Seven steps for a great ROI Model

If you are selling a business software solution, a good ROI model has always been an important part of the sales process.  It allows your customer to build a business case for buying your software. It is even more important now than ever before, when the budgets are tight and fewer projects are getting funded.  I have built several ROI tools for my clients in my 15+ years as a marketing consultant to software companies.  Here are some important tips:

1. Design your ROI model for use by pre-sales engineers in sales cycles.  This allows you to build a business case in nearly every deal, using the data they have gathered during the discovery process. 
2. Ensure the model can be driven with inputs that are not too difficult for customers to provide. For many customers, their current process may be manual and hence they may not have too many metrics around their process.  If you ask for a lot of details, you may not be able to get enough information from them to calculate ROI.  On the other hand, if your model is driven by very basic information, it may look too trivial or generic, and your customer may not take it's output seriously.  It is very important to have the right balance – I typically analyze several pre-sales discovery notes, speak to a couple of pre-sales engineers and post-sales consultants and interview 4 to 5 customers to understand the right balance before creating a ROI model for my clients.
3. When you are asking the user to enter current metrics in the ROI model, have the field default to industry best practices data (i.e. what you have seen at your other customers), which can be overridden by customer.   This ensures that if the customer cannot provide the right metric you are looking for in their current manual process or the improvement they expect to make by implementing your technology in their environment, you have a valid default number to plug in (that will be acceptable to the customer).
4. I have always seen less push back from customers, when you use a range (a high number and a low number) for either current metrics or for capturing future improvements from your technology, rather than a single number.  
5. The end goal is to quantify the inefficiencies from manual process, as well as capture the value of the benefits (revenue increase or cost reductions) from automated process, and the investments required, so you can use this data to show Net Present Value of savings from your technology.
6. Since the ROI model is in Excel, what I have seen works is the first tab contains all inputs, the next set of tabs contain costs and savings (or revenue increase opportunities), one tab for each process improvement opportunity area.  The last tab of the model contains a chart showing $$ value of savings over a 3-year period, as well as NPV.  Such a layout is easy to understand and also at a glance highlights the various benefits from your software.
7. Never send the ROI model to your customer to fill in the data – have your sales engineer work with the customer to create the ROI for their environment

A good ROI model can accelerate your sales cycles and help you improve your win rates.  Don't short change the creation process, but at the same time, don't make it too complex.