Advertising Data Analyses, Part 2
Note: This is a continuation of my last blog post on marketing metrics, please find it here: https://daria-morgan.medium.com/marketing-data-analyses-part-1-e934d87b26a2
My last blog post walked through a few informative ways to examine FB/Insta ad spend with a focus on how performance differed across audiences and ad assets, as well as how to see if certain types of ads resonated with certain audiences or were generally well liked.
This blog post focuses on how get a better understanding of your performance, how it has changed over time, and answer the big question why ad performance has changed over time.
View 1: Spend by Ad Set over time and overall CAC
I start with this view because it really helps you visually understand how your ad spend has changed over time, where that ad spend is going, and your overall performance over time. (I also think it’s a beautiful chart) In the above example (all data is made up) you can see that their overall ad spend has increased materially recently, and that this has not really impacted their overall CAC. They reduced spend on many of their old ad sets, and are now relying heavily on Ad Set 1, Ad Set 4, and Ad Set 6. If I was working with this account, I would be impressed with how they were able to basically double spend without a large impact on their performance.
View 2: CAC over time and key drivers of CAC over time
I really like this next view because it is a great and simple summary of the overall CAC health of your ad account as well as shows you how each main driver (CPM, CTR, CPC, Conv Rate) has changed over time. For example, in the below, you can see that the 14 day weighted average CAC is now at $60, and that this is a bit higher than it has been historically. Next, you easily tell the story of what has driven that increase:
1) CPM’s (green) increased in the middle of this period but have come back down to starting levels,
2) However, at the same time CTRs (blue) have fallen, especially recently, perhaps because of the increase of spend (seen by the thickness of the lines)
3) This has driven CPCs (yellow) higher
4) Generally increasing CPCs have bene slightly offset by conversion rates (purple), which rose a bunch in the beginning of this period yet have been decreasing recently
5) The combination of these factors have led to a slowly increasing CPA (teal)
View 3: Drivers of CAC by ad set:
Next, I show a quick review of what drives each ad set’s performance and how each metric compares to those of other campaigns. I had this same analysis in my last blog post, but decided to include it again in case you are finding this post first, and because it is one of my favorite ways to dig into ad account performance. From this view you can see that Ad Set 4 and Ad Set 1 receive most of the spend, and that Ad Set 8 has bad performance driven by a weak CTR and weak Conversion Rate.
View 4: Breaking down driver performance on the ad set level
The first view can also be duplicated for any individual campaign or ad set, to also get a good sense of each one’s performance story. For example, in Ad Set 1, you can see that CAC has shot up recently largely driven by a decrease in conversion rates, as well as a decrease in CTR. (Again, I made up this data so if it looks weird that’s because it is).
View 5: See CAC, spend, and change over time by campaign
One view that I created that I find really helpful shows CAC not only by campaign, but also shows the relative spend by campaign through each line’s thickness.
What I like a lot about this view is that it answers a few key questions you may have while analyzing your marketing data such as:
1) How are our ad campaigns performing?
2) Are we doing better or worse than a few days / weeks ago?
3) Have we been able to scale up spend on high-performing campaigns?
4) Which campaigns are driving good performance, and which are underperforming?
By looking at the view below, you can quickly and easily tell a story around what has happened in this ad account in the past few weeks. (Again, all this data is made up so please forgive me if some of it seems unrealistic). For example, you can see that in this account, Ad Set 1 became one of their top performers in mid-December, and that they increased its spend (as seen by the increased thickness of the line), and as they increased spend its overall CPA became worse. You can also easily tell that much of the other spend goes to Ad Set 4 and Ad Set 6, and that their performance and spend has been relatively constant. Furthermore, Ad Set 3’s spend has been relatively constant, yet performance dropped off — perhaps this is a sign you should dig into the ad set’s frequency or refresh ad assets (images, gifs, etc) there.
View 6: Individual drivers (CPM, CTRs, CPCs, Conversion Rates) by campaign
Next, I will show a few other views very similar to the CAC one that show each metric by campaign. The purpose of these views is to get an overall sense of the changes of each one — e.g. CPMs may be up across the board, simply because it is a holiday season. These graphs also help you easily see if one individual campaign is acting differently. For example, in the CPM chart below, you can see that Set 3’s CPM shot up suddenly (again, to not show any brand’s actual data all of this is made up). You can also see that as you scaled up spend in Ad Set 1, that your CTR has fallen. Perhaps this is a sign that you should examine that ad set’s frequency and potentially refresh images.
Conclusion
This post adds to my last one (link here: https://daria-morgan.medium.com/marketing-data-analyses-part-1-e934d87b26a2) and shows a few great ways to not only visualize but also analyze your marketing data. This post shows how I think you should examine trends over time, and easily read into why each campaign’s CAC has changed.
As always, I’m always interested in hearing any feedback or new ways you might know of to show data!