SoundCommerce Marketing unifies first-party data to give you an accurate view on your return on ad spend (ROAS), customer acquisition costs (CAC), customer retention costs (CRC), and more. This can help you determine if your marketing campaigns are acquiring new customers, generating leads, and contributing to revenue.
This article explains how SoundCommerce Campaign unifies cross-functional data from digital marketing, web analytics, and e-commerce platforms.
In this article
- Requirements
- How we model the data
- Attributed ad spend
- What about unattributed ad spend?
- Unattributed revenue
- Clean data in / clean data out
- Marketing overrides
- Module ERD
Requirements
SoundCommerce Campaign Module
How we model the data
To get a comprehensive view of total orders, ad spend, and attribution, we join three different types of data, described as follows.
We understand that most businesses have numerous media and order data sources. To keep things simple here, we refer to these data sources by their functional category.
Digital marketing data
Your media data tells us what you spent on advertising and messaging platforms. Examples of these include digital marketing (Google Ads, Bing, Criteo), social media (FaceBook, Pinterest), email marketing (Klaviyo, Listrak), and SMS text messaging (Attentive).
Reports from media channels regarding attributed order counts and order items are often biased and unreliable. Ad platforms frequently calculate conversions within a specific time window (for instance, 30 days), so you might have two or more platforms claiming a conversion for the same order. To give an example, after viewing your ad on Pinterest, a customer goes to your site and places items in their cart, but delays ordering until days later when they see your post on Instagram. Both media platforms claim the conversion, which translates to inflated conversion metrics.
Orders data
This category groups all orders coming from eCommerce, retail stores (POS), and apps. This data provides accurate order counts, but without additional programming logic and effort, orders can't be reliably attributed to a specific ad campaign.
Google Analytics (GA) data
For each conversion, GA tracks the source (the site that claimed the customer attribution), the medium (how the customer got to that site, whether from an email, click ad, social media, and so on), and a related transaction ID. The source/medium values are the sites and campaigns that refer customers to you, and are key to determining attributions. The transaction ID lets us loop in the e-commerce order and the customer record, giving us the true ROAS and CAC for each transaction.
Occasionally, orders can't be attributed. This is caused by direct traffic or poorly configured UTM tags in your source system social and marketing campaigns. Direct traffic means the customer got to your site by typing the URL directly in the address bar of a browser, or by clicking a link in Skype, Outlook email, a PDF, or a mobile app. In these situations, Google marks the source/medium as “direct / none” and determines it to be non-attributable.
Attributed ad spend
Finding attributed ad spend is a matching game. First, we match ad spend to specific orders (order_id) for that day. Next, we match the same ad spend to a source and medium in Google Analytics so that we can credit the order, which is now a paid attribution.
Attribution ad spend uses the following logic:
- Source and medium match
- Campaign ID match (currently limited to Google Ads)
- Campaign name (if the UTM parameters have passed the campaign name and matches a media source campaign)
What about unattributed ad spend?
Unattributed ad spend is paid advertising that isn't associated with an order, transaction, or conversion in Google Analytics.
If you launched three facebook campaigns, and one of those campaigns landed two orders, the ad spend for the winning campaign would be split between those 2 orders. All the other campaign ad spend would be unattributed.
|
Ad spend |
Orders |
Attributed Orders |
Unattributed ad spend |
|
|
Campaign A |
$500 |
2 |
$250 (cost per order) |
|
|
Campaign B |
$500 |
0 |
0 |
$500 |
|
Campaign C |
$500 |
0 |
0 |
$500 |
In most cases, we can attribute ad spend to roughly 80% of orders. The remaining 20% of unattributed ad spend is distributed evenly across all orders placed through digital marketing channels to derive your CAC rate. This fully loaded, or burdened CAC, gives you a more realistic metric for forecasting future CAC than just your attributed ad spend.
Unattributed revenue
Ad spend is considered unattributed only after fields for source, medium, and campaign have been verified as correct and consistent, and all attempts to match the order_id have failed.
Unattributed revenue is the total revenue that can't be tied to any paid or organic marketing. For this reason, unattributed revenue is kept separate and isn't distributed across your campaign expenditures. You can find unattributed revenue on its own row in the dashboards, and in the attribution dimension (adspend order attribution view) labeled as No conversions.
Similarly, unattributed orders are any e-commerce orders not reported in the attribution data.
Clean data in / Clean data out
Producing a strong data model requires clean data -- free of duplication, inaccuracies, and inconsistencies. Maintaining this kind of hygiene in large data sets often boils down to following best practices when configuring your tools. For instance, any time you set up a new campaign or tracking tool, use simple, concise names. Use these names consistently across your other tools and channels so that data translates seamlessly between them.
UTM tags are essential to measuring the effectiveness of your digital marketing campaigns. UTM tags are snippets of code that get attached to the end of the URL on web pages you're promoting. When implemented correctly, UTM tags ensure that conversions are correctly tracked to your campaign. To make sure your UTM tags are working the way you intended, test them out before you deploy them. Review this Google Analytics article, Best practices for collecting campaign data with custom URLs.
Marketing overrides
Some attribution issues can be resolved by renaming and cleaning up field values from your media and marketing data sources. To handle this, we created an override tool that allows our clients to rename fields to create a cleaner aggregation.
When inconsistencies or variations occur in the source and medium fields that prevent us from combining like fields, we recommend that you override the original values using this tool. This is especially helpful in streamlining attribution data when a media channel employs several different source names for the same campaign. For examples and instructions, see How to use the Marketing Override sheet.
Module ERD
The ERD below illustrates the relationships between the views that comprise the Core module. This module contains the following views:
- Standard Views
- BI Views
Click on the above ERD diagram to see a higher-resolution version.
Comments
0 comments
Please sign in to leave a comment.