“This one of the core reasons why most dashboards are ‘crappy’, i.e. they are data pukes that provide little in terms of context and even less in terms of actionable value.”

In the past, we have offered tools to make data presentation as clear as possible (chart chooser, Excel chart cleaner). Sometimes clean isn’t enough; a more dramatic approach is needed.
[...]

My added comments are as follows: swap Quadrants 3 and 4 - the outcomes are usually more powerful on foot of the consequences, so read it last. this is especially true if you feel the following ideas merit inclusion.

Also, in the Actions/Steps Quadrant, it is frequently useful to include the specific intended end-result from the action/step. Often, the end-result is unclear or unspoken and this alone can result in wasted effort and misdirected steps. So, the guiding question is: “What specific results are we after?” and hold the actions/steps being considered up against it. This also allows for a more creative approach and the mining of less obvious but better-formed action choices. After all, the outcomes should dictate the actions/next steps.

Finally, you may want to include a brief Cost Benefit Analysis of the proposed actions - this is where you can heavily (and professionally) score on getting the right “influencers” to put their weight behind your recommendations.

I have been looking for helpful examples in web metrics reporting and came across this site…

http://dashboard.imamuseum.org/

I was impressed by the style and simplicity of the various reports featured on the site. To get an idea of how the museum displays this data, I took a look at the source html code. My question is this:

Would anybody be able to shed light on what sort of computer application allows the user to produce web widgets like the ones shown at the website above. I know that Apple allows you to with Dashcode, but I’m interested in learning about others that produce similar results. If anyone can offer any input or suggestions, I’d really appreciate it.

My company is in the process of revamping our website and I feel that widgets in the style of imamuseum.org would be perfect.

Thanks!

For those of you charged with presenting your company’s metrics, this is a must read. Kushik’s article recommends a better way of presenting numbers (financial, sales, web or otherwise) through the use of an “action dashboard” that provides meaning to the reader.
[...]

It’s true for *all* reporting, not just dashboards. When executives get emails with just a report of sales, unrequested, I suspect it goes in the trash.

I challenge my team to ask the question “So What?” when looking at data, and use the answer(s) to that questions when crafting the narrative.

Thanks for yet another well structured instruction plan that is enormously useful.

Typically my experience is that drill downs are an excuse for not actually putting in enough time to dig for the insights and providing answers. So the report writer will take the easy way out, create a wiz bang flash driven excel backend infinite drill down dashboard. The problem is that the Decision Maker will need increasingly ever more context the more they drill down, context that they won’t have.

So my recommendation is that, atleast for Senior Executive dashboards, the Analyst should spend time understanding key trends and developments and include the insights / context in the other quadrant, thus reducing or eliminating the need for drill downs.

Of course each case is a little unique but I hope the ability to provide drill downs is not a excuse for not doing the hard work up front.

Would you agree?

-Avinash.

That’d help u have a good look at the person ;) and not be disappointed (yes you are married but looking is still ok :)

HIPPOS usually dont have the time to read more than 3 slides anyway. What I like hear is that you’ve basically taken 3 slides and combined them with the metric chart to create one all inclusive slide.

Very good post.

I think that the reason that so many organizations build autmated data only dashboards is that it is easier than building the dashboard that you propose. It takes time to absorb the data, analyze it, come to conclusions and then tweak and present them. It’s much easier to just pump out the data.

My 2 cents…

My 0.015$ (yes less than 2 cents) cents on metrics and dashboards.

* For the site as a whole a few key metrics do tell the story. The roadblock i hit is when owners of different parts of this site try to align their KPIs to mirror the “strategic” goal. I think it is on your blog or maybe not but i totally agree that not all parts of the site serve the same purpose so trying to align all KPIs is a mistake.

* To cater to the “i want that too” crowd I end up with multiple charts on the dashboard but highlight and give my perspective on the relevant site section KPI’s only - still a one pager. Not ideal, but i at least get a chance to get my point across without having to fight the “align with the strategy” mantra.

* Also, have taken to removing the scale and just presenting a trend with superimposed Last year trend.

Just some thoughts.

Thx for the insights.

The problem that I see with it being insight heavy, is how often is that going to change over time? If the meat of the dashboard doesn’t change often, how engaged are they going to be in that dashboard as time goes on if those insights change infrequently?

As such the biggest challenge is often to reduce the number of KPIs and to provide “thicker” interpretations.

Having said that, however, I think we should also avoid the opposite mistake of throwing away too much information. In my view, it is perfectly fine to include several charts/tables in a dashboard as long as they all serve to “explain” your main KPI.

In your dashboard, for example, one could have shown the total cart abandonment rate in the upper left quadrant and then used the remaining three to provide some explanations. One way of doing so would be to break down the total abandonment rate to your four key customer segments. Another way would be to show which traffic sources or campaigns had contributed most to the positive/negative trend for this overall metric.

The idea would be to build a single display which shows not only how well the check-out process is performing, but also why this might be the case.

Thanks again for an interesting post!