Higher Reliability

Errors in spreadsheets appear to be frequent and are potentially costly.

• 94% of spreadsheets deployed in the field contain errors, when averaged over several studies
• 5.2% of cells in unaudited spreadsheets contain errors

Source: A Critical Review of the Literature on Spreadsheet Errors, by Stephen G. Powell, Kenneth R. Baker, and Barry Lawson, Tuck School of Business, Dartmouth College, December 1, 2007.

• Uses readable symbolic variables and formulas that reduce errors
  • Example: profit = revenue expense not C11 = D52 - Sheet1!E20
  
  
• Uses fewer formulas – often 1% - 10% as many
  • Each formula applies to part or all of a variable table
  
  
• Re-usable structures that capture model logic
  • Dimensions (e.g. a list of products) and time series
  • Vastly fewer manual operations, flawless automation of many tasks
  
  
• ModelSheet automates manual cell-level operations
  • Each formula applies automatically to all or part of a variable table
  • Eliminates whole classes of errors, such as copying formulas over too many/too few columns
  • Automates worksheet layout, after you specify general layout

This view shows the variable editor for revenue. One formula listed below the table generates all values in the table. The formula tells you in English what it means. Pink cells (really one logical cell in this case representing the entire table) contain independent data or formulas. Values in white cells are determined by data and formulas in pink cells.

This is the formula view in Excel for the revenue table, where revenue = (1- discount %) x list price x sales units. Each cell has its own formula. The formulas are expressed in cell addresses so their meaning is not obvious.