Before you begin the installation please note that if you are planning to track sales and your “thank you”
page (which is displayed right after a successful sale) is located on a secure site, you would need to install Tracker on a secure site too (accessed with “https://” instead of “http://”). A secure site where Tracker should be installed can be different from the one where ordering pages are located.

If you are not planning to track sales or your “thank you” page is located on a non-secure site then you can ignore the above recommendation.

If you are upgrading an existing Tracker installation, please go to the Upgrade section for detailed instructions.

The zip

Please download the zip file with all Tracker files if you haven’t already done so. You can download the file directly to your server or to your local computer, whatever you prefer.

After you will have the zip file, you will need to unpack it preserving the directory structure inside of the archive (all modern archive tools do this by default). On Windows you can use a tool like WinZip to do this, on Linux or Unix based server you can use an unzip command if it is available:

unzip tracker.zip

If you’ve decided to unpack the archive on your local machine first, then you will also need to transfer all unpacked files and directories to the server. You can do this with any FTP client, for example CuteFTP on Windows. We advise to first create a separate directory on your site for the tracker and then upload all the files from the distribution archive to this newly created directory.

Setting permissions

After you will put all files and directories from the distribution archive in a directory on your server, you will need to set up correct file and directory permissions so that Tracker can function properly. You can do this with your FTP program or directly from a command line on your sever.

First you should rename “empty.conf.vars.php” file to “conf.vars.php”. Then make sure that directory “store” and file “conf.vars.php” is readable and writable by the server (on Unix servers this usually means setting access rights 0666 for the file and the directory).

Now you are ready to launch the installation wizard. Just open your browser and point it to the directory on your server where you’ve installed Tracker. For example, if you’ve uploaded the tracker indirectory “tracker”, use an URL similar to the one below

http://www.your-site.com/tracker/index.php

Please replace “www.your-site.com” with an actual URL of your site.

Or, if you are installing Tracker on a secure site, use an URL similar to the one below:

https://www.your-secure-site.com/tracker/index.php

Please replace “www.your-secure-site.com” with an actual URL of your secure site.

If you did everything correctly – you will see the first page of the installation wizard (shown below).

If instead of the wizard you see a server error page please rename the “.htaccess” file which is located in the root Tracker’s directory to “.htaccess.old” or just delete it and try to open the URL mentioned above again.

Step 1 – Choosing a language

Tracker supports a number of different languages. The language that you will choose on this step will be used on the subsequent steps in the installation wizard and in the program itself. You will be able to change this setting after the installation in the control panel.

Step 2 – License agreement

Please read the Tracker license carefully. If you agree to all terms then please check “I accept the license” checkbox and continue to the next step.

Step 3 – License key

If you’ve already purchased a license key for Tracker, then please enter it in the text box on this page. If you are only evaluating the product, then just choose the version you'd like to try and press “Request trial key” button. You will be able to use the product for 30 days without any limitations.

Step 4 – Settings

To complete this step please make sure that you have your database access details at hand (please consult your hosting company about these details if you are not sure what they are).

• Database host – the default value “localhost” should work for most of the cases. You only need to change it, if your database is located on a different server from the one where you install Tracker.
• Database port – the default entry “3306” is a standard port that MySQL uses. Unless you have a non-standard database setup, you don’t need to change this.
• Database name – a name of the database that Tracker should use.
• Database username – a name of the user that has access to the database which you’ve specified in the “Database name” field.
• Database password – a password of the user that has access to the database which you’ve specified in the “Database name” field.
• Tables prefix – this is a prefix that would be used for all database tables that Tracker would create in your database. This is a useful setting when you are using one database for several different products, or even for several different installations of Tracker. If you are unsure about this setting, just leave the default “ns” entry intact.

Step 5 – Registration information

On this step you are creating an administrator user account that you will use after the installation to login to the control panel. Please fill in all the fields and continue to the next step.

Step 6 – Company information

• Company name – the name of the company.
• Company description – a short optional company description, only for your information.
• Main domain of the site – the main domain of the site that you are going to track, for example “www.mytools.com”.

Step 7 – Confirm entered information

Step 8 – Installation

At this step, the installation wizard creates the required database structure and populates it with the data that you’ve provided. If no problems would be encountered, you will see a message that installation is complete. If the wizard would display an error instead, please try to resolve the problem yourself or contact Guys for assistance.

Please note that when this step is complete you can’t go to previous steps any more.

Remove the “install” directory

After you complete all steps of the installation, please delete the “install” directory located inside the directory where you’ve uploaded/unpacked Tracker on your server. Keeping this directory imposes a potential security risk and we advise to remove it as soon as you successfully complete the installation.

