Hacker News Reader

Comment by daigoba66 | Hacker News Reader

Once upon a time I liked Jira, and most of the products shipped by Atlassian.

While Jira has large admin overhead, it can be configured to make things simple for users. And I still think most hatred of Jira ought to be directed at management for imposing bureaucracy and control through the tool, instead of the tool itself (I see this first hand as my current gig has grown from “startup” to “enterprise” in a few short years).

But Atlassian has made a lot of questionable product decisions over the last many years, mostly in terms of UI/UX. What were once very much developer-centric interfaces are now geared to general enterprise product management. Somehow it’s becoming _less_ functional over time with each major revision.

View on HN

One of the lenses that we look at things through when making product decisions is making it hard to do things that end up feeling like paper cuts for developers all day long.

I've always seen Jira as the "manager's tool" and we're trying to make Clubhouse the "developer's tool."

It's a simple decision that I think (and I'm obviously biased) has had some profound effects on the way we've built Clubhouse.

I think the challenge with this is what happens when you’re in a growing product org.

At a certain point it makes sense to have product managers. Maybe for a while you keep optimizing for developer usecases, but that’s no longer your customer (even if they are most of your users.)

Larger revenue deals come from teams with more people and those are the ones that need more manager-focused features or where there manager is the buyer. From a go-to-market standpoint, you might put a ceiling on your revenue if you limit the addressable market to those teams that are still product managed by developers.

Not that that’s bad! Just a thought.

> At a certain point it makes sense to have product managers

I would just hire more engineers who have a good product sense and give them more power. That was Facebook's approach (initially).

if you’re building a product for engineers, that makes sense. your product managers are then former engineers.

if you’re building a product for normies, you want to hire normies as your PMs.

This idea that understanding how to build a product means you do not understand why or how someone would want to use it is baffling to me. Yes, some engineers have no product sense and no taste. But there are plenty that do. As is the case for all non-engineering fields as well.

Our product managers user clubhouse constantly and love it.

Club house has done a great job of organising information into stories, epics, milestones, projects.

All task manager products have some variation on these components but I've found clubhouses organisation of them to be the best.

It allows you to very quickly view the data differently depending on the context: stand up, sprint planning, PM trying to figure what's going on and what should be next.

My gut is that it turns out the way developers want to organise there work is actually a really sensible way of doing it, and so it scales up the management stack well.

Just to clarify, we think a lot about product managers, but when there's a tug-of-war between making life easier for developers vs making life easy for PMs we lean towards the former.

Since you said "we", some feedback:

I detest jira. The software fills me with loathing. We also pay them a little shy of $5k/year, and will be paying $12k/year by the end of next year. I've been begging one of your competitors to let me into their private beta. So I'm close to being an active buyer for software like this. Also, I can personally make the purchase decision.

I went to look at your site, and clicked on pricing because I'm scared of the word free. I want to pay for this type of software so I have reason to believe it will be there in a year.

However, it says I have to go to the enterprise plan to get SSO. Monetizing security is shitty of you all, and I don't want to have to take a call to get the price. I just bounced.

Last bit is so true, stop wasting my time and just tell me the costs. It just indicates that either you're too expensive or you're bad at concisely making a case for your product. Both are a red flag.

"Call for a quote" for software needs to die.

Personally also hate the let’s charge extra for SSO but at least with Clubhouse I can login with Google (G Suite account) and use the free product, as that’s what I just did a couple of hours ago to kick the tires.

I'm in same spot. Willing to pay (ie, routinely pay reasonably big money just for Microsoft VLSC stuff and a bunch of other specialized software 10K+ annually and willing to do $50K investments in licensing), but want to have some $ around pricing.

The problem is there are too many players in most spaces.

I literally DO NOT HAVE TIME to handle all the inbound / DRIP / follow-up sales marketing, go through your qualifications process etc. I need to do a quick pass - and pricing is an obvious part of that.

