4 ms·

This is a bit of a meta-discussion (apologies to the people at Tara, they did a fantastic job!), but is the market really large enough for the insane amount of

by dvt 6y ago

This is a bit of a meta-discussion (apologies to the people at Tara, they did a fantastic job!), but is the market really large enough for the insane amount of productivity tools that have been released over the past few years? From Airtable, to Monday, to Hugo, to Tara, to Asana, etc., etc. It feels like there's a lot of stagnation here, also. If I was building a product management tool, I would certainly not make it anything like Jira. Jira's a bit of a "dirty word," anyway -- it's a monolithic system that seems to suck more productivity than it generates. Where's the innovation? The last truly revolutionary productivity tool was Slack, and maybe Dropbox before that. I don't mean to throw any shade at the Tara people -- it seems like a ton of work went into the product. But I guess I just don't really get the state of the market (even though I work as an engineer and am often-times force-fed these kinds of tools).

rubyron 6y ago
Software is “eating the world,” as they say, and selling blue jeans and pick axes to gold miners is a perennially solid way to make money.
iba99 6y ago
At it's core, we find it to be an Apple OS vs Windows OS problem. You can either build highly complex project management systems, with 50 different types of subsets, customizations and setup instructions (ie Windows?) or try to design software that just works through simplicity in design and set defaults based on your workflow (ie Apple!). It's an interesting conundrum and one that we think B2B software rarely gets right. Why can't B2B software inspire delight in the user and just work? We've decided to have a laser focus on teams that ship early and often, and help with their workflows, by innovating through design. That being said, we still have a long way to go.
- oaiey 6y ago
  The problem here is that processes in enterprises are never default. Depending on the company the requirement to these processes are always different. Tools like Jira reflect these complexity on us during customization and are rated by their capability to handle them.
jillesvangurp 6y ago
I think Tara is on the right track by doing a deep integration with Github and Gitlab. IMHO project management tools need to track progress automatically where possible and grab relevant information related to pull request status, CI/CD feedback etc. The Gitlab and Github issue trackers do that and aren't that bad for managing large scale projects actually. Their weakness is that they are inherently tied to a single source repository and most teams and organizations would have many of those. Fragmenting information across multiple issue trackers is not ideal. IMHO Microsoft is on to something by grabbing Github and deeply integrating it into their stack. CI/CD is where the action is and a lot of that is cloud centric and a great opportunity for upselling things like Azure. I've been keeping an eye on the github issue tracker and they've added some nice features there related to project management. There's also codespaces, project discussions (in beta currently) and a few other features that basically indicates to me that they want to be a one stop shop for anything a development team needs. An org level issue tracker that aggregates information from all source repos and their issue trackers would be a great move for them. Jira has plugins for github integration that I've never seen working properly; most of them you need to pay for to even find out if they work or how. But the core issue is that this is clearly an afterthought for Atlassian and their tools are generally UX challenged resource hogs that are poorly integrated with each other even. They don't do github because they are still flogging the dead horse that is bitbucket; which is thankfully a lot less commonly used than gitlab and github. Basically they have a lot of me too products; I like exactly none of them. Trello used to be kind of nice and they sort of bolted it onto Jira in a weird way. We are currently on Asana and kind of liking it but the lack of github integration is not great. What's also not great is the lack of markdown support and things like code samples. But it gets the job done and the non techies in our company like it too; which is important for us. For a lot of projects I do, the release process these days is centered around github releases and github actions. E.g. we have internal libraries that we tag that get subsequently built and pushed to package repositories. A new release typically addresses pull requests and closes issues. It signifies the movement of those issues from doing to done (i.e. deployed/live/etc.) and reviewed (pr approved). The act of creating a branch can signal the act of starting a task. I consider work in progress pull requests a good practice for these as you get to have early feedback and discussion around it. Any tool that forces me to do task transitions manually is fundamentally wasting my time. Releases have release documentation associated with them. Compiling a list of "what did we actually change" manually is likewise busy work you should not have to waste brain cycles on. So deep integration with tools is the next logical step for a good issue tracker. Most existing ones don't do a great job of it. That's the market opportunity. Atlassian is ready to be disrupted. They got too big and bloated and they are not on top of their game for a long time. There are a lot of dinosaurs using them and a lot of misguided PMs insisting that the Atlassian way is THE way. IMHO they probably should but can't afford to gobble up Gitlab and that's actually a good thing. A reverse takeover would be a great lateral move for Gitlab actually. Also Google or Amazon could do worse then attempting to buy either (or both) and mirror Microsoft's strategy of owning everything relevant to running a software project in house.