The Tracker does lots of work behind the scenes automatically. It prepares the requested reports, archives the reports, removes the old statistical data, renews the visitors geo data, browsers and operating systems data, processes the tracking requests that were logged when Delayed tracking was enabled.

We strongly advise to setup the cron script right after you finish the tracker's installation.

When a user requests a report in Tracker and decides to add it to the preparation queue – the report will be prepared as soon as the command line script will launch again on the server. That’s why it is preferred to launch the script with smaller intervals.

The cron script setting will require your skills of server configuration (or presence of a person who knows how to configure your server).

It is not enough just to check the “Enable the cron script” option in the settings. It is also necessary to add a record to the crontab of your server (if it is Linux, FreeBSD or any other Unix-based server) or setup Scheduler in Windows.

Setting up Scheduler on Windows might be challenging but without it Tracker will be rather restricted in its capacities.

It is necessary to setup the script intervals so that it is launched every minute. It does not mean the script will do something every minute, since it has built-in logic which decides if any new tasks have to be done. But if the script is launched every minute it has the best chance to do all the new tasks in time.

In the section “Administration / Cron Settings” you will find the line to be added to the crontab. However, the line may differ a bit for your particular server. In the very beginning of the line there is a 'php' command which tells the server that it is necessary to launch PHP interpreter. It is possible that your server needs to have '/usr/bin/php' or similar command instead. Please consult your system administrator if the default suggested line doesn't work for you 

On the same page you can see if the cron script is enabled or not. Under the main program menu you will find a message “The cron script was last launched at…”

The second important setting is “Maximum CPU Load”. Based on this setting Tracker will be taking care of not overloading the server by its operations. And if the load exceeds the specified value,  Tracker goes to a sleep mode for a while, until the load is below the critical level again. This works for all processes – Caching, Archiving, Old statistics removal, XML libraries processing.

The load limit you specify depends upon your server. We advise you to use the value equal to the total sum of processor cores of your server. That is, if you have Dual Xeon, for example – put “2” in the form field.

It is also possible to set the load limit on Windows servers. You should install the free of charge package LoadAverage that gives an opportunity to see average load parameters in Unix style on Win32 systems. 

In the corresponding field you should specify the complete path to GetLoadAvg.exe, so that Tracker could run the program and get the current load information from it. Path example: 'C:Program FilesArkane SystemsLoad AverageGetLoadAvg.exe'.

You can also specify the time of day when the cron script should better not be working. It will be launched but will not perform any tasks.

PHP CLI

For the command line script to work properly, it should be executed with PHP working in a CLI mode (command line interface). Depending on your version of PHP this might require using a special PHP executable (php-cli.exe, cli/php.exe, etc). Please refer to the documentation on the PHP site for the details related to the version of the PHP installed on your server (the location of the PHP CLI executable differs in different versions of PHP as well as on different platforms):

http://php.net/manual/en/features.commandline.php

Alternatively, you can ask your hosting company about the exact location of the PHP CLI executable on your server.

You can check if the php executable that you are currently using is a CLI version by running it with a “-v” parameter like this:

php –v

or like this:

/usr/bin/php –v

The output should look similar to this:

PHP 5.0.5 (cli) (built: Jan 5 2006 16:25:09)
Copyright (c) 1997-2004 The PHP Group
Zend Engine v2.0.5, Copyright (c) 1998-2004 Zend Technologies

If you will see “cli” in the brackets immediately after the PHP’s version then you are running a CLI executable – just what you need! If, however, you don’t see “cli” in the output then you are running a wrong PHP executable and you need to find its CLI version.

The main advantages of PHP CLI:

1. PHP doesn’t output any default HTTP headers (which do not make any sense when the script is run from the command line).
2. Several limitations from the php.ini file are ignored, like max_execution_time, for example. If you will notice that the command line script fails to finish the work and is killed by PHP because the maximum execution time configured in php.ini is exceeded – then you are NOT using PHP CLI and should correct this.

Logs

The cron script logs its operations.  There are three types of logs available to you:

1. Report preparation log
2. Statistics removal log
3. Statistics archiving log

In each of the three you will find useful information about Tracker operations in a background mode, how long they took and what the server load was.

Cron Processes Monitor

Starting from version 3.0.1 there is a special tool available to you that monitors background processes. The list of processes includes all the cron script runs no matter if it performed tasks or not.

There is a button “Complete” next to uncompleted processes. By pressing it you’ll force the process to finish. If the process is really running, indeed, it will stop as soon as it completes the current operation. If the process has already finished (probably with an error which caused the process to be still marked as uncompleted in the log) then within five minutes such process will be marked as complete automatically.

If you are planning to track sales with Tracker and your “thank you” page (that is displayed right after a successful sale) is located on a secure site (accessed with “https://” instead of “http://”) then you would need to make additional arrangements before you would be able to generate proper secure tracking code. Insecure tracking code located on a secure “thank you” page will make a warning about “insecure elements on the page” to appear in the visitor’s browser.

