Crowd Wise: ongoing learning

Here at Rhizome we’ve recently completed a piece of consultancy work for the Fairtrade Foundation.

We used the Crowd Wise consensus process as a central plank of the work – running five regional Crowd Wise conversations with grassroots fair trade campaigners to develop options around grassroots membership of the Foundation, prioritise the options and then move towards agreement.

Crowd Wise is a relatively new approach, and it’s still developing. This is the first time it’s been used across a series of events. Naturally we’ve had a chat about how that worked and the lessons we learnt.

So what was the main difference about how we used it here? Previous Crowd Wise sessions have been a single event. Usually the options under discussion had been created in advance, but in a few cases the session time had to be used to identify possible options, discuss and develop those options and then reach an agreement. Given that some of these sessions have been quite short, that’s a tall order. But it was also cleaner. One event and it’s done and dusted.

Here we had five events, so we could take our time, and gently nudge our way towards agreement. Alongside Foundation staff, we drew together five options based on the views we’d heard in the first stage of the consultation (through interviews, surveys, meetings and so on). The longer time frame meant that in the first four events we could concentrate on shaping options, merging them and then de-merging them, prioritising and so on, leaving the final event to pull together the threads of the previous four.

That’s not to say that each individual session had no time pressure. We had about 2 to 2.5 hours for each session, and could easily have filled more time.

So it was a more spacious process, but one that left us with some choices to make.

  • For example, we chose to allow the options to evolve between meetings rather than present each group with exactly the same choices to discuss. Often this evolution was subtle – a shift of emphasis within the same rough parameters.
  • We chose to leave one particularly ‘weak’ option in the running. Partly this was in deference to our client, who had proposed this option, to ensure it got a fair hearing. But it also did no harm to leave it there. It was part of the wide package of information that stimulated discussion, thinking and ultimately decision. On the other hand, if circumstances had been only slightly different we might have ‘culled’ it, as a visible sign of moving towards consensus.
  • We chose to involve Foundation staff in the conversations between each event about the evolution of the options. This collaboration had real strength. The Foundation have a much more intimate and connected relationship with the process because of it. We also benefited from their perception of each event – it wasn’t possible for us to hear every conversation. But it did take time – theirs and ours. On balance though it was the right thing to do.
  • We kept all the stakeholders informed through a project blog. After each meeting we blogged about the meeting, the key conversations and the outcomes. We then gave an indication of how we were developing the options for the next event based on what we’d witnessed at that event. This gave the project transparency and integrity. And of course allowed readers to interact and leave comments, some of which gave us very useful feedback on how the process was working from a participant’s perspective. We made useful changes to the information participants were given based on these comments.
  • We chose to do the final event differently. The first four involved small groups exploring one or two options each – strengths and weaknesses and potential for mergers between options. Each group then put the case for the options it had discussed to the full group before more discussion and a final vote (using a consensus voting system). Whereas in the final event each group was asked to consider all options and specifically look for a way forward that took the best from all options.

Most of these choices were intuitive ones. There’s no Crowd Wise rulebook, which is why we want to share these thoughts here as part of the process of developing a body of thought and hopefully best practice.

Other observations and learning

  1. Part of the process in this case was about developing trust – trust in the process and in the wisdom of the group. It’s Crowd Wise after all. We had to trust that each group would pick up, to an extent, where the previous group had left off. And we had to trust that the final group could discern the developing consensus of their peers from the previous events and pull together a solution that fully considered their views. It worked wonderfully.
  2.  The process also confirmed the power of collective process to change individual perspectives for the better. At each event we started with a quick vote on the options before any discussion had taken place. This gave us a benchmark, and gave the group a chance to get a sense of how the consensus voting process worked. It allowed us to clearly see the changes of mind that the group had undergone. For example in several sessions one option ranked highly in the opening vote but came last in the final vote.
  3.  We were reminded about the power of the individual within the collective. There were several examples of individuals making contributions that shaped the process and outcomes as a whole. One person making a distinction between having access to the Foundation’s AGM and access to the Board (where he considered real power to lie) had a significant impact on the conversation at that session and the two that followed. Another participant shared an idea about randomly inviting campaigners to sit on a national committee in order to involve people who would never think to stand for the position. A simple idea, but one that leapt out as useful and found its way into the final report. The Crowd Wise environment seems to nurture this possibility both because it’s based on face-to-face meeting, but also because it explicitly values exploration, innovation and creativity.
  4. We learnt the importance of carefully setting the context. Despite a reasonable amount of background material being available ‘out there’ in one form or another, face-to-face briefings prior to the Crowd Wise session on the background and purpose of the consultation made a significant contribution to the quality of the event.
  5.  We were reminded that groups will talk about what interests them – not all of the questions posed to the Crowd Wise sessions elicited a response. Some fell flat and it became clear that in these cases campaigners were happy to leave the detail to the Foundation. Our job was to pick up on this and not force the question, leaving the group to talk about the issues it had energy for.


At Rhizome we bring very different perspectives to consensus building. I have been using and then facilitating and training folk in formal consensus for getting on for 18 years. Perry, who developed Crowd Wise and facilitated the sessions for the Fairtrade Foundation, is twenty years a facilitator and tells me that until he developed Crowd Wise, he knew little about this field, and his concerns were more about how to create dialogue in a world that thinks that conversation in politics – however small the ‘p’- means debate.

If I hadn’t been aware of Crowd Wise, I could have proposed to the Foundation a method called ‘Small to large group consensus’ which shares several of the characteristics of Crowd Wise. But in reality I probably wouldn’t have done so because it would have felt like imposing my consensus ideology on them (I’m certain that they are interested in consensus in that ‘broad coming together’ sense of the word that is in common usage, but probably less so in the formal consensus process with its non-hierarchical values and its particular process quirks like blocks and stand asides).

Traditionally consensus works well among co-operatives and non-hierarchical activists who are committed to its values and for whom any form of voting can be suspect. Crowd Wise, by contrast, may need less commitment to consensual values, which makes it more suitable to situations in which there is some form of hierarchy. It may also appeal more to those happy to use (consensus) voting in the cause of speeding things up. The dividing line – if indeed it really exists – is unclear. We’ll be exploring it further at a training course run by Talk Action on 26th January 2012. You’d be most welcome.