‘Dashboard’ is not a dirty word: Napkyn’s defense of an overused term
There are a lot of terms that get badly abused – to the point where they are annoying to hear. A few personal examples would be the words ‘cloud’, ‘big data’, ‘guru’ and using the label “2.0” for anything other than a software release. The word “social” is border-line with me right now as well.
The word “dashboard” is not on my hate list, but it seems to be on a lot of other people’s. Everything has a dashboard now, from software products, to consulting deliverables (we do them weekly for our clients), even video games and LinkedIn have dashboards in them.
So, I get why the thought of another dashboard would make many people want to barf, but it doesn’t make me think they are any less vital to being successful with data.
A few points below that you should consider before you give up on dashboards (I had a great Karate Kid reference I was going to make, but I have been informed that I need to dial back on the 80s film references…).
The obvious point:
A dashboard you don’t love wasn’t built properly: A good dashboard is supposed to provide ‘at a glance’ understanding of something you care about, giving you the ability to have increased understanding you wouldn’t have achieved any other way. If you are indifferent to a dashboard, it sucks. Get rid of it and have one crafted to your exact needs.
Less obvious but critical points:
1) Dashboards create commonality of language and goals: Ever notice that sales and marketing people use different words to describe the same thing? Ever notice that they don’t tend to get along? A well crafted (and agreed on) dashboard has the ability to create significant alignment in an organization, not just between different departments, but between different levels of the org chart. In helping a senior stakeholder build an exciting and relevant performance dashboard, we are very educated on where to focus our analysis on their behalf.
2) Dashboards decrease “weaponized” analysis: In organizations without proper executive dashboarding (so most of them…), the lack of common language, goals and structure creates “analysis anarchy”. This means that execs often ask the analysis to provide reports and data to support an idea or initiative. We call these reports ‘weaponized’ because they are only being created to provide data support to help win an argument, i.e., “Build me a report that shows how marketing has been wasting money on leads for the last 6 months”. Competing on analytics is supposed to be an external activity, not an internal one. It’s very hard to weaponize your reports when the whole team ultimately works again an agreed on performance framework.
3) Good Dashboards maximize analyst value and contribution: the lack of structure around data consumption is the Petri dish that dumb questions grow in (would that mean ‘bad culture’?). Many of the ad-hoc questions that web analysts get are based on a lack of overall understanding about digital. We find that instituting an appropriate and valuable weekly performance dashboard cuts down on ad-hoc requests significantly, and increases the quality of the questions being asked. Good questions tend to be harder to answer; your analyst will still be 100% utilized, but purely on creating insights of high business value.
If you are making plans for 2012, and someone comes out against using a dashboard, realize that they are tired of the abuse of the term, not the value proposition.
Just call it something cooler, like a ‘executive performance visualization’. At the end of the day, a well executed dashboard can have both corporate and competitive effects that are far reaching.
Cheers,
Jim
December 7th, 2011 at 12:11 pm
Great post, Jim. I think you should have went with the Karate Kid reference though
. Long live 80s movies!
December 7th, 2011 at 12:37 pm
Jim:
Interesting post, having publicly stated my distaste for ‘dashboards’ the real issue is not the dashboard itself. Rather, the implied dependencies to create functioning dashboards.
The hidden nugget is in point 1:
My frustration with dashbords in general is that companies are often internally inconsistent in their business rules for measurement, classification, validation and valuation of the key metrics which drive a dashboard.
Having built out a solution for a large scale website which monitors the tagging implementation and validates alignment with business rules which have been developed, I know this can be done.
Take a simple metric such as bounce rate at company XYZ, do the business units at XYZ all understand bounce rate the same?
Is bounce rate at XYZ visitors who visited the site for less than 60 seconds?
Do we discount visits of less than the page load time, or some other cutoff, and only consider visits between say 20-60 seconds bounces?
Or, more precisely in GA, if a person makes only one GIF request during this prescribed time frame?
Who is continuously watching the tagging implementation for the web property to ensure consistency of data collection?
Is anyone looking over product’s shoulder to catch them implementation tracking for some new feature, which creates a second GIF request, which marks a visit non-bounce when it is a bounce?
Suppose bounce rate goes up, then what?
What is the response to an increase in bounce rate?
Will anyone create action items from that delta?
Building out dashboards that lack process for:
Business rule alignment
Continuous tagging validation
Processes for handling changes
Seems a bit like putting the cart before the horse.
Clients, whether internal or external, want to have the flashy charts with as little effort on their part as possible.
This, of course, can be accomplished. But if I make a dashboard lacking consistency in any of business rules or repeatable measurement, what have I created?
Paraphrasing Edward Tufte:
Hammering out the process for data intake, storage, processing, analysis and presentation is usually close to voluntary root canals for clients. However, just like a root canal, when you need it you need it and it needs to be done ASAP.
Michael D. Healy
http://michaeldhealy.com/
December 8th, 2011 at 8:18 pm
Two comments to a blog post that not only aren’t from my mom, but are from some analysts I really respect. Christmas is early this year.
@Bill – A good Analyst is like Mr. Miyagi. In getting senior stakeholders to continually perform seemingly trivial tasks over and over (review weekly traffic conversion summaries), they will unknowingly develop the basis skills of analysis kung fu. (or not…but I needed the metaphor to work). So Actionable Analytics = Wax on Wax off…or something.
@Michael – Kickass feedback, and based on your points I think we are violently agreeing. Most of my secondary wins in the post are the outcomes of a dashboard that has come from a process like you mention in your final point. Getting to that ‘holy grail’ report sure isn’t something you can get from one meeting, it happens organically over time, as ideas are socialized and multiple stakeholders buy in. But when it all comes together – the impact on both the bottom line and the organization at large are significant.