One of the main questions that always arise when I meet fellow marketers – maybe it’s not really a question but more of a general wonder – is “how do I know that my campaigns are bringing value to my organization?“ And in the online marketing industry, understanding the incremental revenue generated by campaigns is the eternal conundrum.
More from PostFunnel on CRM:
Opted for an In-House CRM? Here’s What You Should Know
Power Stepping your Way to a Full Marketing Journey: A Quick Guide
How is Your Organizational Structure Harming Customer-Centricity?
Wouldn’t you love to know the monetary value generated from a campaign you executed, to make sure your efforts had a measurable impact? But what incremental revenue should be measured?
The historic answer has been to select an attribution model (i.e. first-touch, last-touch, multi-touch) for customer acquisition, and use the revenue resulting from that acquisition as a proxy. However, as customers have become more sophisticated and competition has risen steeply, marketers are shifting away from acquisition-focused marketing towards retention-marketing.
Today, smart marketers look at revenues generated from existing customers as a stronger measurement for incrementality generated by marketing activities. This shift comes with its own challenges – on how to attribute the specific monetary value generated by a specific customer, to a specific campaign?
Over the last few months, I’ve been working hard to crack this riddle. Thanks to Optimove’s clients that focused on creating test vs. control campaigns, I have a solid ground to perform this test and share with you the steps of this research, and some of the conclusions I came up with.
For better measurement – experiment
Optimove followers have heard us pontificating about the importance of the test vs. control method many times before. It’s simple: the optimal measurement starts with treating your campaigns as experiments.
Every time our customers are executing campaigns through a target group, we encourage them to choose the test and control option. We split this group of customers into two groups which differ in only one way – one group (the vast majority) will receive the campaign, and the other will not. This test vs. control method eliminates alternate explanations for the campaign results and thus gives us the best indication of whether the campaign was successful or not.
Fine-tuning the formula
When examining the CRM Contribution of various Optimove clients, we wanted to find a KPI that can be measured among these clients. There are two important points to consider:
The first, in reality, not all of our customers are measuring all their campaigns using the test vs. control practice. Let’s say that you calculated a CRM Contribution of $100K this month but measured only 80% of your campaigns. You have to base the assumption that the uplift generated from 80% of these campaigns also reflects the uplift in the campaigns that lack a control group. In this case, it will affect our calculation as follows:
Revenue uplift / %Campaigns with control. Hence – $100K/0.8 = $125K.
Also, it won’t be reasonable to compare the uplift that one client had – a client that has an organic yearly revenue of $100M – to one that has only $10M in annual revenue.
Therefore, we decided to add another denominator to the calculation: CRM Contribution will be a CRM Contribution percentage, and the final calculation will be: (Revenue uplift / %Campaigns with control) / (Total revenue).
How to strike ten…
By comparing various clients’ performances, we can see our clients’ CRM Contribution score. In order to have a valid sample of clients, I set thresholds that would exclude clients that might skew the results of our experiment, based on these characteristics:
1. Seniority – Included only clients that launched their site at least 6 months ago.
2. Annual Income – Set the threshold to exclude smaller clients.
3. Coverage Ratio – We’ve taken into account clients that have high coverage rates of their customer database.
4. Test vs. Control – Examined clients that used this measurement method on at least 33% of its total campaigns.
Examining the results, it was clear that there are massive variations amongst our clients, when comparing the CRM Contribution. Now the goal was to find the metrics that explain the differences.
We ran research of clients’ first-year performance and tried to find the most correlated features to point on future uplift. Of dozens of attributes taken into account we came up with the following key factors to success:
1. Customer Base Coverage – It makes sense, but you’ll be surprised how many customers are being left out on a weekly basis and are underexposed to campaigns. We learned that covering as much of your customer base as possible will lead to higher uplift.
2. Number of Target Groups – Granularity matters. Sending out campaigns to everyone is nice, but the ability to send the right campaign at the right time, to the right group of customers, is extremely effective. The clients who worked hard in order to create various target groups also had better performance in CRM Contribution.
3. Number of Channels – Clients who used many channels outperformed those who used only one execution channel. Communicating through the right channel leads to better performance.
In conclusion
In this blog, we shared the research made to understand CRM Contribution. We saw the need for CRM Contribution KPI and the components of it. We learned what metrics have an impact on this KPI, and how we can work to increase our CRM Contribution.
In our next piece we’ll drill down on the magic numbers derived from our research and try to answer the many questions that undoubtedly arose:
- “How many target groups are needed?”
- “When should I reach the goal of two channels?”
- “What is the portion of customers that I would like to target on a weekly basis?”
- “Should I consider all of the customers in my database for the coverage ratio?”
Stay tuned.
