Comparing Conversion by Acquisition Source

Overview

PetBox is a fictional eCommerce company that sells monthly subscription boxes containing animal care products. Currently, they have four marketing channels through which users are acquired:

  • Paid - Users acquired through paid search.
  • Email - Users acquired through email campaigns.
  • Content - Users acquired through campaigns on Facebook and Instagram.
  • Direct - Users arriving directly to the site by searching or typing the address.
  • Referral - Users acquired through a link from an affiliate partner.

PetBox wants to evaluate the conversion percentage of users acquired through each of these channels. This analysis will determine which channels will receive a greater investment of marketing resources, and which channels are underperforming.

Question

What is the most valuable marketing channel by conversion percentage? 

Open the Funnel Tool

Select Create New Analysis, and then find Conversion and Drop Off in the dropdown menu. Selecting Conversion and Drop Off will set up a pre-configured query to identify points of friction between events.

comparing_1.png

Build the Query

The three-step funnel will include the events Site Visit, Blog View, and Subscribe.

comparing_2.gif

Next, let’s collapse the query builder so we can focus on the funnel.

comparing_3.gif

This is a three-step funnel showing the conversion from Site Visit to Blog View to Subscribe over the last 7 days. The percentages in the webbing between steps indicate the percentage of users that have moved from one step to the next -- keep in mind that your data may vary from this example.

The percentages in the webbings should be interpreted as:

  • 22.78% of users who visited the site went on to view the blog.
  • 45.05% of users who viewed the blog went on to subscribe

The total conversion rate of 10.26% is displayed above the funnel. Percentages will change if the date range is adjusted. For example, conversion percentages across 30 days will be different than those across 7 days. 

Break out by Marketing Channel

To break out our results, let’s add a User or Event Property. In this case, we’ll use the event property Marketing Channel. Breaking out our funnel by this property will group users based on the marketing channel they were acquired through. To add these breakouts, re-expand the query builder and select Group By underneath the first funnel step, then choose Marketing Channel under the event properties tab. 

comparing_4.gif

 Select from First/Last/All

We’re interested in understanding the marketing channels these users were acquired from.

comparing_5.gif

  • Select First next to the event property Marketing Channel in the query builder. First will only display the channel that first led users to the homepage within the specified date range and will exclude marketing channels users were re-acquired through later on.

Run the Query

comparing_6.png

comparing_7.png

The table below the chart is broken out by events along the X axis, and by marketing channel along the Y axis.

For example, the first row tells us that of users who were acquired through Paid search:

  • 43,484 completed Site Visit.
  • 10,309 completed Blog View, a 23.71% conversion from Site Visit.
  • 5,551 completed Subscribe, a 53.85% conversion from Blog View.
  • The total conversion rate across all steps was 12.77%

Compare Conversion Rates

Let’s click the row corresponding to the largest count in user signup, which is Email.

comparing_8.gif

Clicking a row in the table expands the selection. All conversion rates within the webbings and top line update to reflect the selected campaign source.

The total conversion rate is 10.26%, while the conversion rate for users whose acquisition marketing channel was Email is 8.34%. Now, let’s compare that to the row with the highest (non-direct) conversion rate, which is Paid (12.77% conversion).

Observation

While email is the largest source of users, users from paid search convert at a higher rate. PetBox should explore investing additional resources into paid search campaigns.

1 users found this helpful