To generate secure tracking code you will need to make Tracker accessible from a secure site (it doesn’t matter if the secure site where Tracker can be accessed is the same as the one with the “thank you” page or it is different, the only requirement is that it should also be secure).

You can check if Tracker is accessible on a particular secure site by trying to open its control panel, using a secure link similar to the one below

https://www.your-secure-domain.com/tracker/index.php

where “www.your-secure-domain.com” is the secure site where Tracker can be accessed and “tracker” is the directory where Tracker is located on the site.

Secure site setups can be very different from server to server. If you are not sure how to make Tracker accessible on your secure site or if you don’t know whether you can use a secure site at all, consult your hosting company.

After you’ve made Tracker accessible on a secure site you should go to “Administration” page in the control panel and then click on “General settings” link located in “Program settings” section of the “Administration” page.

On the page that opens, please find a setting called “SSL-protected URL of the tracker”. In the input field to the right of this setting you should specify a secure URL of the directory where Tracker can be accessed on your secure site. For example:

https://www.your-secure-domain.com/tracker

To check if you’ve entered a correct URL, you can click on the “Open URL to see if it is correct” button that is located next to the input field. If you will see Tracker control panel in the new window (even the Login page) then the URL that you’ve specified is correct.

With this your SSL setup will be complete. You only need to do this once for all of the sites that you are going to track.

After you have at least one site added to Tracker, you should get the tracking code for it and put it inside the pages of the site. There are 2 ways to get the code.

If you are still on the page with the list of the sites, you can click on the “tracking code” link located next to each site’s name.

Or you can get to the same page by clicking on the “Settings” tab and then clicking on “Get tracking code” link. You will see a page similar to the one presented below.

There are 3 types of tracking code: to track visitors (their paths on the site), to track sales and to track actions.

Tracking visitors

First, you will need to get the code for tracking visitors. In order to do it, you will need to choose a site for which you want the code to be generated. Then you should select the code invocation type (we recommend using the default “Javascript” type because it can collect more data about the visitors than any other type). And finally select “Track visitors” and click on the “Generate” button.

After this you should see the tracking code inside the textarea. You should copy it and place it inside the pages of your site. If you’ve chosen a Javascript or HTML invocation types then you should place the code inside the BODY of the pages (between <BODY> and </BODY> tags). If you’ve opted to use PHP tracking code, you should place it inside of your PHP pages before any content was sent to the browser.

If possible, you  should place the tracking code on all pages of the site. If for some reason you can’t do it, you should place the code on all the landing pages (where visitors will land after clicking on your advertising links) and on the action pages (which, if opened by the visitor, trigger an action – see “Actions setup” for additional information about actions).

If you’ve configured more then one site in Tracker, you should generate and place the tracking code on all of them. If you are going to use Tracker as a general statistics package only you don’t need to setup anything else. However, if you want to also track ad campaigns, actions and sales then please continue reading the following sections of this guide.

Tracking sales

The second type of tracking code is used to track sales. Generally, the instructions to get the sales tracking code are the same: choose the site, choose the code invocation type, this time select “Track sales” and then click on the “Generate” button. After you will do this, you will see the code that should be placed inside the “thank you” page, which is displayed right after the order is successfully completed.

To get secure tracking code suited for placing on secure pages (for example, secure “thank you” page) please refer to the “Secure URL setup” section. If you’ve configured a secure URL for Tracker previously, then the tracking code that you’ve got is already secure. The provided sales tracking code will track an event of the sale, which is enough to calculate conversion ratio.

If you also want to track the total value of the sale (which is required for calculating ROI), particular products that were sold, id of the order in your own sales system or any other additional information then you will need to manually change the code using the instructions provided on the same page, under the textarea with the tracking code.

These instructions are for advanced users and you can safely ignore them if you are not sure how to use them. However, if you will decide to use them you should be prepared to specify dynamic  parameters in the tracking code, such as total value of the sale, product’s name, product‘s quantity, product’s value and id of the order in your sales system. You should consult developers or maintainers of your sales system for recommendations about these parameters: whether they are available to you on the “thank you” page and how you can use them with the tracking code.

All of these additional parameters are optional; you can use only one of them if you want, ignoring all others. Or you can use all of them if so you desire.

Tracking actions

The third type of tracking code is used to track actions. However, before you will be able to generate tracking code to catch actions you will need to create at least one action of a proper type. Please refer to the section below for a detailed description of actions. To get the action tracking code you should choose the site, choose the code invocation type, select “Track actions” and then click on the “Generate” button. If you have created at least one action of a proper type you will see it in a new drop down that will appear to the left of the “Generate” button.

