Idea Conversations: What Comes Next?

[shirenv]

Over the last few weeks, our team has brought cross-functional folks from across Instructure and customers from our community together to chat about the functionality requirements of Idea Conversations.  Our goal is to get to the root of everyone’s needs from Idea Conversations and have a well-rounded discussion on how we can build an updated solution that fulfills those needs and creates actionable insights for our product strategy.

What you’ve told us is needed for the next version of Idea Conversations: 

• A safe space for community members to share what they need help solving in order to promote teaching and learning with Instructure products
• Communication channels that consistently assist us in understanding the current problems at hand and to provide ways for us to ideate together on how best to solve them
• A more present and consistent product voice
• A way to provide the why behind product prioritization decisions, even when we can’t prioritize it

So what comes next? 

On November 30th, we will:

1. Turn Idea Conversations to ‘read-only’ mode for ~60 days (through the end of January)
2. Highlight the most common trends our team has identified from all the open ideas
3. Provide an outline of what you can expect for the next version of Idea Conversations

“Wait, what? Idea Conversation is moving to read-only?” It might seem counterintuitive, but it’s what we need to complete our analysis of the existing 4,446 open ideas and efficiently transition to a scalable new solution by early 2023. We recognize there is very rarely one solution that solves everyone's needs, and the ultimate solution will best balance opportunities for input with a scalable way for active product team participation. 

Further, in an effort to be as transparent as possible and provide context, strategy, and insights on a regular basis we will be sharing a monthly Product Blog that spotlights use cases across our products and provides updates on any changes in product strategy or prioritization. 

We appreciate all of the input we have received so far and are looking forward to deeper opportunities to collaborate. We are working hard to develop the tools and processes which will make this process simple, fun, and enriching for our entire community. And while we are in this transition, please continue to leverage other resources available to you throughout the Community, our Support teams, and your CSM / CSM team.

[leward]

Thank you, thank you, thank you @shirenv! I have been so frustrated with Idea Conversations for many years, particularly since the most recent community migration to the Khoros community platform. At one time, the number of votes for each suggestion displayed on the search results page and it was also possible to sort suggestions by the number of votes. This was incredibly helpful to our institution for a variety of reasons and provided a level of product transparency that has since been lost. I also would advocate for simple binary voting rather than multiple stars. But regardless of what form the new idea conversations takes, I really appreciate your willingness to make BIG changes in response to customer feedback.

[james_whalley]

You may also want to check the Idea Conversation precursors, as I remember some of these same ideas that had hundreds of upvotes suddenly only had dozens of 5 stars. It would be nice to recapture all of the upvotes that were lost in the last update.

[chriscas]

@shirenv,

This is fantastic news! I think a there is a lot of frustration with the current process and the lack of transparency about what Instructure does with the posts in the Ideas area. I'm thrilled you've outlined both "A more present and consistent product voice" and "A way to provide the why behind product prioritization decisions, even when we can’t prioritize it" as things customers want. I know it will probably be difficult to make everyone happy, but those two things would go a long way for us.

I think it would be great if up/down voting could return, as the current star system isn't very intuitive, but I don't think the current community platform supports the up/down and I know there are also issues with voting like that.

Overall, having some kind of communication from product on the most popular requests on a regular basis (whether quarterly, yearly, or whatever would work best), is something we'd love to see. Requests that have been around for 5 years and garnered a lot of interest (whether by stars, by votes, by comments, etc) should at least get some official acknowledgement. Even if this means telling users "we won't be including this because of X" or "we like the idea, but can't work on it in the next year" or something, I feel those responses would be better than radio silence.

Again, Thank you for trying to wrangle in the idea conversations beast! Very much looking forward to what you come up with for the future!

-Chris

[brian_mullins]

A great move. I think more moderation, making sure clear use cases are provided would also be helpful. In addition, moderators need to be much more proactive in merging the same or similar ideas. Also, tagging should be more encouraged to make search easier.

Star system stinks. Simply move to a vote system. We don't need downvotes. We just need to be able to say yes to something we want. Ideas (powerbi.com) is a great example. Clear and concise insight into the most wanted changes.

Thanks for the work on this!

[chriscas]

Just a quick comment on @brian_mullins reply... I think downvoting is actually important for the process. Instructure doesn't seem to like coding in too many options, so when they do implement ideas, it tends to be done globally and affect all users. Upvoting is probably used much more, but if there is an idea that would cause issues for certain people or institutions as a whole, I think we need to be able to downvote to show that.

-Chris

[audra_agnelly]

I miss the old platform for a number of reasons, one of them being the shift that happened with the Ideas space. In the old platform there was a process where an idea collected votes over a period of time and then moved to categories along the lines of "under consideration", "reviewed and not gonna happen", etc. That wasn't always great, we know there are ideas that sat around after that voting window or ideas that had massive amounts of votes but never moved forward into development, but now it feels like all ideas just go into the ether with little to no acknowledgement from Instructure...are you considering a posted idea? is the idea too difficult to implement for the payoff? is the idea even being read by someone on the product team?

I hope in this process that CSM teams are brought into the ideas discussions. Many of the ideas posted to the communities are pain points that we admins are discussing regularly with our CSMs; they know what features users across accounts are asking for.

I also hope that moving forward, there is some weight or focus groups looking at the needs of K12 users. If we are only looking at a voting process with ideas, those that are more specific to K12 needs will always have fewer votes than higher-ed-specific or more general use ideas.

