Blog: Claudia Imhoff Subscribe to this blog's RSS feed!

Claudia Imhoff

Welcome to my blog.

This is another means for me to communicate, educate and participate within the Business Intelligence industry. It is a perfect forum for airing opinions, thoughts, vendor and client updates, problems and questions. To maximize the blog's value, it must be a participative venue. This means I will look forward to hearing from you often, since your input is vital to the blog's success. All I ask is that you treat me, the blog, and everyone who uses it with respect.

So...check it out every week to see what is new and exciting in our ever changing BI world.

About the author >

A thought leader, visionary, and practitioner, Claudia Imhoff, Ph.D., is an internationally recognized expert on analytics, business intelligence, and the architectures to support these initiatives. Dr. Imhoff has co-authored five books on these subjects and writes articles (totaling more than 150) for technical and business magazines.

She is also the Founder of the Boulder BI Brain Trust, a consortium of independent analysts and consultants (www.BBBT.us). You can follow them on Twitter at #BBBT

Editor's Note:
More articles and resources are available in Claudia's BeyeNETWORK Expert Channel. Be sure to visit today!

I read an article recently titled "Dashboards: The Key to Breaking the Dependency on BI" by Steve Ricketts. In it, he claims that you don't need to build data warehouses -- they are complex and hard to implement. All you really need is a dashboard to present "vital information in an at-a-glance graphical format..."

Oh puleeeeaaassse - spare me from this overly naive, simplistic and unfortunately erroneous piffle.

First off, let me state that Mr. Ricketts is the VP of Marketing for Bowstreet, a dashboard vendor. That should be enough to cast doubt on his article but I will continue...

While I certainly see a role for dashboards in today's BI environment -- they make very good presentation tools -- they do not solve the myriad of problems with the data they present. He compounds his errors by stating that dashboards solve the problem of IT stovepipes by using their standardization of the underlying application and portal server architecture as well as the dashboard user's interface. No argument there but this assumes that the data is already in a form that is integrated, cleaned up, and easily accessible. In my years of working with all sorts of industries, governments, non-profits, etc., I have NEVER seen one that had its data in such pristine shape. Mr. Ricketts must live in a world that is foreign to me -- and most other organizations.

What the article promotes is nothing short of putting lipstick on a pig. It annoys the pig and doesn't make it prettier. By simply throwing a dashboard on top of disintegrated, unimproved data, you are giving everyone a quick way to view chaos. And chaos is not good for decision making.

To get the most out of your dashboard technology, you must have the data integrated and of good quality. This is not easy. It is not simple. It is not quick. Nothing worth anything really is. But you also have something that is useful. Put your dashboard on top of that, Mr. Ricketts, and you really have something.

Yours in BI success,

Claudia


Posted November 29, 2005 9:20 AM
Permalink | 12 Comments |

12 Comments

I agree. A dashboard merely surfaces the data model beneath it. Garbage-in/garbage-out.

This is a great post - thanks Claudia! I was astonished at Steve's article. Even quite small enterprises these days often have quite complex data integration just to handle such simple issues as international formatting, never mind any attempt to disambiguate or consolidate semantics across multiple systems. Dashboards simply do not help in these cases.

Steve's proposal - to the extent that I understand it - requires executives to be experts in their data, not just in their business. The executive using the dashboard has to not only understand her business needs, but also the semantics, lineage, integrity, accuracy and latency of the data behind her dashboard.

Personally when i look at the dashboard in my car, I want to know how fast I'm going. A self-service portal offering me the pulse rate of the distance sensor at the transmission tailshaft, the gear reduction, the differential drive ratio and the tyre diameter is not going to help me much.

All of this may be the right information - but I am not qualified to validate it. It may be in the right hands - but my job is driving the car, not designing or servicing it. It may even be available at the right time - but if I have to construct the version of the truth that I need, ad-hoc, whenever I need it, I'll likely spend more time on that than on doing my real job.

Sooner or later, I'll take my eye off the road and careen into the ditch of lost opportunities, or I'll see the blue flashing lights of the auditors in my rear view mirror as they pull me over for non-compliance.

donald

I was surprised at the article, but not the sentiment. My last tour through a tradeshow floor turned up several dashboard vendors who marketed around the concept that "A dashboard means no data infrastructure". Yes, and you can knit with one needle.

I think they're desperate for differentiation. Many of the products were of a me-too nature, and the literature reminded me of the Excel-versus-Lotus feature wars.

I'm impressed that you made it through the article. I skip them if the first few paragraphs stray into the land of the marble deficient.

Thanks to all of you for your comments but especially to Donald Farmer. You put it far more eloquently than I ever could. Maybe you could start writing for me?

Just keep blogging!

If there is no data warehouse where does Steve suggest we get the data from? From the transactional databases?
But then we still have to integrate databases to get an integrated view of the business.
May be he is thinking of datamarts? if so that could work.

Building scoreboard/metrics apps based on a dimensional data mart/data warehouse offers greater benefit/advantages, however, is not an absolute necessity. For example, Cognos Metrics Manager can create a metrics extract from an .iqd file that is a SQL file connected to a relational database, not necessarily a data mart/data warehouse.

From this perspective, the arguments by Steve Ricketts, in my humble opinion, are still valid.

Please correct me if my understanding was wrong.

Another goof example is a dashboard/scorecard/metrics app can be created using Cognos Visualizer by using an .iqd file alone.

Note that Dashboards/Scorecards are a different kind of animals than Query/OLAP Reporting Tools.

Trust IBM can't be wrong acquiring Bowstreet, a dashboard/portal vendor... hopefully;-)

I am one of the team members in a portal project, we are linking to several relational databases (not dimensional) to expose data to the end users. We have not gone through the exercise of building a data warehouse and it is working OK.
What are the challenges that you think we will be facing if we do not use a data warehouse?
Thanks
Fatima

Fatima,

As long as your operational systems are in sync with each other (data fields are the same, content is the same, IDs map across perfectly, etc.), then you are fine. It is a rare company though that can say this. Most often, the operational systems were built in complete isolation from each other, meaning that IDs are completely different, codes are different , formats do not match, and redundancy in data appears in spades. Unless you have a robust mechanism to integrate this data together (i.e., some form of ETL processing, data quality processing, and a place to store the integrated, cleaned up data), then you will very likely be presenting misleading or even erroneous data through your portal to the business users.

If you have other questions, please post them.

Claudia

Totally agreed with Claudia.

Most of the times a data mart/warehouse is the first step absolutely necessary.

Sometimes it depends.

We think visually. Dashboards are just the visual representation of data - the content. Hence rightly said Content is King while the GUI is queen.

Leave a comment

    
   VISIT MY EXPERT CHANNEL

Search this blog
Categories ›
Archives ›
Recent Entries ›