Concerns about feedback and feature request tracking

[Oshyan]

Hey, we were getting pretty heavily off-topic in the other thread, so I split this to here. Continuing the discussion from Customized Entity Views:

I would love to see a feature in Fibery itself that could enable taking public feedback directly!! But I think for now that’s a big ask.

I know from following ClickUp that Canny can also be quite messy, and needs maintenance. At most basic level it requires changing of status (accepted, in dev, etc.), which also happens here, at least sometimes, by changing titles. This is a reasonable way to do it in Discourse, although tags might be as good or better. Canny also has poorer search, and this may be partly what contributes to lots of duplicates and diffuse feature requests. The ClickUp feature board is hardly a model, e.g:
https://clickup.canny.io/feature-requests?search=database

You’d think that search would surface some really meaty discussions around the future of ClickUp’s arbitrary data handling, wouldn’t you? Plenty of unvoted requests, overlapping requests, unclear requests, etc. and no way to filter/refine. And I can’t change the sort order either. The default view of the main page also doesn’t seem to allow changing sort. Canny is hardly ideal to me, or at the least it - like Discourse - is best when well managed.

If Fibery suddenly switched to it, I don’t think it would substantively address a majority of your concerns, with the possible exception of the desire for a clear indicator of an idea/feature’s “popularity”, because Canny does have the single vote-up button for the feature itself and not just the individual post. Discourse can have this too though, as you know. It’s an open question why Fibery team doesn’t want to implement it, maybe they have confidence in their existing approach, maybe they have some concerns about the triviality of “vote up” interactions (I’ve heard developers express this before), I don’t know.

I did just look and I do think that Fibery uses the official Discourse hosting, which at base level is $100/mo. If they’re using that package, they don’t have access to the “Topic Vote” plugin. It’s $300/mo to get access to that in the next package up! So they’d have to triple their hosting costs for the forum. It comes along with other benefits, of course, but I can understand them not thinking it’s worth it just for voting if that’s the case. But also looking at that page/plugin description I just realized it might not do exactly what you might want it to: users get a limited number of votes. If you could replenish that on say a monthly basis, it might actually solve some of the potential concerns with a voting-based system, but I’m not sure. Maybe I should test it, I have a self-hosted Discourse instance on Digital Ocean now…

Anyway, I do know that there are more tools that could be used in Discourse for good feature management out-of-the-box. Sub-categories or Tags would both help with the organization issue, for example. There is also really good thread merging and splitting (both occasionally used here!), among other things. Ultimately though they’re all just tools, and they require staff time to do the actual management. That, I think, is the main limiting factor. Which is why I raised the possibility of them onboarding mods from the user community. There are of course risks and challenges in that, so I don’t pretend it is trivial or a panacea, but it might help, and is worth considering. That or make it part of someone’s job to be sure to fully manage the forum content, if it seems worth the time and effort. We have to accept, though, that it may not, and this forum is only one source of feedback they get. They also have a lot from Intercom which is invisible and may take priority…

[mdubakov]

@Oshyan Thank you for starting this thread, I wanted to do it myself yesterday but here you are

Let me provide my perspective on the topic:

1. I didn’t see any single tool that is good enough for viable feedback accumulation. I used UserVoice heavily in my previous company and didn’t like it. Canny is very similar so I don’t like it either. The problem with all these tools that they enforce data model, and in Fibery we use quite different data model for feedback management (more on this below)

2. I don’t like votes, since votes are easy and it takes nothing to click and vote. I do like verbal requests though since they demand at least “some” effort.

3. The main goal of the Fibery Community is to provide and discuss ideas/problems/issues, and that is what happens here. This is not a place to prioritize features or display what features are more important. That is for us to decide.

How we prioritize features internally?

We accumulate feedback from 3 main channels: Intercom, Calls + Meeting notes, Discourse. As a result, we have verbal feedback that accumulates into a number of references and we rely on this number to identify feature importance. Here is the list of all Ideas/Features/Insights and User Stories sorted by the number of references in the last 6 months:

As you see, we do try to work on top requested things. There are some exceptions. For example, permissions improvements were on top for a long time, but we can’t start them since we didn’t have a good conceptual solution. It took use about 6 months to find it and now we are starting permissions implementation. Form View is on top as well, but so far we don’t think this is a must-have feature for Product Teams (maybe we are mistaken, but that’s what we feel right now).

We have the capacity to work on 4-6 major features simultaneously.

That is not all. We also have many small things that have 3-5 references and have no chance to be on top, but they can be bundled into some initiative like “Better Getting Started experience”. People express problems differently, but that is our job to analyze them, generalize them and bundle into features. I’d say this is not what we are doing very good at this moment, we still somehow prefer to work on large missing Features from the image above, but we are thinking about this process and going to improve here as well.

Ideal solution

Ideally, we should just share some parts of Fibery and allow people to communicate on Ideas and Insights, in this case we will have a direct match between our terminology and set a stage where you can communicate more efficiently.

Here is the video how we categorize and prioritize feedback.

However, this is a very hard technical and even conceptual problem and I don’t know at what point we will be ready to solve it.

[Matt_Blais]

Thank you so much @mdubakov - I love hearing about your process and progress.

I think everything is perfect, except you are 50% understaffed

[Oshyan]

I very much appreciate the detailed reply, and I largely agree with @Matt_Blais : it seems like you’re doing the best you can with the resources you have. It is good to know that you feel you could be doing better in “bundling” work and distributing effort between big missing features and other things, though. I think that might help address some of the concerns that drove the original discussion, as it does seem like some small, “simple”, and seemingly quick to implement features (like [PLANNED] Number Field Formatting Options - Currency) remain unimplemented for a long time.

