Freshmarketer now allows you to pause any particular variation in an experiment. This in turn will stop that variation from receiving new visitors.

This option is helpful

• To redirect all new traffic to other existing variations
• To study how other variations perform
• To stop allocating traffic to the low performing variation

Example:

Let us assume that in an experiment containing 4 different variations, one particular variation performs poorly compared to other variations.To stop sending visitor traffic to the low performing variation, Freshmarketer provides you with an option to pause that particular variation. Now, the traffic is divided and reallocated to other active variations.

Understanding “Pause Variation”

Once a visitor is shown a variation, they will continue to experience the same variation for the entire experiment. So, when a variation is paused and a visitor converts (goals), the conversion will be taken into account for that variation even though it is paused.

Example:

Let us assume visitor A is shown the variation A. After few days, the variation A is paused. Now, when visitor A visits the site after five days, since the variation is paused, the same variation A will be shown (provided the cookies are not deleted). If the conversion goal has been set as Track engagement and if Visitor A converts, the conversion will still be accounted for the paused Variation A. Thus, the reports will show that a conversion happened on a paused variation despite not receiving any new traffic.

In the end, pausing a variation will only prevent the variation from receiving new traffic and it will reallocate its share of traffic among other variations. The reallocation calculation has been explained in later part of this guide.

Locate the “Pause” option

This guide will help you in locating the Pause option present in Freshmarketer.

In the “Preflight” or “Summary” tab inside Freshmarketer

• After pausing the experiment, in Preflight tab, under Set Traffic section, pause a variation by clicking the Pause option to the right of all the individual variation. The paused variation can be made active again by clicking Unpause option right to the paused variation.

Before pausing a variation:

After pausing a variation:

• Click Save Changes to save the reallocated and updated traffic engagement. Click Undo to revert back the changes.

• In Variations section, the traffic percentage and status of all variations along with original is displayed. If a variation is paused, then it is shown as Paused right to that particular variation.

Note that the traffic percentage of the paused variation becomes 0 since there will be no new traffic. The paused variation’s traffic will be split among other variations and gets updated in the fields.

Calculation of reallocating paused variation's traffic to other variations

When a variation is paused, the traffic allocated for that variation will be divided and reallocated to other variations in the experiment.

Example:

Let’s consider an experiment with an original and 2 variations. The traffic allocation for them before and after pausing variation are as follows.

When the Variation2 is paused, the traffic percentage allocated for the same, 25%, is divided and allocated to Original and Variation1.

The traffic allocation calculation is explained as follows.

Step 1:

In the given example,

Step 2:

Traffic allocated to paused variation (T1) = 25

[From RESULT 1in Step 1]
Total (T2) = 12.5 + 6.25 = 18.75

Now, Total = T2 - T1 = 25 - 18.75 = 6.25

In the given example,

Step 3:

In the given example,

The remaining 1% traffic is allocated to either of them, randomly.