If you need to generate tracking code for another action and it is also present in the new dropdown, then just select it and click on the “Generate” button again. After this you should see the tracking code for the selected action inside the textarea. You should copy it and place it inside the page of your site which if requested by a visitor should trigger an action.

The procedure of upgrading Tracker to a newer version is even easier then the initial installation.

But before starting with the update we recommend to backup your database and the conf.vars.php file (which is located in the main Tracker’s directory). In case if something will go wrong during the update process you will always be able to go back to the previous version of Tracker.

Update the files

First you should extract all the files from the distribution archive and upload them to the same directory where you have originally installed Tracker on your server. You should upload all files and directories except directory “install” (don’t upload it, it’s only required for initial installation).

Please note that as soon as you will upload the new files, Tracker will stop tracking automatically. When you will finish with the update, tracking will automatically start again. This is done to protect the data in your database from corruption because a new version of Tracker might require an updated database structure before it will be able to function properly.

Launching the upgrade wizard

Now you are ready to launch the upgrade wizard. Just open your browser and point it to the directory on your server where Tracker is installed. For example, if the tracker is located in a directory “tracker”, use an URL similar to the one below

http://www.your-site.com/tracker/index.php

Please replace “www.your-site.com” with an actual URL of your site.

Or, if Tracker is installed on a secure site, use an URL similar to the one below:

https://www.your-secure-site.com/tracker/index.php

Please replace “www.your-secure-site.com” with an actual URL of your secure site.

If you did everything correctly – you will see the first page of the upgrade wizard (similar to the one shown below). Please note that you should be logged in as an administrator in order to see the upgrade wizard and perform an update.

Updating process

To start the updating process just click on the “Update software to version” button and follow any additional instructions that a particular update might have.

If you have more then 1 update available for installation you will see an additional option “I'd like to run all updates in one step” that you can turn on to run all updates in one step, instead of going through them one after another.

If update to a particular version requires significant database structure changes you might be asked to perform an update via a command line script. In this case, please note, that you need to use a CLI version of PHP to run the script, otherwise the update process might not finish due to PHP killing the updating script when maximum execution time specified in php.ini will be reached.

You can find more information about PHP CLI in a corresponding section of this manual.

When installation of the updates will be finished you will see a message confirming this. You can go to the Tracker’s control panel after this, the upgrade process is complete.

In order to start measuring your ad campaigns effectiveness you need to setup the campaigns in Tracker and generate ad links for them.

You can manage your campaigns in the "Campaigns" section. Here you can create, edit, group campaigns and specify their costs — as a total or per click.

Once a campaign is created, you can start generating links for it. In the ad links you can specify keywords that should be tracked when visitors will be clicking on the links; you can also generate redirecting ad links that will track a click and redirect the visitor to the destination landing page (when redirection is not used, Tracker will expect that a landing page has a standard visitor tracking code inside); and you can also combine an ad link and a split-test link in one link which can be used both to track a click and to send the visitor through a split test.

After you will put the generated links in your ads Tracker will begin gathering campaign statistics. For a thorough statistics analysis you can use such features of Tracker as the reports constructor, log reports and the visitor paths report.

In order to begin gathering natural traffic statistics you only need to put the visitors tracking code on all the pages of your site (or at least on the pages which you want to track). Natural traffic statistics will be available in the reports constructor, log reports and the visitor paths report. 

To track sales, you need to put special sales tracking code on the final "thank you" page of the ordering process (the page which is displayed to the visitor after the purchase was completed successfully). After that, whenever this page is displayed a sale will be tracked automatically by Tracker. You can also specify additional parameters in the tracking code — they would be saved with every tracked sale.

Add sales manually

It is possible to add sales manually (Settings → Add a sale) and edit already tracked sales (All reports → Sales log → Edit a sale). If you will specify an ad campaign, a keyword or a split-test for a sale which you add or edit, then the sale will be attributed to that campaign/keyword/split-test and this will update corresponding numbers in the paid traffic statistics.

You don't have to specify the ID of the visitor for a sale, in that case Tracker will create a new visitor for you and will log a sale for that visitor automatically. If you will specify the visitor ID and a particular campaign for the sale then Tracker will check if the specified visitor had a click tracked from the campaign that you've specified, if it won't find any, then a click will be added automatically to the visitor path — this will change the numbers accordingly in the paid traffic statistics.

There are four ways of tracking actions and action targets on the site.

Actions could be sometimes referred to as "events" in other statistical software. Tracker uses the term "actions" everywhere in the interface, so if you are looking for events — actions is what you need. 

Also, in Tracker a sale is not an event or an action — it's tracked separately using special tracking code. You can read more about sales tracking in a separate section of the manual.

