Advertising Campaigns and ROI

MBI allows you to easily marry advertising cost data and revenue data from your database. This helps you identify which campaigns have the highest ROI. This article explores a few different methods of evaluating your campaign performance.

Prerequisites

User acquisition campaigns

Campaigns targeted at user acquisition can be measured from many perspectives, including:

  1. The number of new users acquired of campaigns
  2. The conversion rate from registration to purchase of campaigns
  3. The ROI of campaigns based on average user lifetime value (LTV)

Analyses (1) and (2) above are explored in a separate tutorial on identifying your top marketing channels. Here, you explore analysis (3) to measure campaign ROI over time. This answers whether users acquired from a particular campaign generated enough lifetime revenue to cover the cost of acquisition.

NOTE

This example assumes that all campaign costs were exclusively used to acquire new users. In reality, your campaign cost is also shared with acquiring unconverted visits, repeat purchasers and such. By assuming that all cost is used to acquire new registered users, the resulting ROI accounts for the worst case scenario (highest cost per acquisition). You can be sure that your actual ROI is higher than your calculation.

Example: Assuming that you spent $20 on a campaign that generated 10 new users and 10 repeat buyers, your actual cost per new user is $1. But, with the assumption that all cost went to acquire new users, the cost per acquisition is $2.)

1. Start by creating a chart that segments your Ad Cost by Campaigns:

  1. Create a Metric that sums your spend over time

  2. Go to Data > Metrics

  3. Select Add New Metric and select the Adwords... table that is recording your AdWords cost data.

  4. In the metric editor, give your metric a name (for example, AdWord Cost)

  5. Using the dropdowns, perform a Sum on the adCost column in the Adwords… table (Change) ordered by the date column.

  6. Click Back to Metric List at the top and go to any dashboard.

  7. Create a report that segments spend by campaigns

  8. In any dashboard, click Add Report > Create report

  9. Select the Adword Cost metric that you just created

  10. Set the Time period to All-time, and Interval to None

  11. Under the Group by tab, add campaign as grouping field, and click Add All in the box.

  12. This report shows your all-time AdWords cost by campaigns

2. Create a report that counts new users by campaigns:

  1. In any dashboard, click Add Report > Create report
  2. Select the New users metric that counts the number of new registered users over time
  3. Set the Time period to All-time, and Interval to None
  4. Under the Group by tab, add campaign as grouping field, and click Add All in the box
  5. This report shows your all-time registered users by campaigns

3. Create a report that segments average user LTV by campaigns:

  1. In any dashboard, click Add Report > Create report
  2. Select the Average lifetime revenue metric that calculates an average user’s lifetime revenue
  3. Set the Time period to All-time, and Interval to None
  4. Under the Group by tab, add campaign or utm\_campaign as grouping field, and click Add All in the box
  5. This report shows your average user lifetime revenue by campaigns

Finally, calculate campaign ROI by bringing together these three analyses in one report:

  1. In any dashboard, click Add Report > Create new report

  2. Add as input, use the three metrics used above. Each is assigned a letter (for example,[A], [B], and [C])

  3. Cost: Add the metric AdWords cost - this is variable [A]. This returns cost by campaigns.

  4. Users: Add the metric New Users - this is variable [B]. This returns the number of users by campaigns.

  5. LTV: Add the metric Average Lifetime Revenue - this is variable [C]. This returns LTV by campaigns.

  6. Click the hide icon beside the word Chart so you can focus on the table

  7. Now use Add Formula to combine these metrics, as follows:

  8. ROI: Enter the formula (\[C\]-\[A\]/\[B\])/(\[A\]/\[B\]), if [A] represents Ad Cost by Campaigns, [B] represents New users by campaigns, and [C] LTV by campaigns. This returns the ratio of (average user LTV - average cost per acquisition) / (average cost per acquisition)

  9. Avg Return per User: Enter the formula [C]-([A]/[B]). This return the average margin made on a user by calculating (average user LTV) - (average cost per acquisition).

  10. CPA: Enter the formula \[A\]/\[B\]. This returns the actual campaign’s cost per acquisition.

  11. Other potential metrics to include from AdWords data include sums of Impressions and adClicks (from AdWords data), along with the total number of orders made via a particular campaign.

  12. It may also be interesting to calculate the ROI based on LTV 30 days and 90 days after a user registers or makes a first purchase.

  13. Feel free to click and drag your metrics and formulas to reorder the columns of your report

  14. Name your report and be sure to save as a table.

Product campaigns

Are you running product-specific advertisements? If so, you can measure ROI on those campaigns by calculating revenue / cost for specific product(s).

NOTE

This example assumes that all campaign costs were exclusively used to generate purchases of specific product(s). By assuming that all cost was spent on generating purchases, the resulting ROI accounts for the worst case scenario (highest cost per purchase). You can be sure that your actual ROI is higher than this calculation. Example: Assuming that you spent $20 on a campaign that generated 10 new users and 10 purchases, your actual cost per purchase is $1. Under the assumption that all cost went to acquire new users, the cost per purchase is $2.)*

Before you start, submit a support ticket to join the following dimensions to your line items table (sales\_flat\_order\_item, order\_item):

  • Order’s source (if you only track referral source at the user level, then join user’s source)
  • Order’s campaign (if you only track referral source at the user level, then join user’s campaign)
  • Order’s medium (if you only track referral source at the user level, then join user’s medium)

1. Now start by creating a chart that returns revenue per campaign for specific product(s):

  1. In any dashboard, click Add Report > Create new report
  2. Select the Revenue by items metric that calculates revenue at the line items level
  3. Set the Time period to All-time, and Interval to None
  4. Under the Filter by tab, add product name 'IN' Product A, Product B, Product C, …" and include all product names targeted by your campaign separated by a comma (for example, product name 'IN' yellow t-shirt, red t-shirt, blue t-shirt)
  5. Under the Group by tab, add order's campaign or order's utm\_campaign as grouping field, and click Add All in the box
  6. This report shows you the revenue for specific product(s) by campaigns

2. To calculate ROI, you again combine metrics in one report:

  1. In any dashboard, click Add Report > Create new report

  2. Add the Revenue by items metric, following the filter and group by directions from the campaign for specific product(s) report above and click Hide beneath the metric’s scalar value

  3. Now add the AdWords Cost metric, following the filter and group by directions from the Ad cost by campaigns report you explored in the User acquisition campaigns section above; then click Hide beneath the metric’s scalar value

  4. With these metrics in place, add formulas:

  5. ROI: Enter the formula \[A\]/\[B\], if \[A\] represents Revenue per campaign for specific product(s) and \[B\] represents Ad cost by campaigns. This returns the ratio of (Revenue for specific product(s)) / (Campaign Cost)

  6. Return: Enter the formula \[A\]-\[B\]. This returns the average margin made on a user by calculating (average user LTV) - (average cost per acquisition)

  7. (Optional) Revenue: Unhide the Revenue by items metric to see revenue for specific product(s) per campaigns

  8. (Optional) Cost: Unhide the AdWords Cost metric to see the cost for campaigns

  9. Give your report a name and be sure to save it as a table

3. Repeat steps 1 and 2 above for each of your advertised product(s) or product group(s).

On this page