Are software professionals smarter than other business professionals? At times, we seem to think we are but let’s examine that premise more closely.
Consider this. National restaurant chains spend millions of dollars test marketing new menu items. For example, Burger King, McDonald’s, Dunkin’ Donuts, Olive Garden, and Taco Bell invest boatloads of time and money trying out menu ideas before deciding to make the items available throughout the country. Here is a brief sampling of menu items and related services that have been test marketed recently:
- Burger King – home delivery
- Dunkin’ Donuts – Glazed Donut Breakfast Sandwich
- McDonald’s – McWrap sandwich
- Olive Garden – small plates menu
- Taco Bell – Waffle Taco
They often try out ideas in test markets, limiting availability to a small number of restaurants. Of course, there’s a problem with that approach — just because something sells in one area of the country doesn’t mean it will sell in all areas. So there are other techniques they employ. In fact local restaurants also engage in test marketing using these techniques. For example:
- Daily Specials – These menu items may simply be existing ones offered at a reduced price. They may also be new ideas that the restaurant wants to try out. If a special item has broad appeal, it will be offered as a regular item.
- Limited-Time Offerings – A new menu item is offered with the stipulation that it won’t be available for long. If it’s not popular, it gets dropped. If it is popular, it returns as a regular item.
Restaurant professionals have learned that they need to engage customers in extensive test marketing to identify menu items that sell well. No amount of laboratory testing or focus-group sampling provides the kind of real-world feedback they need. It’s much easier to tweak a menu item in a test market or a trial run before recipes and menus are finalized.
Should the software business operate similarly? We often seem to feel that we can develop software in isolation and deploy it to everyone without obtaining some kind of real-world feedback first. Why is that? Are we smarter? (I don’t know about you but I’m not.)
Agile concepts like the minimum viable product (or minimum marketable product), beta testing, and frequent deployments exist to help us get the software right. We can never be sure of how the user community will react to new or upgraded software. By soliciting a reaction early in the development process, we reduce risk and end up with better software systems.
We don’t have to be smarter. We simply need to work smarter.