The problem with spreadsheets
Read whitepaper - Quantrix & Excel: 3 Key Differences
For many years, the spreadsheet has been the ubiquitous tool for developing business models. Every accountant, financial analyst and business planner has become familiar with the spreadsheet way of working.
However, nearly everyone who has used spreadsheets for business modelling will have encountered problems, as their spreadsheets have reached any significant size and complexity.
At Planning Models, we use Excel for various modelling and analysis tasks. But for models that reach any reasonable size, or for any model that needs to cover multiple business dimensions, we have found that Quantrix Modeler offers significant advantages, in terms of reduced development effort, greater flexibility and a far lower risk of modelling error. Quantrix tackles the following principle limitations of spreadsheets.
Lack of design and layout structure
Perhaps the greatest end-user appeal of the spreadsheet is that it is so flexible - model development can start with little or no preliminary design effort. But the spreadsheet environment provides no real structure for designing and laying out your model. You can get started quickly, but soon find that you need to self-impose some structure and discipline to the layout, as your model spreads into multiple worksheets. Multi-sheet spreadsheet models can quickly get "messy" and difficult to develop further, or to change, when the underlying business requirement changes.
Formula copies everywhere - the risk of model error
Even with layout disciplines in place, spreadsheet errors often occur, due to mistakes in formula copying. Spreadsheet formulae each calculate only one cell. However, nearly every model requires formulae to apply across many columns or many rows. In a spreadsheet, you achieve this by copying the formulae across the columns and rows to which they apply.
This process is not so onerous when you first develop a spreadsheet. But when you have to change the model, it becomes tiresome, as you have to change each logically unique formula and then repeat the copying process to all other cells that need to be updated with the new logic. It is very easy to make a mistake in selecting the formulae to be copied from, to do the copying inconsistently across multiple spreadsheets, or to forget about and overwrite individual cells which have some valid variation from the general formula.
It is this copying and recopying of formulae that lead to many high profile spreadsheet errors. The only alternative is to carry out very disciplined and structured testing. Because each copy of a logically unique formula could have an error, testing of spreadsheets is a very onerous process and is seldom done thoroughly.
Lack of transparency
Due to non-meaningful cell references, model formulae can become difficult to understand. More complex formulae can refer to multiple cells and the cell references are not inherently meaningful. This makes it hard for anyone to understand their own, let alone anyone else's spreadsheets. It is very difficult to change spreadsheet models safely and efficiently. You either have to invest in detailed analysis, documentation and change control processes, with lots of retesting effort, or to abandon the model and start again.
Lack of dimensionality
A spreadsheet basically has two dimensions: rows and columns. But in many models, you want to apply the same model logic across multiple dimensions: product groups, market sectors, business units, distribution channels etc, and to sum across these dimensions. To do this in a spreadsheet is a convoluted process. The modeller can represent a third dimension by copying the model across multiple spreadsheets, (with even more copying and risk of error) or within the row dimension, creating an analysis within extra rows. Beyond a third dimension, things can really get out of hand!
Quantrix Modeler solves these problems
Because it has a multidimensional structure and logic formulae which applies across rows, columns and other dimensions, Quantrix Modeler eliminates these modelling difficulties. Model formulae exist separately from the data, without the need for risky copying. Quantrix formulae generally refer to row or column variable names, so that, for example, a single formula, such as "profit = revenue - costs", applies across all time periods and also across any other business dimensions that need to be modelled, such as product groups, regions, market segments, or channels. All this without copying formulae.
The formulae in Quantrix Modeler refer to the actual business variable names, rather than to obtuse cell references. This makes it much easier to understand. Because there is only one copy of each logically unique formula, models are much easier and faster to change and retest than spreadsheet models.
The payoff is not only in faster, less error-prone model development, but also in much easier and more reliable model changes. Working with spreadsheets, changes in product structure, market groupings or time horizon would mean recopying formulae and carefully checking the model again, Quantrix can handle these changes more easily and much faster, without any formula changes or copying. This radically improves the usefulness of the model for dynamic planning and decision support.
To see a practical introduction to these benefits, we recommend the 30 minute introductory tutorial. Also, we would recommend the following whitepaper:
Quantrix & Excel: 3 Key Differences
Previous page: Purchase Quantrix
Next page: Contact