[brian_mullins]

@chriscasAgree that some things might have a negative impact on other users. Here's how I'd hope an upvote-only process would work. If I find an upvoted idea that would negatively affect me, I'd add a comment to the thread listing additional use cases showing the negative impact. My expectation for any vendor offering to hear input from users is that they would not just take an idea and run with but instead choose ones to implement based on the votes and listen to/involve the community to figure out how to best implement with the least impact.

Anyway, I can dream, right? 🙂

@audra_agnellyAbsolutely right about CSM's being more involved.

[Mikee]

This is both excellent and concerning 🙂

The current ideas process is pretty clearly broken, and the links both ways into and out of the engineering teams seem somewhat constricted from this end.

As part of a mult-layered support structure here to serve a student population of near 50,000 there are often features that simply don't scale to the extent we need - even fundamental things like the gradebook, and speedgrader. While voting is brilliant when most of those who are engaged here have similar needs, I'm concerned it could lead to a much stronger K-12 focus for the platform which would be categorically disastrous in the longer term for us.

Your transition to a more platform-like approach is brilliant, and we're really looking forward to seeing what the future holds.

[rake_9]

Adding my voice to those expressing desire for more moderation, interaction, and insight from the product team.

At my large university, I frequently hear some variation of "we have to be a big enough customer that they'll code what we need!" One of my responses to this kind of statement (after some skeptically received assurance that we are not, in fact, Instructure's biggest customer) is to share several Idea posts that are similar to what the user wants but that have gotten very low support or have a number of comments explaining the problems with implementing the idea. This worked better with the older upvote/downvote system, because it was easier for a user to see that only a handful of other people thought their idea was useful.

Many very smart people don't understand how application development works. While I don't expect Instructure's product team provide end-user education on how they do their work, I would love to be able to share articles that can help people understand that what seems like a simple and obvious request isn't so obvious or simple.

[mpfaff1]

As a Canvas admin and faculty support staff: It's been disheartening to see good ideas "Open" since 2016. What does an "open" status even mean with people still voting on it after 7 years? It's hard to find motivation to submit and rate ideas when hundreds of votes disappeared with the old system.

Current Product priorities lack transparency and feel out of touch with education. For instance, Cross-listing is a popular, fundamental feature, but has long-documented FERPA concerns that almost made our college abandon it. We were explicitly told by Product that: 1. Instructure has its own interpretation of FERPA on this matter, and 2. fixing this issue isn't popular/requested enough to be a priority.

This was news to us. FERPA compliance isn't the kind of high-impact feature that gets faculty rushing off to vote on Idea Conversations, but we're clearly not the only institution that's addressing this as a compliance issue. One that we could potentially lose federal funding over if not treated with care. We couldn't help but wonder if Instructure thought that should take a back seat to confetti animations. What "feedback" were they acting on when they instead decided to completely overhaul every core Canvas feature almost simultaneously?

While admins are outnumbered, we collect our own feedback from faculty and students and put forward concerns impacting a wider audience, creating hours of extra work for faculty or worse student outcomes. Not all schools have the time or talent to code their own workarounds for these: our admins might be office managers or librarians. Which is why I advocate that admin and institutional requests should have their own path to approval, their own lines of (public and transparent) communication, and be accepted or declined based on their merits, feasibility, and feedback from other schools. Not just a popular vote, or "talk to your CSM," or promotions for premium, "white glove" services.

Canvas as an LMS does a lot of things right, but it's getting harder to believe that EdTech has any educational interests left. So far Idea Conversations have only been asking, "How do we solve user's problems?" Instead, the focus of the new platform needs to be answering: "Here's how we're solving educational problems."

[mwakefield2]

I applaud the effort to do better information gathering.

Thoroughly agree about the "user" vs "educator" (and "learner!") focus, after many years of watching missing essential functions languish in favor of novelties developed by, um, developers. If at least a few of those devs actually had to teach several courses for a quarter, semester, or year using their system, I think they would better understand the nature of the needed priorities (especially ones that have been solved by/exist in other LMSs for decades in some cases).

Sorry to say that after all these years of waiting, I am skeptical that this process improvement will result in any meaningful change in the product itself. Educators (users) in the trenches will likely still remain at the bottom of the communication food chain.

Time to go back to grading the final project files I have to download from "New Quizzes" one at a bloody time (and still not be able to mark up). /rant

[BradMoser]

Thank you for addressing this roadmap and ideas/development struggle. As a Canvas user since 2012, I have seen the iterations come and go. I feel like more recently, the suggestions and influence of users (teachers) have decreased and it appears there is not a rhyme to why one feature gets developed while another more requested doesn't. 

I am hopefully that your analysis of the 4000+ ideas will yield some more focus and needed improvements. Thanks for listening. I have had one of my very own ideas to Canvas selected and developed. It feels good. Hopefully, we will get back to that for all users. Thanks! 

[mbflatau]

I came here looking for the feature request page and forgot this was temporarily disabled. We can't be the only K-12 district with this issue, so I figured I should just post this here anyway to see if anyone else has an idea for a workaround to combine quizzes (built-in New Quizzes) into each other (into one big master quiz). Wish that "Add question to Item Bank" was an automatic feature because if you forget to click that button, you have to go back in and edit each question to push it into an Item Bank manually. Tried to do a course export (quiz only) and import all the quizzes into one big Item Bank - but discovered the file that Canvas produces doesn't even integrate into an Item Bank (it's producing a too-old version of a QTI file to import). Just hitting the wall everywhere I turn - any suggestions? Aside from manually going into every question and adding it to an item bank?