Hello friend,
(welp, thought I scheduled this to send on Tuesday, but I just scheduled it to publish. Here you go!)
Today we’re talking data + dashboards. I’m not a super technical marketer but do know how to get my hands dirty in data. I did, however, get some great advice from my friend Alex Sofronas who is an analyst on the Advanced Analytics team at DirecTV. He writes about several ways to ask marketing questions and build dashboards around them.
Spending time on your data and dashboards is a tricky space to play in at a startup. Looking at marketing data and finding meaningful insights can be a dangerous rabbit hole. Everything is moving so fast that even if you find an insight in your data, it may be irrelevant in a couple weeks, or simply a fluke.
Plus, there can be misalignment with leadership in which metrics actually matter. This is an exhausting situation for marketing pros.
Consistent execution and testing is much more important than wading in a sea of data, hoping your answer to great marketing will jump out at you (pro tip: it won’t).
All that being said, understanding and keeping tabs on your data is essential for any business. Do you need to hire a marketing analyst for your 10 person company to dive into your data? Absolutely not. Should you take out an hour or two every month to make sure your metrics are properly tracking? Absolutely.
In order to make analyzing data easy, you can build a dashboard!
If you use Google Analytics, for example, everything you see can technically be considered a dashboard. Though, for the sake of this newsletter, a dashboard is a place where you intentionally place reports focused on specific KPIs.
Some examples of marketing dashboards:
Executive dashboard
Purpose: high-level reporting on top marketing KPIs
Can include overall website traffic, Emails sent/opened/clicked, total lead conversions, maybe meetings booked
Email Performance
Deliverability rate, Open Rate, Click-Through-Rate, Emails sent, Unsubscribe rate
Broken out by campaign, quarter, month, etc.
My overview dashboard for UTM Link Manager
This is where I track high-level activity for UTM: campaigns, traffic sources, conversions, etc.
You can get the template and add it to your own GA account here: https://analytics.google.com/analytics/web/template?uid=1KPm2UliRm29t1_g2wohSA
How I think about building marketing dashboards
There are lots of ways to build a dashboard. Ultimately, it comes down to the story you want to tell.
So first, you need to understand how to communicate data:
have a strong understanding of your goals
know what your audience cares about (the people hearing about the data)
anticipating fears, delights, and of course, questions
Your dashboard should be designed so it’s easy to deliver the information in a way people will understand. This is the basic structure you’ll find in most research reports, analytics decks, etc.
Direct point/insight (ex. Our customers love our new product)
Supporting point 1 (ex. Product engagement is up by X% as noted in this chart)
Supporting point 2 (ex. Social sentiment around the product is mostly positive as shown in this chart)
supporting point 3 (ex. Initial survey data shows a 90% recommendation score)
Then, just rinse and repeat across whatever insights you want to deliver.
What order should you put the reports in? After all, a dashboard likely doesn’t exist for a single presentation. It should be a living thing that evolves as your marketing evolves and should be able to provide value whenever someone opens it.
For marketing dashboards, it can look something like this:
Volume: understanding counts in the data, will likely include conversion rates, volume shifts (% or numeric), averages
How much traffic did we get?
How many meetings did we book?
How many emails delivered?
These questions and associated data are often the most important numbers people care about and tend to be the most foundational. Volume data helps contextualize everything else. If no one’s showing up, you’re not getting any meaningful inputs for other reports in the dashboard. Or if you’re seeing a rise in traffic, that is an indicator to find out where and for how long.
Time: much of it will be volume data over time
What is our traffic by marketing channel over time?
What is our blog’s bounce rate over the past quarter?
What was our revenue for each month in 2021?
As mentioned above, these questions can help answer how long a trend has been going for and be a great way to indicate when a positive or negative event occurred in your marketing.
If you’re sharing an insight that’s based on data from the last week - but you’re doing an annual report- then that’s not going to be very helpful.
Sentiment: great section to add more qualitative data to your overall dashboard. Often based-on social media + Social listening data, can also be customer satisfaction, support, and other feedback surveys
How many support tickets did we receive this month? How many received a negative rating?
How has our brand sentiment shifted on social over the past year?
What were the replies to our new product announcement?
If you have a story to tell with data, build your dashboard with the story in mind.
In many cases, this “story”, will directly align with your strategy. For example, if your marketing strategy is focused on Top of Funnel activity and generating demand, your dashboard will reflect that. There will likely be more emphasis on certain areas of your marketing engine over others in the dashboard.
This next section includes many more specific examples of marketing questions + directly how to answer them using data. It may be a bit more advanced and I encourage you to stick around.
How to Build a Dashboard from Scratch using Data Exports
Start by mapping out the levers of the business.
Maybe you can shift spend between channels, increase spend caps on specific keywords, and increase spend on specific audiences. That becomes your baseline to build the dashboard from.
Think of some questions to answer around those levers.
For example:
what is my cost per sale across all channels?
Which keywords are driving most of my traffic, and have the biggest upside with increased spend?
Which audiences are showing greatest CAC:LTV ratio?
From here, map out which variables you need to answer these questions.
See if you can get total spend and total sales by channel by month. This will allow you to build a trending line graph with different channels and their sales efficiency.
If you want to look at keywords, come up with a success metric like close rate or engagement rate and rank order keywords by that success metric.
From there, overlay a metric like impression share that can inform which keywords have room for increased spend.
For CAC:LTV, you need spend and revenue per audience, common audience delineations between channels, and common audience delineations between pre-purchase and post-purchase customers.
Once you know what data you need to inform the business decision-making, the next step is to think about the user and what filters or additional variables they may need.
If you work at a big company, you may need to split results by area of the business that somebody may oversee. Or a geographic limit to traffic and spend in a particular country. Or a product interest that we can track in both our marketing and purchase data.
Then, map out the data you need in a table and make sure there is no inherent duplication in the data you want to show.
For example, if you receive customer data by the minute, but want to report at a weekly level of aggregation, then you have to transform the minute-by-minute data into a monthly aggregation. Otherwise you will have all of the monthly variables repeated hundreds of times, while the minute-to-minute data remains unique.
Tools like SPSS, SQL, and Python are great for data transformation to create a finished dataset.
Once you have your dataset ready, you’ve done 80% of the work.
The final step is to visualize the data in the easiest way possible.
One of the best tips for visualization is writing out steps on how to use the dashboard so that the user knows exactly what to do to get to a customized insight for them.
You can literally write:
Step 1: Filter to your country/area of business.
Step 2: Select top audience and analyze keyword performance, and so on.
But ultimately, if you’ve done your job right, the visualization is very straightforward.
It’s the data you have brought together, all with a focus on the business lever, that is delivering the true value.
That’s all I got for this week! Really hope this was helpful. What marketing issues are you facing? Would love to hear about them - can offer advice, talk about it here or happy to just be the listener to a good ole rant.
See you in a couple weeks!
-Connor