Dashboard Design Mistake #2 - Not adding comparison values
Imagine you run sales for a company and you see the following metric on your dashboard:
How should you react to that number?
And how about this?
Unless you happen to be intimately familiar with the history or targets for sales or average call duration, those numbers will raise more questions than they answer.
You won’t know whether those numbers are good or bad, or whether they are normal or unusual.
And more importantly, you won’t know whether any action is required.
Without comparison values, numbers on a dashboard are of little or no value to users. And if users don’t derive value from checking your dashboard, they’ll stop looking at it.
Yet all too often, dashboard builders fail to include comparison values.
The good news, though, is that this problem is easy to fix.
What are comparison values?
Glad you asked... they’re values that compare current numbers to either historical values (today’s sales compared to yesterday’s sales, for example) or to targets (average tech support call duration today, for example, compared to the company target for the duration of tech support calls.)
Either way, those comparison values put today’s numbers into context:
What kind of comparison values should you show on a dashboard?
There is no set rule about what to use; however, there is one guiding principle that applies across the board: Whatever you show has to be meaningful to users. Sounds like common sense - but as the old saying goes common sense isn’t very common.
As a general rule, most comparison values fall into one of the following categories:
- Comparison against a set target - for example:
- current average customer satisfaction rating vs. a target rating set by the company.
- Comparison against a preceding period - for example:
- today vs. yesterday
- Comparison against a prior equivalent period - for example:
- this week vs. the same week last year
- Comparison against a trailing average - for example:
- today vs. an average of previous 30 days)
- Comparison against a projection - for example:
- current cumulative total vs. expected total for this point in time
- Comparison against another metric - for example:
- new account activations vs. account cancellations.
Sometimes, you will need to use both a comparison against a prior period, and a comparison against a target to get a target that will be truly meaningful to users.
For example, if your organization’s management has said that it wants to increase sales revenue by 10% over last year, and you know that your sales are highly seasonal, then a good comparison value for today’s sales might be the sales for the same day last year, plus 10%.
Are some comparative value types better than others?
Comparison values based on organizational targets are generally more meaningful to users than values based on historical data, since they say whether current values are “good” or “bad”, as opposed to just better or worse than in the past.
If an organization has failed to set targets, then the dashboard should probably compare current numbers to historical values, since that will still be useful.
If it’s difficult to come up with a meaningful comparison value for a given metric, then it is unlikely that anyone will need to act on it if it changes - which means it doesn’t belong on your dashboard. Learn how to define your organizations KPIs.
What if you know that your users are intimately familiar with the history and/or targets for a given metric? Do you still need to include comparison values?
Why yes you do!
First of all, those knowledgeable users will not always be the only ones consulting the dashboard. The dashboard has to have meaning to new hires, or to people filling in during vacation or sick leaves. Without comparison values, less-experienced users probably won’t know what to make of the numbers they see.
How to increase the speed a user can scan a dashboard
Adding design elements like coloured text and shapes will instantly draw users’ attention to metrics that are significantly off-target or in an abnormal range - dramatically increasing the speed users are able to scan a dashboard.
Again, it’s all about making the dashboard relevant and useful to users – something that good dashboard design should always keep in mind.
Originally published November 26, 2015, updated Oct, 07 2019