When I have to work with a group of stakeholders to help them prioritize features I've had great success with the dot voting technique which really fosters engaged discussions. I've found this approach to be really useful when dealing with stakeholders who have a lot of ideas. It helps focus the conversation and surface the features that the group believes are the most important. Scheduling two sessions - one for discussion and one for voting - makes the process more efficient, and the tangible element of the voting session makes it easier to understand the stakeholders' perspectives.
Here is a summary of the dot voting process from Wikipedia
The dot-voting process includes the following steps:
- Participants are each given a set number of dot stickers (as decided by the facilitator)
- They place dot stickers next to options presented that they like (they may place any number of their dots on any number of the options)
- Options with the most dots at the end of voting “win”
Variations Include:
- Using different color dots to signify different values, e.g. green for "like" and red for "dislike".
- Using different colour dots for different types of participants e.g. blue for management and red for staff.
- Uote with negative dots (resistance votes) to find the highest group acceptance. Options with the least dots at the end of voting "win".
Dot voting for product features
I've found this useful in enterprise products where the stakeholders usually have a lot of ideas about what they want. Too many ideas! To focus the conversation and surface the features that are important, free ranging discussions can be unproductive. Everyone has different thoughts, but they are hard to capture.
What I like to do is schedule two sessions. In the first session I go over all their suggestions that have been collected during interviews and discuss each one for 2-3 minutes. It's important to have as many stakeholders as possible participating!
During this meeting some ideas will get rejected by the group immediately and some new ideas will be generated. Encourage an exploration of different points of view while being careful not to devolve into a 30-minute discussion about an y single feature.
The purpose of the first discussion is to make sure that everyone involved understands what each feature encompasses. This will make the second session go smoothly.
The second session is for dot voting. In the voting session every participant votes for the features they find most valuable by placing one of their stickers on the feature. When I used to be in the office, I liked to print a large headline or short summary on the top half of a sheet of paper and a 1-2 paragraph description on the bottom half. After folding each sheet in half I'd gather everyone in a conference room and lay the sheets on a table. Every participant is given a certain number of stickers to use as votes and told to vote for the features that they think are important.
Folding the papers in half makes the voting sheets take up less space. Printing the summaries in large type makes them easy to identify from a distance and having the descriptions available for quick reference helps eliminate confusion that stakeholders may have about the feature.
Getting everyone around the table and moving around fosters discussions that provide a lot of insight into how they are thinking about each feature. I like to have help from other PMs to take notes of things that are overheard during this meeting. Having the features physically represented on paper allows the participants to move them, group them, hand them to each other for discussion, and makes the process more tangible. I find it useful to slowly move the features that receive no votes to a different part of the table making them easier for everyone to quickly check for favorites that got overlooked.
Since going remote I've recreated this in Miro/Zoom. I don't find that the discussions are as lively and insightful, so I've modified my approach to be more interactive. For example, if I see someone vote for a feature, I might ask someone else a question about the feature. It takes more input from me, but the end results are just as useful.
Once the voting has completed, I stack rank the papers in order after removing those that did not get any votes. Once all the features are ranked the conversation can naturally progress to a discussion about timelines.