Tutorial - Creating Seed Ideas

What are seed values:

Seed ideas are ideas that are entered before the collaboration is live for all participants.

Sees ideas are usually entered by the Collaboration creator, but may be entered or suggested by the executive sponsor of the collaboration.

Why create seed values:

The earliest participants will not have a ranking exercise unless open-ended questions have initial seed values
Important to seed 3 to 7 diverse ideas if your audience is a smaller than 20 people so every participant has something to rank
If you have a hypothesis or assumptions to test, use seed ideas to test the hypothesis.
Creating seed values often validates you have a well-worded prompt
- If seed values consistent with prompt are easy to conceive and representative of your expected results, you probably have a well-worded open-ended prompt with adequate context
- If seeds values are difficult to conceive, not clearly articulated, too banal, too general, or are irrelevant to your expected results, refine the prompt accordingly

Examples - Using Seeds to Test a Hypothesis

Example - Resource Planning

A group that facilitates volunteer activities for "green" causes wanted to validate their assumptions about what activities motivated volunteers. We created a collaboration that prompted for activities of interest and seeded the session with the top 5 efforts they had planned plus 2 ‘negative’ responses they assumed were barriers for their volunteers. Participants ranked up a few of the seed ideas. Others seeds were not ranked up. Participants added their areas of interest and, ultimately, the top idea was an unexpected area of focus. After the collaboration, the group knew which activities were of most interest to their audience, and knew exactly how to invest their time and resources to have the greatest impact.

Example - Corporate Meeting

A corporation was spending significant time and money to bring together employees for a live event. They created a collaboration to ask for the most important topics to put on the agenda. They seeded the prompt with their planned agenda topics to see if those ideas would get ranked up or if employees were more interested in different content. Of course, after the collaboration, they were able to refine the agenda, but they also had enough information from this 'pre-work' to delve deeper into specific content, creating more value from the time spent in-person.

Alternatives to creating seeds on your own:

Using a seed generator or chatGPT to get ideas for seeds is encouraged, but insure the resulting seeds are appropriate for your collaboration.
A great alternative is a 'Soft Launch' (see best practices for Soft Launch). In a soft launch, a few trusted collaborators enter initial ideas before the collaboration is open to all participants.

How many seeds:

Creating 3 to 7 seed ideas is encouraged, but having up to 12 is allowed.
Except for Try for Free applications, seeds are encouraged, but not required.
Try for Free applications require at least 3 seeds for each prompt before clicking 'Publish'.

Properties of high-quality seeds:

Specific, Creative response to the prompt
Informs subsequent actions after the collaboration or tests a hypothesis
Concise, clear sentence; neither a generic phrase nor a lengthy compound sentence
Reflect a range of diverse responses
- Consider how the responses of participants with somewhat opposing views might be reflected in the seeds
- You may have ‘negative’ seeds, or ideas that are counter to the desired response, if needed, to spark inspiration or confirm the negative view
- If you have an idea about response distribution, keep the seed ideas proportional to the anticipated results. If not, a good rule of thumb is not to exceed 1 ‘negative’ seed for 3 positive seeds.

Litmus Test to identify a good seed:

Is the seed a relevant, relatable idea that would inform subsequent actions?
Consider: If a seed idea was either not ranked as important by participants OR ranked as Most Important/Most Popular, would that insight provide value to you or indicate a clear action? If so, the seed is likely a good seed.
Example: Planning an Agenda for an All-Hanks meeting
- Banal seeds like "The CEO should speak" may get ranked up as important, but provide less value than seeds that provide more information.
- Being overly specific creates a seed idea that is harder to resonate with participants. "The CFO will present the annual budget, investor outlook, and the new expense reporting process" is probably too specific for a seed.
- Better seeds might be, "The CEO will describe the key challenges we are facing this year" and "The CFO will present key points about our budget and financial outlook". If either is ranked up, that should be considered for the agenda. If not, the topic should be altered, minimized or dropped in favor of topics that were ranked higher.

What if the seeds are not relevant or do not resonate with participants?

Trust that:
- Ideas that resonate with participants will get selected as important and ranked up by participants
- Ideas that do not resonate with participants will not get selected or ranked up
- Participants will enter and rank up ideas that resonate with the group beyond the initial seed ideas

The problem with banal seeds

Banal seeds are very general phrases that are likely true, will resonate with participants, but are not particularly informative. The issue with banal seeds is that they tend to get consistently ranked up in the lower portion of the ranking. While this is not 'wrong', it would be preferred to have that slot either 1) taken up by a more meaningful idea or 2) at least be entered by a live participant. Seeds like "I don't have time" are probably true, but not actionable. A seed like "We have too many other areas of focus to take on this project now" states the cause of the availability issue and identifies a barrier that can be addressed.

Specific consideration for prompts related to Rating Questions:

All responses, including seeds, may be selected for a ranking exercise, regardless of the rating the participant provided
There is only one 'list' of seeds for one prompt. Seeds are not unique to or specifically designed for low-score responses or high-score responses to prompts following a rating question.
Ideas, including responses, will be selected for a ranking exercise somewhat proportionately to how they are entered.

Do seed values influence outcomes?

Seed values influence outcomes just as any other idea submitted by a participant
Ideas that are entered early in the collaboration, seeds or otherwise, may be seen by more participants and have an opportunity to be ranked up more often than ideas entered later
The size of the group will diminish this affect. For a smaller group of participants (under 20), early ideas submission has more of an influence than in larger groups of participants