by dvt 6y ago

rubyron 6y ago

Software is “eating the world,” as they say, and selling blue jeans and pick axes to gold miners is a perennially solid way to make money.

iba99 6y ago

At it's core, we find it to be an Apple OS vs Windows OS problem. You can either build highly complex project management systems, with 50 different types of subsets, customizations and setup instructions (ie Windows?) or try to design software that just works through simplicity in design and set defaults based on your workflow (ie Apple!). It's an interesting conundrum and one that we think B2B software rarely gets right. Why can't B2B software inspire delight in the user and just work? We've decided to have a laser focus on teams that ship early and often, and help with their workflows, by innovating through design. That being said, we still have a long way to go.

oaiey 6y ago
The problem here is that processes in enterprises are never default. Depending on the company the requirement to these processes are always different. Tools like Jira reflect these complexity on us during customization and are rated by their capability to handle them.

jillesvangurp 6y ago

I think Tara is on the right track by doing a deep integration with Github and Gitlab. IMHO project management tools need to track progress automatically where possible and grab relevant information related to pull request status, CI/CD feedback etc. The Gitlab and Github issue trackers do that and aren't that bad for managing large scale projects actually. Their weakness is that they are inherently tied to a single source repository and most teams and organizations would have many of those. Fragmenting information across multiple issue trackers is not ideal. IMHO Microsoft is on to something by grabbing Github and deeply integrating it into their stack. CI/CD is where the action is and a lot of that is cloud centric and a great opportunity for upselling things like Azure. I've been keeping an eye on the github issue tracker and they've added some nice features there related to project management. There's also codespaces, project discussions (in beta currently) and a few other features that basically indicates to me that they want to be a one stop shop for anything a development team needs. An org level issue tracker that aggregates information from all source repos and their issue trackers would be a great move for them. Jira has plugins for github integration that I've never seen working properly; most of them you need to pay for to even find out if they work or how. But the core issue is that this is clearly an afterthought for Atlassian and their tools are generally UX challenged resource hogs that are poorly integrated with each other even. They don't do github because they are still flogging the dead horse that is bitbucket; which is thankfully a lot less commonly used than gitlab and github. Basically they have a lot of me too products; I like exactly none of them. Trello used to be kind of nice and they sort of bolted it onto Jira in a weird way. We are currently on Asana and kind of liking it but the lack of github integration is not great. What's also not great is the lack of markdown support and things like code samples. But it gets the job done and the non techies in our company like it too; which is important for us. For a lot of projects I do, the release process these days is centered around github releases and github actions. E.g. we have internal libraries that we tag that get subsequently built and pushed to package repositories. A new release typically addresses pull requests and closes issues. It signifies the movement of those issues from doing to done (i.e. deployed/live/etc.) and reviewed (pr approved). The act of creating a branch can signal the act of starting a task. I consider work in progress pull requests a good practice for these as you get to have early feedback and discussion around it. Any tool that forces me to do task transitions manually is fundamentally wasting my time. Releases have release documentation associated with them. Compiling a list of "what did we actually change" manually is likewise busy work you should not have to waste brain cycles on. So deep integration with tools is the next logical step for a good issue tracker. Most existing ones don't do a great job of it. That's the market opportunity. Atlassian is ready to be disrupted. They got too big and bloated and they are not on top of their game for a long time. There are a lot of dinosaurs using them and a lot of misguided PMs insisting that the Atlassian way is THE way. IMHO they probably should but can't afford to gobble up Gitlab and that's actually a good thing. A reverse takeover would be a great lateral move for Gitlab actually. Also Google or Amazon could do worse then attempting to buy either (or both) and mirror Microsoft's strategy of owning everything relevant to running a software project in house.