Is ROI for an Agile project different?
I read a question in a linkedin group wondering how to measure the ROI of an Agile software project. Business value (ROI) is a basic ingredient of the Agile approach yet Agile does not directly attempt to give guidance on how to measure that value. That is probably because business value has nothing to do with the methodology used to create the project. Business value of software is likely to be different by the end of a project than what you envisioned at the start. I’ve been doing custom software projects since 1987 and have NEVER seen one, even a 4 to 6 week project, where everything is the same at the end that it was envisioned at the beginning. We used waterfall, more or less, for most of my career but we ALWAYS left room for discussion and change throughout the project. One thing that Agile does is make it easier to start delivering the value to the business sooner. Instead of building the database, then the data layer, then the app layer and finally the GUI layer; Agile builds a sliver of value complete, one sliver at a time. As soon as the first releasable part of the software is done the business should start recovering ROI. The risk, something CFO’s are very interested in, is spread through the project in small chunks rather than realized all at the end. As the project progresses the team learns more and it is likely that MORE business value will be achieved than was originally planned. That is not to say you should make the original decision hoping for that outcome though. The expected ROI should be sufficient to justify the risk before ever starting.
What should be looked at for ROI? In my opinion, it should start by looking at how the project will enhance the revenue of the company. Either by making it easier for the salespeople to gain new customers or by increasing the “stickiness” of current customers, or increasing customer engagement and revenue per customer. Second it should be measured in lower costs which usually means more work with less increase in staff. Finally it can be measured in decreasing risk of key employee knowledge leaving the company. By this I mean if you have complex processes that are known by only a single office manager or other key employee and you are scared of the consequences of them leaving then you decrease risk by getting those processes into a program.
Every CFO and company will have their own measure of what ROI is worth the risk. In my personal experience programs return full cost between 12 and 36 months. In many cases once the program is complete, small enhancements can return multiples of investment in a fraction of that time.