Ironically, we have dual fiber options at our location. One sales guy - here are our prices. Not cheap. Other sales guy - we offer a broad array of services yadda yadda would like to schedule a time to meet. We already have VPN / 24/7 on-call help for issues with it etc etc. Called first guy back, said give us a gig, static IP etc - send me contract and signed within an hour. Perfect. I don't even care if we are paying more. I'll look again in 2 years and if I can get someone to give me an actual price I'll make a decision then.

The call for quote folks are HARD to deal with when you change things like seat counts etc. - just be prepared to spend a LOT more time on the licensing piece which is harder to delegate than the folks with clear upfront pricing.

Well you're nailing it IMHO. I'm currently working at a big enterprise place to pay the bills while working evenings and weekends on a startup with a friend. Going from JIRA in my day job to clubhouse at the startup is like a breath of fresh air.

From the article:

“Schrader and Clubhouse CMO Mitch Wainer believe Clubhouse can maintain its organic growth by staying hyperfocused on designing for product managers and creating simple workflows that keep engineers happy.”

my workflow with jira is to click on a box, then move on to another one and start writing something, at this point the first decoration triggers a page reload and it wipes what I am currently writing.

There are just no words for how bad this UX is.

But it's also really fucking slow!

There is a setting to disable keyboard shortcuts in jira. It don't seem to do anything, but at least there is a setting for it. My guess is that the jira ticket to implement that didn't include what the setting should do, only that the setting was required

this is almost exactly my workflow, but before clicking on a first box I additionally wait several seconds until the page loads. Also, sometimes I extend that workflow with silent cursing and contemplating self-harm

As a full time software engineer I actually like their recent UI/UX changes. I go to Jira for a couple of simple things: to figure out what to work on, ask clarifying comments and update ticket status.

All the other features with velocity, burn down charts, workflow, etc I could care less about. Which makes sense, I'm not a manager. For me the new UI/UX means I don't have to think about using the tool, which is what I want out of a ticketing system. Funny enough, it really seemed to be management who was doing most of the complaining about Jira at my last gig.

i’ve used jira for nearly a decade and seen it grow over that time. i’ve worked in orgs that embraced it, made it a central part of business process workflows, and therefore paid super close attention to making it usable for the company. i’ve worked at places where it was a simple backlog tracker.

i think jira is a best in class tool for the way it is customizable. the ui has changed over the years but not greatly. the rolling and apis have gotten much better. but the way different teams even at the same company can use the same tool is extremely impressive. i think people who say it’s over complicated and bloated are correct, but at a certain size the workaround for that is multiple competing tools. one team on trello. another on pivotal. and that’s a much worse experience.

I have been on teams where an obscene amount of time was spent discussing Jira.

Not "how should we do things?" but "how should we do things on Jira?" where it was abundantly obvious that using Jira compared to literally anything else was worse. We could have hired someone full time and paid quite well to sit on a stool and move post-its around on a wall and it would have been better, faster, and cheaper.

Jira is fine when you have whole teams of people whose entire job it is to care about things like Jira who make it good enough for the engineers to be generally disinterested, but it is toxic for small companies, especially those which contain certain personalities or just inexperience. You get sucked in to endless discussions and revisions (which generally make things worse) of your Jira process with this idea that when you make it perfect work will be so much better but the result is a Sisyphean engineering team spending ever more time rolling the Jira rock over the hill.

There are a lot of cases where one guy with a pad of paper or some sticky notes on a wall (or literally nothing) are much better than Jira, but Jira is one of those cargo cult things people do because they are supposed to and it doesn't seem to add value unless you ignore any customization and just leave it or you have massive resources to put to developing and maintaining the process.

> We could have hired someone full time and paid quite well to sit on a stool and move post-its around on a wall and it would have been better, faster, and cheaper.

There is no audit log for postit’s. If that is your main use case, use Trello.

On the other hand, hiring a jira expert to configure that monster is great. She should also push back on locking things down to much: Do you really want only managers to reassign tasks...?

When do you need an audit log and what do you need it for?

