The campaign is complete. The month’s work is done. Now it’s time to assess what happened. Unfortunately, digital marketing reports often fail to get under the skin of what worked well or what to learn for next time. In this post, I explore a range of tactics to help you analyse and evaluate your work more effectively.
3 common issues with digital marketing reporting
How many of these do you recognise?
1. Mountains of data
Digital marketing produces loads of data. There’s a heap of different channels and a heap of different metrics. The temptation is to throw all that data into the report and make lots of pretty charts, without it necessarily being useful.
2. Repeating what the chart said
It can be easy to fall into the trap of using the “commentary” section of a report to playback what the accompanying chart or table said. The purpose of the chart is to present information in a more consumable way, so just writing that out in words isn’t adding much value. What we want (which is the focus of this post) is to understand what that data actually means.
3. A lack of context
Digital marketing, and especially social media, is full of moving parts and nuances. And that’s why context in digital reporting is the absolute king. Without sufficient context, how can we know if what happened was good, bad or indifferent?
9 tips for more effective analysis and evaluation
To help avoid those issues and start teasing out insights, I’ve found a great approach is to basically just ask lots of questions of what’s in front of you…
1. How does the data compare with the last time this activity was run?
- Was there anything notably different in the approach?
- Are there external factors (political, environment, social, technological) that might have had an impact?
2. How does the data compare to what you’d expect?
- For this you need some benchmarks – ideally based on your own work, or failing that some sensible industry figures
- If it’s a new activity, then this data becomes the benchmark for next time 😀
3. Do the numbers just look “a bit weird”?
- Take a moment, re-read and be curious… did performance really increase 200% month-on-month? … oh, whoops there’s a decimal place in the wrong place, it went up 2% (just like last month)
4. Is the selected metric meaningful?
- This may require some technical knowledge at times, but it doesn’t stop you asking the right sort of common sense questions
- Is 500,000 organic videos views good? Well, that depends. If the video was a six-minute story and everyone stopped watching at 4 seconds, then probably not. If it was a six-second teaser, and most people watched to 4 seconds, then it probably did a good job
5. Am I comparing apples with apples?
- Particularly with social media, it’s easy for things to get lumped together that shouldn’t, resulting in potentially meaningless numbers
- For example, an average engagement rate (based on likes, shares and comments) which includes posts designed to drive clicks doesn’t really offer a fair view
6. Is something skewing the average?
- If you remove the top 10% and bottom 10% of a data set you’re looking at, what do you see?
- Don’t let one awful piece of content drag down an otherwise strong campaign
7. Is the data set or data point too small to be meaningful?
- Yes, there might be a 500% increase from last month, but if that’s based on going from 2 clicks to 10 clicks, it’s hardly news to write home about!
8. Is there a trend that’s being hidden by looking at reports in isolation?
- A pattern might become very obvious if you look at even just the three previous monthly reports, for example
9. What didn’t happen that you might have expected?
- What isn’t in the data that tells a story?
The art of digital marketing reporting is something that definitely takes time to learn, and the more reports you work on, the quicker and easier the process becomes.
What tips and tricks do you use to ensure your reports work as hard as possible?
Wait… presentation counts as well!
Strong insights need to be complemented with well-structured reports. In this complementary post, I’ve suggested 12 tips for presenting findings more effectively.