Before we look at why data analytics is not a spreadsheet, it is perhaps a good idea to preface this post with saying that we rather like spreadsheets. They can provide good insight without too much effort. Plus, for many straightforward tasks they are a quick, logical and good choice. So this post is not going to be a wholesale rubbishing of spreadsheets – they are very useful.
Having said that, spreadsheets are too often seen as the only solution to tackling data analytics. Deloitte published the results of a survey last year which found nearly 2 in 3 organisations still rely on spreadsheets for insight into their data. This is arguably a historical thing as people have acquired the basic skills to create and use spreadsheets. To this end they are of course very useful. But this tends to make them a default option for any data analytics work – and this is where organisations come a little unstuck. Here’s five reasons why.
Spreadsheets Take Time To Create….
Even for seasoned veterans, building to undertake data analytics can take a long time. It’s a manual process, sometimes cell-by-cell, making the work both repetitive and liable to error. And while some of the heavy lifting can be automated it still requires detailed, repeated checks to avoid slip-ups to validate individual and grouped formulas that have been inserted. The work can also be complicated by ad-hoc changes having a knock-on effect on other areas, meaning formulas have to be re-checked.
…And They Break
Trying to maintain a spreadsheet can rapidly turn into a strange variation of the game of Jenga. Once the spreadsheet has been created it works for the data within it. Changes to the spreadsheet structure or volumes of data can impact on the formulas within it. Just inserting a row or column can result in errors within any analysis. Even sharing with colleagues or partners can be fraught with problems due to differences in applications versions in use.
Spreadsheets Don’t Scale
We just mentioned that spreadsheets can struggle when altered, and this is very true when they are asked to scale up. When adding a single column can throw an entire spreadsheet’s off-kilter, asking it to accommodate multiple additional insertions becomes a long slog to keep the reporting accurate and true. Not only that, but as they grow their performance can degrade too. A spreadsheet which is slow to access or navigate is a source of some pain, both for the owner and any users.
They Are A Fiddler’s Paradise
Because staff often have some experience in using spreadsheets there is often the temptation to try and make changes. This interference is usually well-intentioned but can have disastrous results. It means a spreadsheet can often end up in multiple variations, or (if not locked) no longer functioning correctly.
Spreadsheets Are Better As A Visual Tool
As a final point to why data analytics is not a spreadsheet, one of the great strengths about spreadsheets is how they present you the information up front. A clear and logical spreadsheet can be read it quickly, and that speed is vital. Any calculation and formula stuff is in the behind the data.
None of the above says spreadsheets are no longer of any use – of course they are. But the volume of data created by a typical organisation has grown and evolved. Gaining true insight into that data requires a more powerful and flexible approach, which is where tools such as AI come to the fore. Any initial upfront work involved to program up a way to analyse data is saved many times over from repeated use, scaling up when needed to handle larger volumes.
So, if you’re still relying on spreadsheets to handle your data analytics, perhaps it’s time to ask yourself how much time and resource they really take up to create and maintain, and whether it’s a good point to start looking into how you can gain much better insight into your business data – faster and with more confidence.