’).
Repeat visitorCategorisation by Return visitors and First-time visitors.‘Yes’
‘No’

Time of the last orderCategory for time intervals between orders.‘No order’
‘More than 12 months’

Time since last visitCategory for time intervals between visits.‘Less than 1 day’
‘More than 30 days’

Key figures

The etracker Optimiser Reports consist of the following key figures:

NameDescription
VisitorsNumber of visitors.
PurchasersNumber of visitors who ordered something.
Bounce RateShare of visitors who leave compared to all visitors on the viewed entry pages.
ClicksNumber of all clicks. A click describes the contact of a visitor with the website. A contact here can be both a click on a link and a type-in.
Conversion buyer/visitorConversion rate: visitor to buyer. Share of visitors who ordered something
Konversion Besteller/KlicksConversion rate: clicks to buyer. Number of clicks which led to orders.
Conversion clicks/viewsConversion rate: Views to clicks. Number of clicks when smart messages were sent out.
PINumber of page views.
ViewsNumber of smart message sent out.

Selecting versions for displaying in the chart

Using etracker OptimiserTesting & TargetingProject overview & creationProject details each page can be shown in the graphic. To do so, simply click the corresponding checkbox. The original page will always be shown and is highlighted in black.

Previewing a page

Under etracker OptimiserTesting & TargetingProject overview & creationProject details click on the preview icon in front of the page name in order to view the preview of a page