Scene of the Crime
The last scene of this video has been repeated, over and over throughout my career. Several people huddled over a laptop with someone (usually me) banging out spreadsheet formulas trying to understand a bunch of summary level data, and reconcile it to other people’s numbers. This reality has been created by the fact that we are still using spreadsheet technology that has largely ignored the needs of the business user.
<iframe width=”560″ height=”315″ frameborder=”0″ src=”https://www.youtube.com/embed/5wM28vOe4-A”></iframe>
Before you say, “Here we go, another spreadsheet basher standing on a soapbox”, know that I built my career on using spreadsheets to streamline manually intensive processes and make people’s lives better. My favorite story is when I streamlined an operational accounting effort that took two people five days to complete…every month. The end result? The woman who had been running this manual effort for 10 years gave me a big hug and thanked me because it turns out she had been delaying her retirement as she took so much ownership in the work and didn’t trust anyone else to do it right.
That was 20 years ago.
The unfortunate truth, however, is that a vast majority of business men and women still rely on the geriatric medium to actually run their business. Let me present the facts as to why it’s time to incarcerate the spreadsheet, and spreadsheet add-on’s that put lipstick on a pig:
Data is sourced from many systems
Joining disparate data sets to get the whole story is the only way to get true insights and provide the organization with actionable insights.
- A top US regional bank recently delivered an analytics application that had 43 disparate data sources.
- A global leasing organization solved a reporting nightmare by combining seven systems across the globe transacting in 24 different currencies.
Organizational change is constant
Traditional query tools have worked around the problem for decades by managing hierarchies and cubes which allegedly represent how an organization is managed. The complexity introduced by a simple organizational change, however, results in a backlog of IT work and frustration by the business.
- Recently, a customer built an analytics application sourced from an Oracle general ledger and addressed eight months of backlogged projects in less than a week. This effort included creating a dynamic hierarchy table that allowed the analyst to adjust/change the hierarchy on the fly.
- Another common practice is to add assumption tables to the data model that have a ‘Start Date’ and ‘End Date’. This allows an organization to address the fact that assumptions change and reporting and analytics need to be able to monitor that change in real time.
Drilling to details is always necessary.
Pre-aggregation of data in order to reduce the number of rows was a technique developed decades ago to allow for the fact that historically, there were simply too many rows of data to handle. This is the number one reason we create massive spreadsheets and store this detail outside of IT.
- Nearly five years ago I showed my first Qlik application to my CEO and his first question was “Can I drill to details”? Of course the answer was yes.
- On a recent customer call we joined an 800,000 customer file with 8 million transactions and in less than 30 minutes. By having all of the transactions made available to the analyst (and eventually the entire organization) a transformation out of the spreadsheet culture will quickly evolve.
Be wary of visualization tools that simply expand/extend the spreadsheet problem and don’t have the capacity to eradicate it. Visualization alone is insufficient.
Many people incorrectly place Qlik in the Visualization part of the above spectrum. Our 3,500 Financial Services customers recognize the value of Qlik is having the option to use our analytics platform at any (or all) of the points on the spectrum.