I certainly have answers as to why you would, I'm not questioning the validity of needing one, I just don't think that very many people who use Jira are optimizing for what they want and need. You aren't in the business of having a detailed record of the engineering work you did, you aren't in the business of making burndown charts, organizing process, etc.

Those things are supposed to work to your benefit, you're not supposed to be working for them.

> it can be configured to make things simple for users.

That is the same defense I constantly hear about Jira. And it proves the point - it requires custom configuration to make it easy to use. So people invest time and energy into their instance of it, and build skills on it, and then are in too deep to move away.

I'm fine with Jira improving, and fixing their UX. I'm fine with other products coming up and beating them out. I'd just like to see this industry move forward, and get back to the place where our tools assist us in our work, instead of frustrating it.

I'm convinced at this point that attempt to combine PM planning & projections, IC planning, tracking for management, bugtracking, et c., all under the same roof, are doomed. I think some of those things cannot co-exist with the others without harming them, sometimes mutually.

Graeber calls out one particular cause of his "bullshit jobs" from the book of that title as excessive, misguided, and ultimately counter-productive attempts at highly detailed & granular computer tracking (complete with, necessarily, more rigid and complex processes). I think Jira & friends are entirely part of that effort-wasting, actual-work-harming trap. They let you see when a project has gone sideways but it's hard to tell that your project management software is part of why it went sideways and also part of why it took you longer to figure that out.

> "...most hatred of Jira ought to be directed at management for imposing bureaucracy..."

yup, that's probably right, as jira (and atlassian) has explicitly targeted large (bureaucratic) organizations.

scrum/agile/kanban tools should first and foremost be a lightweight process on top of development to keep folks coordinated and informed. jira can be configured that way while also providing a lot of visibility to product managers and other stakeholders, but most orgs seem to misunderstand both the purpose and the implementation.

with that said, clubhouse looks really slick (in the ~5 minutes i've looked at it). i've also been impressed with the advances that both trello and asana have made for the soho/smb project management segment in the past year or two since i've last looked at both.

I absolutely hate Jira, it's slow, everything takes many steps, overall horrible. That said, I recently tried their next-gen product and it's surprisingly really nice. It doesn't have all the features but it's fast and much more intuitive to use, even though it looks basically the same on the surface.

We have used Clubhouse for a couple of years, it's fine but I feel like the UI is getting more cluttered, there is very little reporting, few new features. For months the only news were updated to their referral program. They just introduced iteration support ️

Strongly considering to move to Jira next-gen, looks very promising.

> It doesn't have all the features but it's fast and much more intuitive to use, even though it looks basically the same on the surface.

The "It doesn't have all the features" part is the kicker here. I agree that Atlassian has fixed a lot of problems with their next-gen projects, but I think they're unusable for a lot of situations because:

1. the missing features (lots of workflow stuff that people rely on, for example) 2. they don't play well with other Jira projects. That's fine if teams are totally independent, but an epic from a classic project can't contain an issue from a next-gen project, as an example.

That said: if you've got projects that can work within the limitations of next-gen, I think it's a good move forward!

Unless Atlassian has made some major changes to the new Jira recently, I'd advise a lot of caution before moving.

My small team is currently on Clubhouse after attempting to use Jira for a few months and finding it to get in the way a ton. There wasn't any configuration locking anything down like everyone has been talking about, but the page load times were atrocious and the UI just felt painfully slow all the time (on a 100mpbs+ internet connection and a laptop with an i7).

I believe I also remember having to move back and forth between various views in the side menu depending on which page I was on - essentially going in and out of the "project" view when clicking into certain features.

Consider it by all means, but do your testing thoroughly. My Jira experience turned me off so badly I likely wouldn't consider using any Atlassian products in the future.

> Somehow it’s becoming _less_ functional over time with each major revision

I would argue that is now the normal case with almost every software. Technology had its peak some time ago. Most things are getting worse. The internet or telephones are great examples.

Reply on news.ycombinator.com