Search Engine Heuristic
As the resident "computer geek" of my family, I am often asked to help solve problems on everything from Christmas shopping to driving directions by "looking it up on the Internet". Of course, for me, that almost always entails firing up Firefox and finding out what Google can tell me about the topic. In some cases, the person asking me to find information has already tried Google so I'll often think of different text strings to enter and end up getting the answer when they could not.
I have found that having searched so often on so many topics, I have started to think about other problems in terms of search engine strings. I've recently become of aware of this and have decided to see if I can apply it to testing. If finding defects is really just a giant search algorithm, then I should be able to fine tune this technique to ask the right questions when testing.
Am I using the right terms when searching for defects?
What am I really looking for? (no really) looking for?
What if I change the terms I'm using? (analogous to changing test variables)
Example Question: "Hey, I saw a this huge dog the other day with floppy ears. Do you know what breed it was?" The search might be something like "Dog breeds, large, floppy ears". For testing an application I might ask myself, "How can I find out if our web interface works the same as the Windows app for new features?" The test might include "[Feature], web, Windows app, [variables]" and then I can build around those terms. In exploratory testing, this type of thinking has also helped me in writing charters.
Maybe I haven't formed this into a completely workabale analogy yet, but thinking in search engine terms has allowed me to better clarify what I'm really trying to find in an application and how to fine tune my tests to find it.