We're Having A Party
Timing: 30 mins
Ingredients:
* At least 10 pages (8.5’ X 11’) per participant
* 1 marker per participant
* Stickers
Directions:
We are having a party, and we need to enlist everybody in the room to create invitation cards (3 per person). Begin by showing an example of what the finished card should look like:
- fold page in half
- draw a happy face on the front,
- write a message on the inside,
- sign the card,
- stamp the back (sticker) and,
- mail the card by dropping it in a box.
Once everybody is comfortable with all of the steps, start the timer and have participants build 3 cards each by completing each step to completion before moving on to the next step; this is known as batch & queue. Stop production about half way through and ask everybody what would happen if we decided to change the color of the paper. How much wasted effort would there be? How does this map to software? Let production continue and note the time when the first card is delivered to the customer and again when all cards are complete. Run the process again. This time, have participants complete a card before moving on to the next; this is known as single piece flow. Again, stop production about half way through and ask the same questions as before. Let production continue and compare the times with the first method. Obviously, the second method is much faster at getting something to the customer, but more surprisingly, the second method is also faster over all. Discuss why this is; if the participants say that it is because they have become more efficient, then run it again with the first method and challenge them to beat their time.
Learning Points:
- By taking a smaller set of requirements all the way to completion, you get something to the customer faster. Conversely, if all the requirements are processed at the same time, changes later in the cycle become more costly.
- Single piece flow is often faster then batch and queue. This is due to the fact that each cross-functional participant can take ownership of a module all the way to completion, reducing overall task-switching and hand-offs.