1. Tracking an action by using a special tracking code.

You can just place a special code on the page, and an action will be tracked whenever this page is opened.

2. An action can be associated with an URL on the site. Such action is tracked using a standard visitors tracking code.

When creating a new action, you should indicate what URL will be associated with it: it can be just a page URL (e.g. http://www.yoursite.com/actionpage.html), or a certain URL pattern (e.g.  http://www.yoursite.com/*/actionpage.html).

The place where Tracker takes the action target from is also specified in the URL. For example, you can specify an URL http://www.yoursite.com/search.php?query={a}, and then the actual 'query' parameter value will be recorded as an action target.

3. Tracking action via redirect.

When creating this type of action, you should specify the site and the page to be associated with the action. This is necessary for Tracker, because any event on the site (whether it is an action, a sale, an ad click or a split test entry point) is always associated with a particular page URL. In the visitors paths and in the actions log reports you will be seeing the specified page as if the action has actually taken place on it.

You can specify the target URL where Tracker should redirect the visitor to after logging an action. Redirect implies that you are giving to the visitor a special tracking link which leads to Tracker, and not to your site.  After the visitor will open the tracking URL, Tracker will redirect him/her to the target URL you have indicated.

(As soon as you fill in the form and save the action, Tracker will inform you what link should be given to visitors). In order to avoid confusion, let us once again define the notions:

1. There is an URL you give to visitors (for example, in order to register file downloading). This URL leads to Tracker, and we will call it Action URL.
2. And there is a target URL where Tracker should redirect the visitor (for example, a particular file address). Let us call it Redirection URL.

You may use dynamic substitution of redirection URLs. Just add the parameter &rurl=redirection_url to the given action URL. If "&rurl" parameter is present in the tracking URL then Tracker will send the visitor to the URL specified in this parameter.

In the example with file downloading you could make the following links on your site: http://tracker.yoursite.com/track/event.html?eid=1&rurl=http://yoursite.com/file.zip . Tracker will register a file downloading action and redirect the visitor to http://yoursite.com/file.zip.

In addition, with such a scheme you can track action targets. Add the parameter &itm=  to the action URL, and all the value of this parameter will be recorded as an action target.

For example: http://tracker/track/event.html?eid=1&itm=Very_important_file&rurl=http://yoursite/file.zip

After creating a split test you can add to it as many pages for testing as you want.

For the already created split-test Tracker will give you a target URL, which you can use in your ad or any other place that should be linking to a split-test.

You can choose one of the two ways of choosing pages for redirection. By clicking the test link, the visitor should get to one of the pages indicated in the test — chosen either randomly or sequentially.

Moreover, you choose if one visitor always gets to one and the same page or it is going to be a different page every time.

All operations the visitor makes on the site after going through the split-test will be associated with one of the test pages, and you will be able to assess the efficiency of each page via Reports Constructor or on a separate page of the split-test statistics.

In addition, you can combine the link to the split test and the link to an ad campaign, and indicate the keywords for this link if necessary. This will be important for ad campaign statistics.

When defining a new funnel, you will find a lot of options available to you.

Every visitor step in the funnel can belong to one of the following types:

1. An ad click
   You can choose a particular ad campaign with a particular keyword or without a keyword, or just a keyword without an ad campaign, or any campaign with any keyword.
2. A split-test
   It can be any split-test, or a particular split-test.
3. A referrer  
   You can indicate just a search engine, or a group of domains (added manually), or a particular host.
   You can also state a keyword. All these fields can stay blank, and then it is going to be any referrer with any keyword.  
4. An entry point
   Any entry point to the site. You can indicate that this should be a direct entry (without an ad campaign, split test or referrer).
5. An event on the site
   It can be a particular event with or without a particular action target, or any event with any action target.
6. A sale
   Any sale or sale of a particular product.
7. A page load

Besides, the first step can be narrowed and be only for the visitors who enter the site for the first time or it is their first step in the session.

Every subsequent step can immediately follow the previous one or appear in the visitor’s path a few steps later.

The funnel statistics will show you not only conversion of visitors from step to step, but also the pages where their path ends.

The report can be time-limited and particular site oriented.

The main principle behind the Reports Constructor is that in any report you can choose one of the result rows and "go deeper”, leaving the current choice as a limit for the next "deeper" report. For example, you choose report “by sites”, then click on one of the sites, choose report “by search engines” and see what search engines have brought visitors to this particular site. You can go deeper until the list of reports available is over.

The Reports Constructor is divided into two parts: “Paid traffic” and “All traffic”. All numbers which you will see in the paid traffic section is a result of visitors arriving from the ad campaigns..  

You can save any report that you've created in the Reports Constructor. After saving it will be available in the section “All reports”. And the most important is that every saved report automatically becomes a part of the Caching and Archiving functionalities of the tracker.

