Adaptation: Feature Idea Process Changes

Community Team
Community Team
30 24 6,402

Feature ideas from the Community help shape Canvas development. All 330,000 of you (!!) have a part in the decisions we make about what to add and update next. The way we’ve solicited and interacted with feature ideas in the Community has evolved over the years. Today, we’re announcing some updates that—we hope—will make the process a little more clear for everybody.  

289980_pastedImage_1.png

About three years ago (April 2015), we transitioned the Canvas Community to the Jive platform and transformed feature idea submission and voting from semi-organized chaos to a transparent and concrete process. We’ve made some refinements since then. You can follow the history here:

 

We get a lot of feedback about the Canvas Studio, where feature ideas live—both formal and informal. Most feedback is positive—yet we’re always looking to improve how we learn about what you’d like from Canvas and how we interact with you about your ideas..  Our promise to our community members is that we will always communicate about changes and explain our reasoning for making them.

 

This year at the annual InstructureCon event we received a lot of very helpful feedback, including:

 

Feedback

Our Plan

Timeline

Submitting new ideas is simple, and six months is the right amount of time for idea cohorts.

The submission, moderation, and archive processes will not change.  How do I create a new feature idea?

Ongoing

The Radar stage is not clear—will something in radar stage be developed or not?

The Radar stage will go away. All current ‘radar’ ideas will move to ‘open for voting’ stage--consolidating the two stages to one.

ASAP

Complete

What happened to the 100 vote threshold?

We get this question a lot, so we wanted to address it again.  One hundred votes made sense years ago, but it’s a very low bar when measured against the number of users in the Community today (+330k).  In February 2017, we moved from an arbitrary number to a percentage metric that better represents the level of community support. You can find more detail in this guide; How does the feature idea process work in the Canvas Community?: How long are ideas open for vote?

Complete

There is no way to know which ideas were considered during the quarterly product prioritization


You want to know which ideas will not be developed and why

The top 10% off all ideas will continue to be considered during our Product team’s quarterly prioritization activities. Each will now receive one of three official comments:

  • Yes, this idea has been prioritized for development during this cycle
  • No, this idea has not been prioritized, but will be considered in future rounds if it remains in the overall top 10% of voted ideas.
  • No, we will likely not do this (and why)


We will create a new, easy-to-find list that shows the current top 10% most-voted ideas. We’ll update the list weekly so that you’ll know what’s in and what’s out as the quarterly Product planning exercise approaches.

Prioritization Comments to begin in October 2018

 

 

 

 

 

The overall 10% list will begin ASAP. is implemented on Ideas

 

Feature ideas are incredibly important to us. Canvas would not be what it is today without them, and they’ll play a big role in shaping its continued development. We appreciate your investment and engagement.  

 

*If you’re interested in learning a bit more about how feature ideas fit into the larger Canvas Product Plan, you might want to watch our InstCarn18 “Why we Community” presentation.

Tags (1)
24 Comments
Community Member

This Prioritization stage sounds really useful, rcarney@instructure.com‌: that will be very informative for the whole Community in addition to people following the fate of specific requests. Thank you!

Surveyor

Love the following idea:

"We will create a new, easy-to-find list that shows the current top 10% most-voted ideas. We’ll update the list weekly so that you’ll know what’s in and what’s out as the quarterly Product planning exercise approaches."

Community Member

I totally agree. This will make a great difference - as will seeing the top 10% list at all times.  Way to go, Instructure Smiley Happy.

Community Advocate
Community Advocate

Thank you for sharing this document with links to the history of the ideation process. It is nice to have a single place to point people when during training I share the value of user contributions and get asked about the details.

Adventurer

I also appreciate "No, we will likely not do this (and why)".  Frequently, the ideas we put forth or support are generated by our own faculty and staff, bubbling them up through us. This gives us the ability to go back to them and say "why". 

Learner II

Well said! Totally Agree.

Explorer III

Thank you for these updates, and for continuing to evaluate and improve how this process is handled. Adding a list that shows the current top 10% most-voted ideas that will be considered for prioritization will be very helpful. Thank you!

Will there be some way to normalize the percentage metric by primary user-base, and not just overall community? (maybe this is already done) Meaning, will Admin-only changes be normalized against number of users in the Admin community group, and a K-12 specific topic normalized against that large user-base?  For instance a feature idea that only directly impacts Admins will rarely/never reach the top 10% most-voted ideas, even if it indirectly, positively affects all users.

Explorer III

Agreed! Having explanations available as to why something wasn't considered is very helpful to those of us that work directly with the users.

Community Member

Actually, I would say just the opposite: admin-only needs are very well represented since the most active participants at the Community are admins.

Overall, though, I don't think the idea of voting is really very effective or even that important -- and after a certain point, of course, people get cynical when you see requests with hundreds of upvotes that don't seem to actually be moving forward. And, as I learned from the past Project Khaki, even getting voted in at Project Khaki doesn't mean an idea will be implemented. I am very curious indeed to see where Global Search now sits on the priority list after having been deferred (with no time frame) in the last Khaki announcement, with no more specific details shared at the Engineering Panel at InstCon.

But what really is important I think is how the discussion that emerges at the Community around an idea can help to clarify that idea. The more people who contribute to the discussion, the more specific and varied the use-cases become. And admins have a lot of insight into a wide range of users, speaking on their behalf and being able to bring those use-cases to the table.

Learner II

This document is fantastic. In my Canvas training I often talk about how one of Instructure's strengths is that it treats listening to its users like you would any other phase of development: necessary, transparent, and open to scrutiny. I'll be able to use this page to show that kind of approach in action, even when it comes to designing how you manage our feedback.

