What a Hancock-up: Excel spreadsheet blunder blamed after England under-reports 16,000 COVID-19 cases; All amid a second wave of the bio-nasty

In circa 1995, I developed a model in Excel to forecast input costs of various components used in a PC. As a sales guy, my objective was to tighten my costing and, hence, pricing so that I could improve my win rates on government tenders. I used ODBC on my Excel worksheet to fetch historical cost data from a mainframe-based inventory management system. I did have to spend an hour or two to learn lookup, pivot table, and other advanced features of Excel but I didn’t – anyway couldn’t – write a single line of code. It took me a week to develop and test the spreadsheet. When it was ready, it worked as advertised. Our win rate on government tenders went up, which was precisely the purpose of the model. My Excel model became the talk of the company and many others used it for similar purposes in their own LOBs.

Over the subsequent 25 years, I’ve had the need for a similar forecasting model a few times, including as recently as last month. Everytime, I’ve gone to techies as the first port of call to get it developed the “proper way” i.e. with some programming language on some database. Everytime, I’ve been told it will take “awhile”, which is techie-speak for “get lost” to an exec whom they can’t rebuff bluntly. If I had to redo such a forecasting model on a tight deadline today, I’d still use Excel. Yeah I know it’s 2020 but c’est la vie. Excel is the oldest “no-code” platform I know!

It’s not only me. Per anecdata, 70% of Fortune 500 companies have ERP but 90% of them use Excel to prepare reports for their Board of Directors!

Excel is for amateurs. To properly screw things up, those same amateurs need a copy of Access. Beware the wrong tool in the wrong hands 

Why single out Access? Bugs happen in Inventory – and lot of other places – even in SQL Server, Oracle, HANA, etc.

An ERP vendor that used SQL Server found that the software failed to update the stock status automatically after material was issued from the warehouse. Engineering said it would take a couple of quarters to fix the problem. Obviously Sales couldn’t stop for six months, so it devised its own workaround. During demos to prospective customers, the presales consultant would raise a Delivery Challan on the system and shout “We have raised a Delivery Challan for 30 units” twice or thrice. More than a few prospects used to wonder why the presales consultant was so excited about such a mundane transaction. Little did they know that his or her voice had to carry over to the next room where a developer would run a SQL Query to manually update the stock by 30 units!

Coming to SAP HANA, there’s this well documented horror story where German retailer Lidl wrote off €500M because the database couldn’t handle a change in Inventory Valuation from Selling Price (SAP’s standard) to Purchase Price (Lidl’s requirement).