The Reports Constructor filter allows you to limit the reports by a certain date or time period. A number of built-in time periods is always available to you and you can always create your own time periods.

Column sums

In many reports you will find summarized data at the bottom.

1. Column sums + Minimum, Average and Maximum values.

These data is only displayed when the number of the rows in the report does not exceed the limit set by you in the filter (by default it is 20). Otherwise it will be impossible to calculate the sum of columns.

The sum is based on the displayed rows in the report. For all the columns where the number of unique visitors is displayed, the sum value can potentially be not equal to the actual sum of the unique visitors. For example, if you are viewing a “site report”, and a visitor has visited not only one but several of the sites, he/she will be counted in several report rows. The number of unique visitors will be correct for each particular row, while the sum of the values in the rows will be incorrect because one and the same visitor will be counted several times.

2. Total.

The very bottom line reflects the real sum that is calculated in the database. This sum counts one and the same visitor only once, therefore the values can be less than in column sums (but they are correct).

Since the calculation of this value requires a separate query to the database and this query will take as long (or even longer) as the full report, we decided to make it optional.

However, in the cached reports this data is available immediately.

Columns

Only a few report columns need to be explained.

1. Hits (same as "Page views”)

It is the number of page views done by all the visitors that satisfy the conditions of the current report and particular row in the report. Note that it is not the number of different pages, but the total number of page views.

2. Unique visitors (in the columns of sales and actions)

These columns reflect the number of unique visitors who have made actions or sales, and not the number of unique actions or unique sales.

3. Conversions

These columns display the relation of unique visitors who have made actions or sales to the total number of unique visitors.

Delayed Tracking requires constant functioning of the cron script.

You can enable Delayed Tracking for the whole system (“Administration / General settings”), and for each particular company ("Administration / Client Management (or the main company edition)"). All tracking requests to Tracker would be recorded into a temporary table and processed later on.

Please note:

1. With Delayed Tracking enabled the data will get to the statistics with a little delay.
2. There might be so many delayed records that Tracker may not have enough time to process them in a background mode, and the records may be accumulated.

Caching ("Administration / General caching settings") is the main tool of keeping Tracker in the operational condition in case of big data volumes. The more information is in the database, the longer it takes the database to perform queries and aggregate the data.

The caching process will prepare reports in a background mode, therefore it requires constant functioning of the cron script.

Caching prepares several types of reports:

1. Saved reports. You can indicate specific caching settings for each report. (Please remember that you can create your own time period.)
   
   
   
   
   
   
2. Ordinary statistics of campaigns and split tests. This statistics is displayed right in the lists of objects. Such statistics is prepared every hour by default.
3. Popular reports. Using the Reports Constructor you can open some reports more often than the others without saving them. In caching settings you can specify how often the report should be opened in order for it to be added to the cache.

"Smart caching" will try to determine how long it takes to prepare this or that report and will decide whether it is necessary to prepare the report in a background mode or it may be prepared and displayed in real time.

Report Preparing Methods

If you have Caching enabled then you might see a message that a certain report is not ready yet when you will try to open it. In this case you will also be presented with two options:

1. Add the report to the preparation queue (requires constant functioning of the cron script).
   This method is preferable as the work will be done in a background mode and the current server load will be taken into account. Also, this method doesn't depend on PHP script execution timeout settings and PHP won't terminate report preparation even if it will take a significant amount of time.
2. Prepare the report immediately.
   Choosing this variant, you will launch report preparation in real time as an http process. Your server's PHP timeout settings might prevent the report from finishing preparing, and the database load can become too big.

Time Zones

In the settings of a company (or companies) you can indicate the main time zone of the company (companies). This is necessary to prepare reports.

If you put a report in the preparation queue, the background process will be using the time zone of the company. The same time zone is used in preparation of the main system reports (for the list of campaigns, split tests and some of the Dashboard reports).

In all other cases (including the case when Caching is disabled) the time zone you’ve indicated in the Personal Settings is used.

Starting from version 3.0 the tracker works only with the binary geo library from MaxMind (http://www.maxmind.com/app/ip-location). You need to put one of these files in the 'store' folder in the tracker's directory: GeoIP.dat (only countries) or GeoCity.dat (countries and cities). Please note that you should keep the exact name of the file.

If you are adding the geo library to the tracker for the first time or updating it with a new version, you also need to click on the 'Update database' button after the new files have been added.

You can also use a free version of the MaxMind's cities geo library called GeoLite (http://www.maxmind.com/app/geolitecity). It contains almost the same information as the commercial version, but the data is less accurate.

Reports archiving will help you keep the statistics history if you are going to clean the database for better system performance.  

