A cannabis-growing firm recently had to issue an embarrassing correction to the financial authorities for underreporting their losses. Reason: an error in a spreadsheet. They’re not alone: for instance, the “theory” behind the Tory government’s woeful austerity programme was based on a spreadsheet error. Errors in spreadsheets are rife: indeed, a study from the University of Hawaii found that 88% of audited spreadsheets have an error.
At General Products Ltd, we develop consonance.app, the leading publishing management system. In professional, modern software development, it’s a given that when we write code, we test it. Each piece of code has a corresponding test that ensures the code works now, displaying and calculating what it should — and, crucially, continues to work in the future when other changes are made. Every time new code is written and deployed, all the thousands of previous tests run, providing assurance that none of the new code has broken anything. Spreadsheets, however, are like the most unreliable code: they have no tests. And no one wants their business to depend on untested, unreliable code.
Another problem: spreadsheets don’t enforce validation. Good programmers write validations to restrict the data that can be stored in a particular field. But spreadsheets, without validations, allow for harmful duplication. Whilst on-boarding new clients to Consonance, we’ve seen some doozies in old unvalidated spreadsheets. Take contact information. We once saw an author’s name spelled 17 different ways. The name was something like Professor J.D. Pope; the spreadsheet contained entries such as “Pope, JD”, “Prof Pope”, “Professor J Pope”, “JD Pope, Prof”, “Prof J.D. Pope”, and on, and on. It made it impossible to know which books were written by which real-life person, and which royalty payments had been made. And it’s why identifiers such as the ISNI (ISBN13s, but for people and organisations) and ORCID have taken off in popularity, particularly in academic publishing where academics’ careers depend on publishing credits.
The danger is even more insidious than mere error and duplication. In an increasingly-complex publishing world, “one version of the truth” of data is a baseline requirement. The moment you put data into a spreadsheet, though, it’s prone to be emailed to eight people, who save it to eight desktops, and edit it eight times each, often, in the case of P&Ls, to get cells to reflect the number a person wants. And if any or all of them are printed out on A3, further scribbled edits are too tantalising to avoid.
Before you know it, there’s real confusion about what the price is, what the revised manuscript delivery date is, what the pub date is — because there’s the system version, and the Printed-Out-Colourful-A3-Spreadsheet-From-Someone’s-Desktop-From-Tuesday version. Add that to the PostIt-Note-on-So-and-So’s-Monitor version, the AI version, the Nielsen version, and the Second-to-Last-Email-in-the-24-Email-Thread version, and it’s a recipe for disaster.
Not only that, but publishers have let themselves become used to reading spreadsheets, nasty, denormalised things that they are, that lead to monstrosities like the ghettoisation of ebooks, when one row contains information about the paperback but also, as an afterthought, the ePub ISBN13. As Edward Tufte wrote in his seminal The Visual Display of Quantitative Information, “give to the viewer the greatest number of ideas in the shortest time with the least ink in the smallest space.” Spreadsheets trample over this notion. A beautifully-laid out advance information sheet is much more readable and interpretable than a 42-column row on a spreadsheet. Publishing deserves its software to be able to present our information in the format that’s right for the viewer, whether computer or human: sometimes as XML, sometimes as an API endpoint, sometimes as an iCal feed to a calendar app, sometimes as a Slack notification, sometimes as a beautiful fixed-format PDF, sometimes as a re-flowable ePub.
So this is a rallying cry to confirm your likely hunch: spreadsheets can be a dangerous way to run a business, and the wise publisher will seek to mitigate the risk they pose. If you aren’t at the stage where you need a purpose-built system, you must build a standard into your spreadsheet and stick to it.
Are your current systems sabotaging your growth ambitions? Are you hungry to implement new business models, but concerned you lack the strong administrative foundations needed for innovation?
We're always amazed at how resigned publishers have had to become to the low bar in publishing management systems. Demand more.