“Are these numbers good? Or bad? Give me a clue!”
“Why is that number so big? I don’t understand it. Or believe it.”
“Should we be doing more or less of this?”
A little glimpse into the mind of whoever you’ve sent your last coverage report too….
I’ve been thinking a lot about how we measure things in our software business. From feature usage data, marketing data to app performance data. There is so much of it. Maybe even to much of it. Analysis paralysis is a thing people.
The tools to spit out numbers are plentiful. CoverageBook is one of them. A tool I’ve helped to make for PR teams to help communicate the performance of their coverage. But just because numbers are easier to come by and present. It does not mean you can slack on the interpretation bit.
Automation buys you some time on the easy stuff. You can use that time in the bank to get better at telling the story of your data. Communication specialists could (and will be) world leaders at this and put some of those annoying data puke “digital” teams with their endless spreadsheets to shame.
So we need to start somewhere. And a good place is to learn about adding context to your numbers.
In my examples, I’m going to focus on mainly output metrics here. i.e. things you can count + measure. Real + proxy metrics can be useful to track progress. They may not help you work out the outcomes of your work. But there can be no outcomes with at least some output after all! If none of the above makes sense. I get it. It can feel overwhelming. Go learn more from the AMEC framework.
But for now…
Here are 5 ideas on ways to make it easier to understand numbers in a report…
1. Add some simple goals or targets next to metrics.
Instead of just saying 12 pieces of coverage.
It’s better to say…
12 pieces of coverage. The target was 5.
i.e. We did good!
Assuming all the coverage was on target media of course and was of some reasonable quality. Go read about Katie Paine’s “Kick Butt Index” for some inspiration on how you could do this.
Setting KPIs is hard. But do try and start somewhere. If you missed the target. Maybe it’s because your client/boss didn’t sign off on the better idea. Use this data to help make the point.
2. Ghost Lap! Add Context through past performance on that metric.
I use a tool called Profitwell to track all kinds of metrics relating to the business. It lets you set targets and see past performance in one graph. I can see how we are doing against a target. Then vs last month all in one visual.
One CoverageBook customer referred to this as running against a ghost lap. Trying to beat your personal best.
It’s better to say or show…
12 pieces of coverage. Last month it was 20.
12 in December. Here is trend for last 4 months.
e.g it’s nearly end of month and goal was 5 bits of coverage. we have 1. what are we going to do. Or is it ok as over last 3 months we have exceeded target?
3. Benchmarking. You vs you. You vs Industry.
Are my numbers any good?
200 social shares. What is normal? Maybe for this client it’s the best they’ve ever had. Maybe it’s the worst. Without giving this context how will the reader ever know?
One way to get started is to answer this is to compare against your self. Rich Leigh shared some great insights here from their own agency Radioactive PR.
Other tools providers have some fantastic resources that could help you to. What is a decent number of engagements on youe Instragram campaign? Go check out this Social Media Benchmarking report from RivalIQ.
I think CoverageBook has a role to play here. More on that another time…
4. What are the “Must See” Metrics?
Some time ago I spotted some wise words on this…
“Gumtree head of communications Fergus Campbell warned there is a risk PR professionals can drown in data and the key is choosing the metrics that matter.
“You need to identify what single data point matters to you and ignore the bullshit noise and abundance of numbers that aren’t helpful to you”
You can do yourself a favour here by taking the time to explore with your client/stakeholders on what metrics matter to them most. Then consider cutting out the irrelevant ones at least in the report you share with them.
If you’re designing your own reports you can make sure to emphasise the most important metrics. Make sure your eye is drawn to it. Don’t let the reader drown in dashboard hell.
Again. CoverageBook could do much better here! Working on it.
5. Looking for trends over time.
Start by making a time-line of when your coverage/campaigns landed. And overlay with other data over time.
E.g. Campaign A coverage landed in February. Campaign B in April.
Look back and compare dates of when coverage landed vs many things.
- increases in searchers for that brand (google trends data)
- sales/lead data
- increases in search engine visibility of a specific website
- increases in stock price
- increases in approval ratings
- increases in donations
- increases in applicant rates
And so on and so on.
This one takes a bit more time. But hopefully, it’s triggering some ideas.
If you’re looking for brownie points you could always proactively use the annotation features of other measurement tools. E.g, In Google Analytics you can add text annotations.
We use annotations for all kinds of things. Product changes. But marketing stuff can be recorded here to:
Anyway. I think that’s enough for now. If you’re doing this already. Well done you. If you’re not. Don’t worry about it. Nobody has mastered this. Not by a long shot. Take the simplest idea on this list and try and make a new habit.
I’ve also helped to create a tool that does a poor job on most of these points above. So while we continue work on that (and we are working on it) I hope at least one of the ideas here makes some sense. And could go a small way to making life a little easier for whoever is on the receiving end of your report.