Surveyor

I'm very happy to see that Instructure continues to refine their feature request process as the community grows. I've been here since the "100 votes" days, and clearly that was not a sustainable metric. I'm especially glad rcarney@instructure.com‌ published this now, because there was something about various comments I heard at InstructureCarn that made me lose confidence in the current process. Since I've been back at work, I just haven't been enthused about voting or submitting feature requests, because I felt like nothing would ever happen. Now, I am re-energized by seeing the new process and how we will have a much better idea about what is happening and why.

I also agree with vanzandt‌ that special consideration should be given to Admin ideas (sorry laurakgibbs‌!). While we admins are almost certainly over-represented in the community in terms of ideas/comments, we are very under-represented in terms of sheer numbers (330k community members? I'm willing to bet only about 1% are admins); and our ideas can be invisible to non-admin users, but highly impactful. Just my 2cents!

Community Member

I meant active members in discussions, fosterl@oregonstate.edu‌... I honestly just ignored the 330K number... but it would be interesting to know how many unique members have voted on one or more ideas in the past year. I'm not really involved in the feature request process (for various reasons, not just because I am cynical), but I'm very active in discussions, and that's what I meant about admin representation, where admins are very well represented and there are not so many full-time instructional people.

Surveyor

I know there is an Idea out there that has more than 700 votes. I just don't know how long it took to get that many votes.

Surveyor

Found it! https://community.canvaslms.com/ideas/7532-allow-folders-in-pages  It has a "score of" 853 ( 925 votes up and 72 votes down). It's been open since November 11, 2016.

Explorer III

Wow! I haven't seen one that high before. Thanks for sharing.  854 now!

Community Member

I would also like ideas streamed according to sector and location. We are in the university sector in Australia and have different needs to US schooling sector. We have had odd feedback from the Canvas engineers in the past about why they have made a decision from a pedagogical standpoint which is not relevant or appropriate for Australian higher ed. 

Community Member

And that one is a good example of a situation where going at the problem feature-by-feature may not be the best strategy, although that is how all the conversation is structured here. There is clearly an architectural mismatch between Files and Pages which is very strange, and I personally hope that instead of a tinker-here and tinker-there approach to content management in Canvas, people will have a chance to take a step back and think about what a real evolution would mean in terms of a new, integrated approach to content management in Canvas, much like the big changes that have happened resulting in new quizzing and a new gradebook.

Community Team
Community Team

The top 10% sort (button) has been added to the Canvas Studio‌ page.  Take a look!

Adventurer

The 10% sort is filtering by tag, "dobrij,", is that intended? This will be a great tool. 

Community Team
Community Team

It is intended, and not a tag we want anyone else to use.  We use obscure tags when we have administrative sorts that we need to perform, and unfortunately we cannot limit a tag to only an admin, or we would.  So, please, we ask that people not muddy the list, and forget about the tag!  

Learner II

Thanks for finally removing the RADAR stage! I always thought it was a smack in the face to see that a feature was put into RADAR, effectively communicating, "Yeah, we know you want this, but we don't care, and we have no plan to ever move it out of RADAR. It's here INDEFINITELY. So stop asking." 

Meanwhile, the only features you do implement seem to have no significant impact ("Allow drag and drop in Dashboard").

My worry is that the only reason you've removed RADAR is because it had poor optics. In the end, I have very little confidence that you will ever implement anything significant that is requested from the community who struggles with your still nascent and very slowly developing product.

Community Member

This seemed like it would be a step forward when it was announced last year, and maybe it is, but I've still been disappointed. Of the top 10 ideas (by votes) none are being developed. 3 of the top 20 are marked as "influencing development", but with that low rate, it may just be that those 3 ideas corresponded with what the developers were going to do anyway, rather than the community actually influencing development in any meaningful way.

Community Team
Community Team

Hey brandon-davies@uiowa.edu

Thank you for sharing your thoughts here.  I can absolutely see where a perception like this could come from. That's why we always go the extra mile to try and communicate how much we value ideas and this process!

As we mentioned above (in the blog), community input it just one source that informs product decisions!  Our team is in continuous communication with product, but we also have a dedicated meeting once a quarter to review the top 10% of voted ideas for consideration to influence the roadmap (current and future). Each individual product team member we meet with earnestly loves so many of these ideas and repeatedly states they wish they could build most of them, but then have the difficult job of prioritization and managing resources.

The bottom line is we all care, and are passionate about creating the best tools for teaching and learning and we love hearing ideas directly from the users about how that can be done.  We can't build all the things (even though we'd like to) so we have to make hard choices.  Sometimes those choices align with the top voted user ideas and sometimes they don't.  Either way, we're always listening.

Learner II

It seems like Instructure may be spending a lot more time marketing their product than bringing important features into existence. If it is a resource allocation problem, perhaps there should be fewer resources allocated to sales and marketing and more resources devoted to product development. Certainly a lot of work has been done to provide a clean, minimalistic interface to students, but you can't stop there. Instructors interact with the software far more than students and have to do much more complicated things with it, and the instructor audience is one the product simply isn't ready for. You need to deliver to both instructors and students to make an LMS work. Until there are these minimal things in place for instructors, Instructure should re-balance its internal workforce toward product development.

Meeting once a quarter about these issues and feature requsts simply isn't enough. Have you seen the long list of issues?

About the Author
Happily teaching online at the University of Oklahoma since 2002.