The most significant development of the Rio IGF meeting was the dawning acknowledgement, particularly evident during the session on Taking Stock and the Way Foward, that the conference format of the annual plenary meetings is increasingly irrelevant to the work of the IGF. It is becoming obvious to many that another three years of the same dry platitudes on openness, security, diversity and access are going to achieve nothing, and that the IGF must move on. Even the addition of a plenary session on the provocative issue of critical Internet resources this year was insufficient to generate any new spark. So many stakeholder groups are now saying that we need to rethink the purpose of the plenary sessions, and to treat them as a venue for the reception of the much more focused and relevant work of the workshops and, especially, the dynamic coalitions. This is something that I have been saying for a very long time. However, I’ve also said that there should be some basic standards of multistakeholder composition and democratic process that a dynamic coalition ought to satisfy before it would be entitled to present its work for the plenary body’s consideration. Furthermore, what is the plenary body to do with the output of a dynamic coalition once it has been received? Clearly there is also a need for the plenary body to be reformed to constitute it as a body capable of effectively deliberating on the material it receives from dynamic coalitions. A much more suitable format to enable the plenary meetings to do this would be to divide it into small table groups, which would discuss the work of a dynamic coalition or workshop with the assistance of a facilitator. The short-lived proposal for “speed dialogues” to be held at the Rio IGF meeting would have been a productive step towards reforming the plenary body along these lines. Finally, once the plenary body has discussed the work of a coalition and the deliberations of the table groups have been collated, what is the next step assuming that a full or partial consensus has been reached? Does the proposal rest there? The IGF’s mandate in the Tunis Agenda says no. It requires the IGF to have the capacity to make recommendations, where appropriate. And that requires one further level of reform. Namely, it must be possible for the consensus of the plenary body to be assessed and formalised in the form of a written recommendation. It is not appropriate that the Chairman of the meeting, appointed by the United Nations Secretary-General, assume this responsibility. Rather this role can most legitimately be assumed by a multi-stakeholder group. The Advisory Group, which is itself appointed by the United Nations and acts only to advise the Secretary General, is not that group. A new IGF bureau (or an “Advisory Group +”, to spin on the Rio meeting’s “Athens +” appellation) is needed, to which the stakeholder groups themselves should appoint their own representatives. The result of these reforms will be policy development Internet-style; that is, developed at the edges of the network, in the specialist multi-stakeholder dynamic coalitions, but being endowed with democratic legitimacy through the endorsement of the plenary body at large, and being cast in written form suitable for promulgation into other fora by a reformed multi-stakeholder Advisory Group or bureau.