A proposal to better manage the ‘Community Feedback Pipeline’

Hello everyone!
I’m playing the messenger here: this post is brought to you by the product-circle. Special thanks to @Jana for leading this one :muscle:

Lately we had quite many proposals that requested community feedback in play at the same time. Paired with no clear timings this often leads to less and/or slower feedback and at the same time, not having the appropriate time to give feedback when needed can feel overwhelming for the community.

The product team came up with a proposal on how to manage these proposals.

We think this can be done by:

  • scheduling posts, to avoid having too many in play at the same time
  • setting clear timings / deadlines for feedback
  • defining clearer ‘responsibilities’ (i.e. who´s feedback is essential for which kind of request)
  • using a template
  • using tools to support the feedback process

How could this look like in detail:

1. Scheduling Posts & Tools

  • We are thinking of creating a separate Calendar for feedback requests, to have an overview of which requests are out, until when and pre-plan next feedback requests. Dates should indicate the entire span of the deadline (from date of post until deadline)
  • Before posting a feedback request, everyone should check the calendar to make sure there is space for another proposal in play
  • We can also add automated Slack reminders

2. Quantity & Types of Feedback

  • How to manage different “sources” of feedback, so that not too many requests are out at the same time?
  • Can we define a certain number of requests (tbd: different sizes and types of feedback)?

We gave it a try:

Types

  • Feature related (e.g. designs for checkout, Stripe SCA roll-out)

  • Community process related (e.g. papercuts, fundraising for features, instance manager circle)

  • Global team process related (e.g. feature toggle, fundraising for features,…)

Number of requests:

  • 1 max per type of request (independent of size, for starters)
  • Before a new request is posted, feedback round for previous feedback requests needs to be closed

Timing

  • 1 week per round (e.g. Monday to Monday)
  • We would create slack reminder 2 days before deadline, pinging the respective handles

Who to ping:

We would create a template here in discourse to help it out. With e.g. the following info:

  • Feature Related: Instance Managers, Support, Product, Dev, Testers

  • Global Team process: Product, Dev, Testers, Design

  • Community processes: Instance Managers, Support

Reminder: everyone is welcomed to add their names and handles here: https://docs.google.com/spreadsheets/d/1vc8W3QHl_oPJk4_oJP_mYJvIQ7wot8kLPlnX2Lp2gTo/edit#gid=0

Ping:

@Erioldoesdesign @Jana @Rachel @Kirsten @lin_d_hop @filipefurtado @sauloperez @Matt-Yorkley @maikel @apb @jibees

@Eugeni @tschumilas @lauriewayne1 @hernansedano @romale @rafat-khashan @berniemabbs @Cecilia-Hn @VPotvin @Jen @lbwright22 @konrad

2 Likes

Hi @Rachel before jump into the solution feedback I would like to know how many are the proposals in queue, and what community feedback means, because maybe some proposals don’t have the expected feedback is because the community don’t think is necessary, the not popularity of the proposal could be a feedback per se.

1 Like

Hello @hernansedano

Some example of proposals that needed community feedback recently:

Thanks for the examples. Looking the numbers about feedback in this proposals I can see this:

Proposal Created Replies Views Users Likes
Core/Non-core Instances and Sysadmin Support 20 .Oct .2020 48 495 11 30
Instance Manager Circle 20 .Dec .2020 15 197 8 3
A new ‘Split’ Checkout - Community feedback on designs 3 .Mar .2021 7 95 7 9
A proposal for fundraising for features 4 .Mar .2021 19 182 8 25
Proposal for Delivery Train and Product Curation Meetings 17 .Mar .2021 2 56 2 3

If this numbers are “less and/or slower feedback” what do you think would be the ideal situation?

also pinging @dthomas

This reminds me of our situation as developers because we have many more issues than we can work on. Our solution was to create a priority list and always pick issues from the top.

