Filter and prioritize ideas

Filter ideas

Filter and clean the list by sorting the list into three piles: obviously useful ides, ideas to discuss, and ideas to discard. You'll find silly things and duplicates obviously arrive in the pile to discard. To filter:

1. Label three areas on your table: keep, discuss, and discard.
2. Split the deck of ideas into two or three piles
3. Hand a pile to a workshop participant
4. Participants start placing the pile into one of the three areas. Place the cards such that everyone else can see them. If someone disagrees with the placement, they're free to move the card to another pile.
5. When all cards are placed, bundle up the "discard" pile and toss it in the trash.

Use simple prioritization on lists of ideas

With a list of ideas written on cards, it's easy to begin prioritization. Select an area on a worktable to place the cards while you're establishing their priority.

1. Label one area "high priority" - usually far left of top
2. Label another are "low priority" - usually far right or bottom
3. Split the deck of ideas into 2 or 3 piles
4. Hand a pile to a workshop participant
5. Participants start placing the ideas on the table where they believe the priority should be - toward the high end if it's high, toward the low end if it's low.

This can work well because it's often easier to prioritize an idea against one or two ideas that it's similar to. Laying the ideas out spacially allows us to easily find and place an idea near ideas of relative importance.

From here it's easy to begin detailed discussions about ideas and priority. It's easy to change our mind and slide a card to a different place on the table. Shuffling prioritization

Alternatively if the list ideas isn't too long allow each individual to prioritize them.

1. Starting with ideas written on index cards, combine them into a single stack.
2. Give the stack to a first participant, and ask her to arrange the stack by priority. The first participant has the hardest job if initially setting priority.
3. When they're complete, ask them to hand the deck to the next participant. This next participant can change the priority as they see fit.
4. Continue till each participant has shuffled the deck at least once. You may need to pass the cards around two or more times till priority settles.

Democratic prioritization

Oftentimes it's not critical that we understand where all ideas fit in priority but rather understand what the top three or four ideas to focus on are. Sometimes you've already spent time arranging index cards into a model that represents your understanding of relationships between ideas and rearranging the ideas into priority order at that point will break the model showing these relationships.

Use democratic prioritization to allow collaborators to vote for the highest priority ideas.

Arrange index cards on a table in an arrangement that that makes it easy to find information. You may already have a card arrangement such as a role model, task model, or any ad hoc affinity diagram that you're working with.

Give each member a number of tokens to vote with and ask collaborators to place votes on the ideas they think are most important. They may place multiple tokens on an idea. As with all democratic processes participants may try to persuade each other to vote differently - that's good. The conversation that occurs helps deepen everyone's understanding of the context these ideas are prioritized in.

For tokens: stickers work okay, or simply making marks on index cards that you vote for. But, I prefer physical objects that are easy to move so that collaborators can change their minds or coerce each other to vote a certain way. The wrapped pieces of candy mentioned as useful in your toolkit work well as voting tokens.

The number of votes to give each member is tricky. Each member should have one or two votes less than they wished they had - to really force them to think hard about what they believe is important. The number of votes is a function of the number of participants, and the number of ideas in model.

For larger models with 50-60 cards built by 3-4 collaborators four votes per collaborator might be in order.

For smaller models with 10-20 cards, a common formula is divide the number of cards by three rounding up, and award that many cards. For instance a model of 14 ideas might allow each participant 5 votes.

Generally 2-5 votes are appropriate for most circumstances. Make sure everyone has at least two votes.

When all votes are placed, total the number of votes on each index card that received votes. You'll find that most arrangements of index cards have 2-3 ideas that are clear winners, and 6 or more that receive votes. You'll find your model likely has "hot-spots" - areas where lots if high priority ideas group together.

Before removing any voting tokens, mark the cards with a symbol such as a star for each vote, four votes get four stars. Use a bright color of pen ink to make it easy to identify the index cards with votes. It's easy in a model marked this way to identify the high ranking ideas, and distinguish the highest ranking ideas from lower ranking ideas and unranked ideas.

Secret ballot democratic prioritization

Some prefer that participants not influence each other's votes. Use the approach above to determine the number of votes for each participant, but ask each participant to write down their votes on an index card first. Then, when everyone has written down the ideas they intend to vote for, ask them to place their votes on the model silently.

Always set context for prioritization and filtering

What ideas are filtered in or out and what ideas are high priority always requires some context or criteria from which to base decisions on.

For example if we were prioritizing features for a product you might set context by asking "what features are necessary to make this product successful?" You might also ask "What features are necessary to make this product minimally useful?" or "What features are necessary to make next quarter's release most successful?" Each of those questions might result in a slightly different priority.

If you avoid setting context, each participant will prioritize based on their own assumed context. When heated discussions about priority occur, often the root of the conflict is based around participants setting priority for different contexts.

It helps to write the prioritization context down on a whiteboard or poster paper where everyone can see it.