Thoughts On Setting Up A Useful Metrics Dashboard For Email Data

When I work with companies using Klaviyo, usually, they have questions about metrics and what to track for evaluating their email program. 

If you search the internet for ‘email marketing metrics’, you will find a wealth of articles online discussing a variety of ways of using the data generated by email delivery systems like Klaviyo (Omnisend, Sendgrid, etc.) to create reports that will reflect the performance of your email efforts.

People who have read those endless articles about email marketing metrics worry about numbers such as open rates, click rates, deliverability, bounce rates, inbox reach, etc. 

I usually begin the discussion about metrics by asking: What do you want your email campaigns to do?

Usually, when it comes to ecommerce businesses, the answer is: bring money into the business so the program pays for itself and makes a solid contribution to the bottom line. 

Non-ecommerce businesses, usually B2B arrangements, have different ideas about what they want their email to do. In their case, it is a tool to bring interested individuals into a non-email driven sales presentation process (demo, sales chat, appointment setting, etc.)

Once there is a clear idea of what the business wishes to obtain from an email program, then I can proceed with creating a dashboard that shows results that will be legible to the business. 

I build dashboards to answer specific questions the business may have about email performance. 

For example, for a specific time period (month, week, etc.) and assuming that Klaviyo data are your primary source for your email campaign dashboard, I usually start by trying to answer these questions:

1. How many sales have campaign emails brought into the business?
2. How much income have those sales generated?

These two questions deal with the heart of the business need. Usually, your email program should be bringing at least 20% of your store’s income per month, depending on what you sell and how you sell it. 

The answers to those first 2 questions set the table for the next set of questions. Based on those results, we can ask further questions to clarify the answers provided for the first 2.

3. How much friction to sales do our email programs generate?
4. What can we do to reduce such friction and improve the results of the programs?

This last question will always come around, regardless of whether the email program is making 20, 40, 60 or 80% of the sales for the site. The assumption is that it can always do more.

At this point, the dashboard needs to display data showing how the email program is performing by answering questions that clarify potential problems with the current way of doing email.

5. How are the Open Rates relating to the Click Rates of our email program?
6. How are the Click Rates relating to Started Checkouts and tracked through email?
7. How are Started Checkouts related to Orders Placed and tracked through email?

Questions 5 - 7 provide you a good way to understand how your campaign-based email program is performing per time period. 

• Question 5 shows you whether people are opening your emails, reading them, and then performing a specific call to action requested in the email.
• Question 6 shows you how likely the performance of a call to action in your email is to lead to a checkout activity.
• Question 7 shows you how likely a started checkout transforms into a completed sale (as tracked via email)
  

This information can help you create potential tests to see if you can move the numbers you see in the data. For example:

• Question 5 Areas of Concern: Focus on subject lines, email messaging, segmentation and messaging per-segment, and call to action testing. Tools such as Klaviyo allow you to test these in campaigns and in automated flows.
• Question 6 Areas of Concern: Examine the connection between the messaging in emails and the messaging on your landing pages/online store. Usually, testing here will involve changes to the website as well as to emails themselves. 
• Question 7 Areas of Concern: Examine the automated follow-up for Abandoned Checkouts to test whether additional messaging, design, special offers, etc. can help close the deal for each campaign sent.
  
  

In looking at these questions, you can bring in additional information that might be helpful to establish that there are no infrastructural problems with the email process itself. This is the place where deliverability concerns may pop up, or bounce rates, in terms of making sure that the email process itself is running as smoothly as possible without infrastructural problems. 

For example, one problem that may arise is the requirement by the business to use brand-approved fonts in their emails. Unfortunately, unless such fonts happen to be web-safe fonts, it is unlikely that many individual computers will happen to have font-families like Inria, Mulish, Wigrum, etc. or whatever designer-selected font was chosen by the brand. This means that, if the text is ‘live’ in an email, it will likely display with web-safe font substitutions when it reaches the recipient's inbox. Sometimes, the email may not be readable at all because of this and have a negative effect on the areas concerning Question 5.

You can solve the specific problem of using brand-approved fonts by making the entire email a series of images, or a single image. The fonts will not be affected by the recipient’s inbox. However, you do have to make sure that the phone version of the email does not shrink the text in such a way to make it unreadable for recipients. You can also solve the problem by sticking to web-safe fonts as much as possible for live text that can be personalized.

But I digress.

Question 5 provides the opportunity to split the information by campaign and by list, so the data will be doing double duty in terms of providing good ideas on what is likely to work with which segment of the audience in your list.

Questions 6 & 7 expand the concerns of the email program to the areas where the conversion is supposed to take place. It is important that the email program not be siloed within the organization and that good communication exists between the website group and the email group so tests can be performed and corrections made in case there are misalignments, but also as a process of continual testing leading to improvements in the answers for Questions 1 & 2.

Question 7 brings into the dashboard data from the automations created to manage specific situations like abandoned carts. The data can be derived from Klaviyo and added to the dashboard to illustrate where things stand. Usually, the same data points are used in both areas so it is easy to see where things can be improved and/or corrected.

Outside of campaigns, data from automations can be used to establish how they are contributing to the answers to Questions 1 & 2. The procedure is the same, and the dashboard information is added in separate areas for clarity.

I usually build dashboards using Google Data Studio, and bring Klaviyo data on a regular basis to update the information during the year. 

If you have data for multiple years, you can expand the power of the dashboard by adding time-based line charts that can compare the state of specific data points at similar times in the past. You can use these to consider potential trends or compare the performance of the business from year to year. However, time-based charts also have to take time-based historical considerations that could have affected performance and had nothing to do with email itself. 

By focusing on answering specific questions, the content of the dashboard is very clearly related to the state of the business (as the business defines it), and using email and store-generated data through Klaviyo provides a simplified data management method for understanding the state of the program, its usefulness to the business, and potential avenues for improvement.
​
There is no one-size-fits-all solution for dashboards, and every business has a different take on how they want to view their email data, particularly as it pertains to attribution. This discussion focuses solely on using Klaviyo and Store data to derive a dashboard to understand the performance of an email-based campaign and automation program. Depending on the complexity of your needs, and your priorities, you may have a completely different approach that incorporates other aspects of the business and makes the dashboard valid for your business needs.