We have been approached by a large organization whose Excel modelling is a little out of control. Like lots of companies, in their early days, they plugged a few formulas into an Excel worksheet to manage their widgets and all went well. As their business increased, more formulae were added and tweaked and the model grew. Soon it was time to expand onto new worksheets so they can tab around and reduce clutter. Today they have hundreds of Excel workbooks with references between them. Ouch. Now it’s a maintenance nightmare.
Fortunately, they also have a large IT department. They approached them for help to clean things up. Sure, they were told, we can re-implement your model in a database and provide you with a fancy reporting tool. The only problem was: it would take about 18 months to implement. Can you wait until then? They could. They could continue to work on their manual Excel model until then.
But what does that do their business? If they continue adapting their Excel model while the IT build creeps from 18 months to 2 years then what happens when the new system is ready? Do they go back to the old business model and leave their new strategies behind? They can’t. They would not be competitive.
You know where I’m going with this. The waterfall is dead. This company needs an Agile process that delivers incremental versions on a short cycle. An IT team with a Project Owner in the business side needs to start pumping out prototypes every two weeks. Sure, the first ten iterations will be garbage and will not ease the Excel pain, but within a year there will be time saved. Most importantly the development will be in lock-step with the business strategy and ultimately the latest ideas will only be two weeks away.
Now, if there were only a way for IT to build their SQL models from Excel without laborious manual analysis….