Anyway, this is a useful thread to refer back to when frustrations around feature X not being available yet come up. And I don’t know of another company in this product space where I can get such direct, high-level engagement with core team members, and that’s valuable even if my pet features are further down the roadmap than I’d like!

[B_Sp]

I appreciate the response, Michael.

When you provided a glimpse into the internal feature weight and categorization here with your link to the live tracking of your guys’ own Fibery Product Area set up, I wondered if you couldn’t integrate that more into a public-facing area where users can actually see the difference between what’s here in the community, and what you guys are actually working on. In particular, the difference in weight between what’s here in Discourse, and the actual weight you have assigned requests officially. You have provided another glimpse into your internal roadmap now, so I am wondering this again.

For example, you showed in your table “Rich Text or multi-line text in Table View” as a top item that is for now on hold. I’m very glad to see it’s even that high up, as this is a big item for my particular use case, that I think a lot of others have supported, but it remained hard to figure out where it was on your guys’ priority list. Now I feel much better seeing where it actually is.

I don’t think Canny is great either, but it does allow those using it to publicly communicate what they are working on, votes aside, with just a few simple statuses like “planned” and “in progress” etc. You guys occasionally manually mark up certain Feature Requests with a similar status, but I can’t figure out when you do this for a request, and when you don’t. And there’s no way to see in one place all the items in the same Status. For example, I’d like to be able to see in one area all the “Planned” items together. There are many teams that even use Trello boards to do this, such as here:

There’s no need to count user votes with a solution like this, it’s just a nice way for the users to see what’s being worked on, what’s next, etc. You have basically teased such a view for users a few times now…

At any rate, I’d really like to have a way to see live a view just like you provided here. Perhaps if you guys incorporated that into the community, it could be the beginnings of your “Idea Portal” solution to match those of ProductBoard and Aha? Count my big vote for a feature like that!

Cheers!

[Oshyan]

I wonder if it’s time to consider bi-directional Discourse integration. I know, more work on Discourse @B_Sp , but hear me out.

If changing a Status in Fibery on a Feature linked to a Discourse Post could then automatically add a corresponding tag in Discourse, you could essentially have what you are asking for. It would not of course be a comprehensive feature/priority list, but of the features discussed in Discourse already and that were linked into Fibery, you could easily have a reflection of what features are in what status internally…

[B_Sp]

If the team could figure out a way to expand on the Discourse integration and we’d accomplish a Public-facing Feature Tracking tool ala Aha and ProductBoard, then I think that’s interesting for sure. Discourse has a ton of cool features, don’t get me wrong. I think if what you’re proposing came with some substantial customizations of Discourse, using those tags for example, it might work!

Either way would love to see Fibery come up with some Idea Portal - type solution, especially as that’s a key piece for Product Teams. Aside from Aha and ProductBoard, Craft.io is another not to be ignored - very solid tool which includes that feature.

[Oshyan]

Seeing if I can reply here from the other topic as it seems more relevant here…

@B_Sp

…and I can’t help but mention yet again that we have here another example of fragmented requests in this community that an organized voting / request tracking approach would solve, right @Oshyan?

I don’t really know for certain how Fibery team is tracking this stuff. I do have a Discourse instance integrated with Fibery though, as a test, and I can tell you that it has a column in the “Topics” table and what it shows is total hearts in the entire thread. I think that may be new information to the discussion here, so think about it for a moment…

Presumably this is either A: because this is what Discourse exposes in their API, B: this is what Fibery implemented because it was easiest, or C: this is what Fibery team did because they felt it was most useful/representative as sort of aggregate of positive sentiment toward overall content of this topic’s subject and ongoing discussion. I can see it being somewhat inaccurate in certain respects, and some might argue that hearts on the 1st post should have more weight, or that the hearts on replies might be misleading (e.g. if a reply strongly disagrees with the topic starter and many people then support that person’s disagreement).

But in general I think they probably use this “hearts column” in some way in their feedback weighting process. My guess though is that it is probably not a direct 1:1 addition to feedback “weight”, they may divide by some other factors, such as number of replies or something. And they may also weight their own highlighted/linked content from the forums more highly too.

Now of course some more transparency in exactly how they weight feedback from the forums vs. other things would be nice for us forum denizens. I think in fact this is the principle concern being expressed here and in other threads like Suggestion: Canny or other alternative for roadmap/feature requests I think a more clear way of showing Discourse-specific “feedback weight” in Discourse would be nice, but even without that I do feel confident that forum feedback is taken into account through this integration, and it seems to provide enough detail that there is no technical reason why that would not be the case.

Having said all that, I do think the topic voting plugin we’ve discussed could provide some clarity in certain cases where hearts don’t necessarily do quite that job (e.g. in the example problem case I described above). Notably that would be in the area of providing more clarity to forum viewers and voters. But the current Voting plugin also has some potentially annoying limitations. Anyway, I would be in support of it being tried here, but I don’t see it as being dramatically better than what they have.

Bottom line I hope you will feel a bit less uncertain and anxious about how and whether feedback is being taken into account here. They have an integration directly into their feedback and dev management system, itself implemented of course in Fibery. And it provides basically all the relevant Discourse data to be prioritized and actionable. We will still see seeming disconnects between our priority here in the forums and what is actually worked on, of course.