As our business gradually expands, we ask ourselves why has no-one tried connecting analysts to software?
One reason is that is that the core of our product, the database, is an original invention. Another reason occurred to us the other day as we heard about the progress one of our major clients was making in understanding and configuring their Azquo implementation. Consider the following two groups of people:
- Excel experts. Accountants, auditors, financial directors, analysts etc. These people can do fiendishly complex things in Excel. We know examples of people carrying two laptops with then as some of the workbooks they work on take 20 minutes to load!
- Infrastructure, database administrators or programmers.
What is notable is the lack of crossover. The Excel experts might do a little SQL but generally they are not interested in programming. The techies might do a little Excel but generally they don’t know what Excel is capable of, the very complex things people are doing with it.
This divide is a huge missed opportunity. Let us say we are programming a highly configurable bit of software. Before Azquo, that configuration would be done either in a fixed user interface, which would require code changes to modify, or in a plain text configuration file, a properties file maybe.
Users would be happy with an interface, they might be ok with a text based configuration file if it is not too complex. As soon as it became complex they would not want to use it, it would be back to the techies to manage it.
With the experience of Azquo we would be strongly in favour of a third option: using an Excel workbook. Programmers would initially not consider this as at first glance they would see hassle – more code and library dependencies – combined with little advantage.
There are advantages though.
The first is cultural – the Excel users are comfortable with Excel so even if the configuration options were complex their initial response at being asked to manage such a configuration would be positive. It’s a spreadsheet – they know what it is.
The second is that Excel is powerful, far more powerful than many programmers know. I wonder how many even know how to make a pivot table?
Often selecting an option from a dropdown will affect another dropdown, that can be easily done in Excel and the users can adjust the options if they are not right. That of course is not possible in a flat text file but it’s only the beginning for Excel. Use named ranges to pump data in and out and now you are really talking! An Excel sheet could slice and dice data according to criteria defined entirely by an Excel expert.
The onus is on the techies to make the first step and to use libraries like the excellent POI from Apache. It is not hard and the potential pay off is massive. The Excel experts are often the people who define business requirements. Excel is effectively a way for them to configure business logic in software, there is no need for the tiresome loop with tech. Faffing about with requirement documents, testing finding out it is not quite right etc. They can just tell the software what to do and the workbook itself is the spec.
Connecting Analysts to Software
As we said above, we have begun to see this happen. We believe the potential here for making useful software that saves a lot of drudge work is staggering. It has not happened because of existing professional culture. The techies are the ones who can change it.
Azquo has produced bespoke applications for the insurance market, multi-channel retailers and call centre operators. We like to solve problems with scalable reliable software solutions – so please do let us know what your company finds tricky and we will try to provide answers.