A key difference is that a dev issue can be picked up by one of many developers. But the community is less distinct. We have one community but not everybody’s input is required all the time.

What do we do with the proposals that are coming up but which don’t have space at the moment? I think that we still want to write them down but we need some kind of queue and then select from that for the next round. So maybe you discussed this already but here is how I imagine to fill this detail:

  • We create a priority list for proposals.
  • A new proposal enters the list at the bottom but can be pushed further up if the deadline requires it.
  • A new proposal can not be pushed to number one at the top because:
  • The top proposal is fixed for a week.
  • The community focusses on discussing the top proposal.
  • When the top proposal had enough attention and is resolved, it moves off the list and the next most important proposal is moving to the top.

Is that how you imagined it? Is this a good way of managing the queue of upcoming proposals?

To your question @Hernando

If this numbers are “less and/or slower feedback” what do you think would be the ideal situation?

If you have a look at your table (thanks for summing up by the way :)),
as example for ‘less and slower feedback’ Instance Manager Circle was posted on December 20 and then it took 2 months until the discussion started.

The Proposal for Delivery Train and Product Curation Meetings actually needed quite urgent action/descision, but it´s been 11days and not much input has been given.

These are two examples, where we think that having deadlines and indicating the urgency or type of feedback needed can help to improve the process.

The ideal situation would be

  • getting feedback from the people whose input is needed - or approval/consent of the ones who should at least be aware of the proposal - in time to move on.
  • after a deadline passed, things are not “rolled up” again (this goes especially for designs) to not slow down progress.

Following @maikel suggestion:

When the top proposal had enough attention and is resolved, it moves off the list and the next most important proposal is moving to the top.

→ Solving/closing posts after deadline. If something related comes up at a later point, a new topic should be opened and refer to the previous one to prevent endless threads.

Very true. That´s what we hope so solve by having the "Who to ping’ section in the proposal template.
Depending on the nature of the proposal, different people or circles are ‘required’ to give input

What do we do with the proposals that are coming up but which don’t have space at the moment? I think that we still want to write them down but we need some kind of queue

Exactly. That´s what we hope to manage with the feedback/proposal calendar. Almost like a ‘backlog’ of proposals.

It seems to me there are 2 related issues to this: 1. who are the people’s whose feedback the product team needs, and 2. timeliness of feedback. We only now have an instance mgmt circle in formation (thanks @lauriewayne1 ). So - no wonder there was a long timeline without feedback on the issues above - we didn’t have an instance managers for the issue to go to!! And - in all likelihood - we won’t have an efficiently functioning instance mgmt circle for some time yet. So - maybe in the interim the product circle can do 2 things: 1. post on the community AND on slack (at least for now) in instance mgmt, and give a deadline for feedback. I’d say 10 days is good. (A week is too short for volunteers who only are doing OFN stuff on certain days.) I know for my own responses - I need a timeline. I think instance folks are juggling lots of incoming with deadlines (proposals, meetings with partners, conferences, demos…) So - things without deadlines don’t get added to my ‘to do’ list. 2. Look for comments from 2 individuals working in instances that are not in the product circle. So - kind of a minimum level of feedback required thing. Most important - we need to make sure that the nature of the feedback request, and the way to respond is very clear. What questions do you want answered and how do you want us to answer?

2 Likes

Alright, so we have put several things in place:

  • along with the spreadsheet with all handles there is now an @instance_managers handle on slack we can use to ping all instance managers. We are looking into discourse groups to hopefully do the same and also handle the @product @design @devs @testers handles here the same way we do an slack :slight_smile:

  • there is a new calendar where we track how many feedback request are in place and when. If you want access to this calendar and haven’t yet, drop me an PM on slack with the email address you want an invite to :slight_smile:

  • some discourse categories do now have the template we have shared in the first post: let’s try it out! :slight_smile:

Let’s use these and iterate! I’m closing this proposal but anyone thinking we should change something is now free to try out the template in a new thread :tada:

2 Likes