On the page of Archiving settings (“Administration / Archiving Settings”) you can specify what reports and for what time period are to be archived.

For example, if you specify that you want to archive a "Year" report, at the beginning of every year  Tracker will archive the report for the past year. A “Month” time period means that Tracker will archive the report at the beginning of every month — in February for January, in March for February, etc.

On the first launch Tracker will start archiving all the old time periods starting from the first day when it started to gather the statistics.

Please take into account that the process of database cleaning will not affect the information that is due to be archived.

For example, if you have specified that you want to archive a "Year" report, the process of data removal will never affect the current year. Even if you will specifically indicate the dates for which the statistics should be removed. (This logic works only if Archiving is enabled).

The archiving process takes place in a background mode; therefore it requires constant functioning of the cron script.  

You don’t have to enable/disable the Archiving function many times. It is enough to set it up once, and Tracker will automatically monitor reports and archive them as and when necessary.

The system of automatic old statistics removal lets you keep the database size, avoiding its uncontrolled growth.  

The process of old data removal takes place in a background mode and therefore requires constant functioning of the cron script. No matter how often the cron script is launched, the process of old data removal will be taking place in the intervals you specify.

For example, you can specify the following interval: to remove old statistics data every month, on the first day of the month, at night, using your time zone settings.

You can remove both complete visitors’ paths and parts of paths that are non-significant for statistics. Besides, you can remove the data that is not used in the statistics (referrers, keywords, visitors, site pages, products, action targets). Non-used data appears when the visitors’ paths have been removed but the additional information remains in other tables.

Every time after old statistics removal the database will still have non-used records left. As the program is meant to work with MySQL version >= 4.0, we can not use methods of automatic removal of non-used records that are available in MySQL starting with version 5.0 and any other “normal” databases.
However, the fact that they are not automatically removed is even an advantage. If you choose to keep non-used referrers, for instance, then the tracker will not have to make a new record in the database when a new visitor comes from such a referrer, and this means faster tracking.

You can specify the time period that should not be removed during old data removal – for example, not to affect the data within a year (starting from the current date backwards) or the last three months. Besides, you can specify a particular time period (from date to date) of statistics to be removed. You can specify separate settings for each particular client company.

Database cleaning will not affect the data to be archived, providing that Archiving function is enabled.

In case you want to launch automatic old statistics removal but doubt if it will affect the data that should be archived -- do not doubt. It will not. On the contrary, the database clearing might fail to remove anything if Archiving function is on but nothing has been archived yet.  

Manual Statistics Removal

Besides the automatic old statistics data removal there is also a manual method. You will find the manual removal form in the section “Administration / Database”. The form only appears in case “Automatic mode” is disabled.  We strongly recommend you to use the automatic mode. There is no big difference between the two modes of removal as the used settings and the code are the same. It is only that the manual mode launches the process in real time via a special HTTP request which means that the removal process will obey your server's PHP script execution timeout settings. If the removal process will take too long it might get terminated by PHP prematurely which may damage the database integrity. Besides, the cron script watches the server load and goes to the sleeping mode if the load exceeds the critical level. The real time removal process does not do this.

That is why we deliberately made the manual removal form not so easy to access.

The only reason why we have made this form at all is because you might not have an ability to setup the cron script on your server.

If you do not have any special reason to launch the manual removal, please use the auto removal mode.  It requires constant functioning of cron script, but on the other hand the PHP timeout settings have no effect on the script running in a background mode and you can be sure the cron script will finish its work.  

If you need to remove statistics as a one-shot job, in the settings you should specify that the removal process should be launched once a month (if it is not the first night of the month, the automatic removal will be waiting for the next appropriate period) and press the button “Run removal process as soon as possible”. Afterwards you can reload the page from time to time to check if the button is active again, or you can use the cron processes monitoring. As soon as the removal process is complete you can disable the automatic removal mode.

Tracker offers you several tracking code variants (Javascript, PHP, ActionScript) which you can easily obtain using a convenient form of tracking code generation.

It is important to know which additional parameters you may use in the code.

Open the table of parameters →

More details on how to add parameters you will find on the page of code generation.

Depending on your preferences you can configure the tracker to use different traffic priorities (“Administration / General Settings")

What does it mean? Let us study a few cases:

1. Equal priority + Last Entry Point.
   Sales and actions will be attributed to the very last entry point in the visitor path just before the sale or action happened (entry point could be an ad click, arrival from a referrer or a split-test).
   That is, if the visitor had entry points in his/her path before an action or a sale was made, then the action or sale will be attributed to the very last entry point.
   If afterwards the visitor gets to the site again (through an ad click, from a referrer or through a split-test), the new actions or sales from him/her will be attributed to the new entry points. The previous actions and sales will stay unchanged.
2. Equal Priority + First Entry Point.
   Regardless of how many times and where from the visitor gets to the site, all his/her sales and actions will be attributed to the very first entry point.
   

If you specify the traffic priority, the situation changes a bit. Just like with equal priority, the tracker will be checking first and last entry points and attributing actions to them, but:

1. If the visitor has an entry point of the traffic type for which the priority is specified, the sales and actions will always be attributed to it. In the same settings form you specify whether it is the first or the last entry point.
2. If the visitor has no entry point of the traffic type for which the priority is specified but has entry points of other types, the actions will be attributed to them. In the settings form you specify whether it should be the first or the last entry point.
   

Split-Tests

One more important note: split-tests always have the highest priority. Regardless if you give priority to paid or natural traffic, split-tests will always have an advantage. However, if the entry point coincides with the split-test (for example, if ad click leads to split-test), they are considered as one entry point and then actions will be attributed both to the split-test and the ad campaign.

Statistics

Statistics that you see in reports is based exactly on this attribution logic. For example, in the sales column of ad campaigns report you will see exactly those sales that during tracking have been attributed to entry points from these ad campaigns.

Change of the traffic priority settings does not affect the already gathered statistics.

If you do not see the expected figures of sales and actions in the statistics, you should check if the traffic priority settings are configured as you want and have a look at the sales logs. If the sales are successfully tracked and displayed in the sales log, but you can't see them in the paid traffic aggregated reports then they most likely have been attributed to different entry points from what you expect.

Site Profiles

Tracker doesn't have a limit for the number of tracked domains inside one site profile. So in theory inside one site profile you can aggregate statistics for any number of domains and sub-domains.

However, this can prove to be inconvenient if you need to view the aggregated statistics of different sites and projects separately. For this you can use site profiles, which effectively are just containers for the aggregated data.

Site Domains

After a certain period of gathering statistics you may find out that the list of site domains contains some records of unknown origin or unwelcome nature.

These may be IP addresses of search engines, or your local domains that you use for site creation, or domains of sales systems on whose 'thank you' pages you’ve put a sales tracking code. And so on.

Therefore we advise you to enable the option "Gather statistics only for the selected domains”. Every new domain which is not specified in the list of allowed domains will be ignored. You can make up your own list of specified domains on the site editing page.

Please note that If a visitor arrives from a referrer whose domain is registered as a domain of the current site, the tracker will not consider it to be entry from a referrer.

Removal of unwanted domains can be done on the company editing page, in the list of sites.

Please pay special attention to sales system domains (i.e. Yahoo store, ShareIt, etc). For the sales to be tracked their domains should be included in the list of allowed domains.

Cookie Domain

In the site settings you can specify on what domain the tracker should set the tracking cookies. It is important in case you use different tracking code types on different site pages.

For instance, you might be using Javascript tracking on www.yoursite.com, PHP tracking on clients.yoursite.com, and Tracker itself is installed on tracker.yoursite.com.

By default the tracker will set cookies without specifying a domain, and they could only be read back on tracker.yoursite.com in case of Javascript , HTML and ActionScript tracking types. In case of PHP tracking, the current domain will be clients.yoursite.com, and the cookies could only be read back there.

Thus one and the same visitor may be identified as two different visitors in case he/she loaded pages with different tracking code types on different sub-domains.

In order to avoid that in the above scenario you should place '.yoursite.com' (without quotes, with a point at the beginning) in the corresponding cookie domain field. Then the tracker will set cookies that could be read back on all the sub-domains of yoursite.com site.

Site Text ID

This option is added for your convenience and is especially useful when you integrate Tracker with various content management systems. In the tracking code you can use a text ID of the site instead of numeric one (if the text id has been specified in the site settings).

Instead of the variable nsSiteId in the tracking code you should put the variable nsSiteTextId, and the tracker will identify the site using the specified text ID.

More details on parameter settings can be found in the tracking code article.

Start creating your own skin with these simple actions:

1. Create a new folder inside skins. (i.e. skins/myskin)
2. Inside the new folder create the following file: myskin/config/conf.skin.php
3. Fill the file with the following content:
   <?
   $SkinConfig['Parent'] = 'default';
   ?>
   

Now you have a new version of the interface (which is fully identical to the default one).

You may take any file from the default skin, copy it to the new folder (keeping the inner path) and change it.

All the files missing in your own modification will be taken by the tracker from the skin specified as parent (parent skin can also be a modification!).

This approach is much better than changing the default skin. Every new version of the tracker might have CSS styles and parts of the templates changed in the default skin, so you risk losing your changes (in case you did modify the default skin instead of creating your own) after installing each product's update.

Please bear in mind that without a White Label license you are not authorized to remove the product’s logo and it should always remain visible in the tracker's interface.