Hey all, I have a question I’d like to trow into this group and get your feedback on.
I’ve been managing a team of technical artists for some time now and the one thing that always remains a difficult problem is how to handle the task tracking/scheduling for our kind of people.
In our case the TechArt team consists of both pipeline programers/scripters and character riggers who deal with distinctive different types of tasks that require their own tracking methods.
Also for a large part of the Tech Artists their time is almost equally divided between feature development and bug fixing/fire-fighting.
In the past I’ve tried various type sof software solutions like MS Project, Excell, Jira, Hansoft and even writing my own solution
and I still feel that maintaining a good overview of tasks and priorities is taking up too much time.
Maybe it’s just the TA in me that tried to come up with the perfect automated solution
Anyway I’d be very interested to get some feedback on this and hear what solutions have been tried other teams.
Why don’t you just set up a website where, at the end of the day, your TA team just puts in how much time they spent on each task? It seems pretty straitforward… maybe I’m misinterpreting the question.
While I do understand the use for task tracking software, I also find that they over complicate things a lot as well unless you deal with tasks that are more repetitious then what TA teams usually deal with.
During the final year at Grin, they started using scrum, Hansoft and other ways to plan peoples work. I managed the TA team there, which was stand alone to be able to jump between projects like the tools and engine team, and the producers never asked me to use any of it. Why? Because they knew we don’t really work by a schedule, and that was not the best way to make use of our team.
Like you say we are fire fighters, which is what we did at least 50% of the time. So we could have planed the remaining 50%, sure, but it would never work anyhow as that’s just an average and was not the same each week. Besides fire fighting we also had a lot of other small tasks pop up on a almost daily basis, which made it even more futile to try and scheduled our work.
Here at Avalanche we don’t have a TA team, and we are merged in with the artists and animators in the project we are assigned to. What we do is state weekly goals at the beginning of the milestone, each day at the morning meeting we report progress on that weeks goal and is there are any issues we need help with. So here do kind of have a task tracking system… but it veeery flexible as a lot can happen in a week during production.
This may or may not be worthwhile reading. It’s not directly applicable but I find it interesting:
Basically there are two things in there which I find very valid, although I have not implemented anything based on it:
Charge bug fix time to the project which created the bugs (as in, add it to the time it took to add the feature originally).
Don’t separate “fire fighting” from “feature development”. Just add it all together to start getting a picture of how long it really takes to get a feature in, including all the interruptions a TA goes through.
At the end of the day, I think you could possibly end with a proper method for realistically tracking TA tasks. This way even if a feature request appears that it will take 1 day, you know the actual task will take 3 days - because 66% of the time is spent doing the “other stuff” (fire fighting, random meetings, etc) that interrupts a TA. Since you’ve started tracking the actual time it takes to get a feature in (as opposed to estimating the time it would take to do without interruptions, and then guessing how much extra “fire fighting” time to add in), you’ll be able to (in theory) develop more usable estimates.
I suppose it’s possible you might get pushback on things (“Why do you say this feature change will take 3 weeks, I think it’s only a few days of coding?”), but you’ll also have a back catalog of concrete numbers to back up your estimates.
Again, I haven’t implemented any system like this, I just wanted to mention it.
I’m in the same boat. We do development, fire fighting, VFX, export & packaging… pretty much everything a TA could do. Tracking websites don’t work if people don’t honestly fill out hours or if they forget to do it. After a while the info becomes rather hazy and not much better than educated guesses. Been there, done that.
The biggest problem with tracking is that most TAs have multiple projects. Some require full time attention, some are just about being on-call. Then there’s the daily fire fighting interruptions. Sometimes artists have to be shuffled around if a certain skillset is required by a certain project or resources are needed. Having a single TA assigned to a single project full time is rather an exception here.
At least now the team members all go via their leads - before I came here TAs would just listen to anyone who said “hey, I need help”, and the TAs might end up doing something totally different than the tasks prioritized by management.
Since we’re a sort of “service” department I now outsourced all the tracking to the in-house clients. They have to create tasks on Hansoft, they have to complain if they feel the progress ain’t good. They, me and the involved artists plan time and give estimates. Ain’t optimal either but as Serguei said, otherwise it’s a full time job and the team ends up being 1 person short.
I’m better at helping my team with making resources available, helping with the tough jobs, choosing the right guy for the job, etc, rather than investigating how many minutes everyone spends doing this or that and digging around in Hansoft.
A coordinator is probably the best solution. Chance is they know much better what they’re doing than you do, making the whole tracking much more effective than you could ever do it. Plus, you’ll be able to focus on what you love doing: tech art!
Thanks for the link Phil and I’ll also have a look at Shotgun. I haven’t given up on the issue yet, but it’s good to see I’m not alone here
I am self managed at the moment. I do have a weekly meeting with all the tech artist. We then discuss what has to be done and who will do it. But that is kinda it and it works great.
Thanks for the reply’s people, it’s very nice to hear about other peoples experiences with this.
Shotgun is something that I have looked at a few times before and it definately has some very nice features.
The main issueas have been so far that I wasn’t able to convince the studio to use this instead of Hansoft and also it seemed like overkill for the type of tasks that my team works on (though I may be wrong there).
I do agree that having a dedicated coordiator would be the best approach for me and that’s what I’m now trying to budget for so let’s see how that goes :):
Another thing that I’ll be looking into is Jira since it looks like this may be a nice system for tracking both features and bugs and generaly keeping a good overview per tool/pipeline.
Anyone experience with using this for TA tracking ?
There are a lot of time-tracking solutions, if that is what you’re after. Rescue Time is one I use for my own productivity goals.
But time-tracking is altogether different than scheduling. Scheduling is always tricky, since you’re essentially asking someone to give a time estimate based on pretty limited information. For straight-up asset generation–creating models or animation–this isn’t so bad. But once you start getting into fire-fighting and bug fixing, you start getting into gray areas quickly. And even very experienced devs can be terrible at giving time estimates, as we often have a habit of underestimating the complexity of a problem or overestimating our speed.
If you’ve already got the personnel in place, I’d second the advice of having someone dedicated to production to handle the schedule and tracking metrics. Otherwise, try checking out time-tracking solutions for tracking metrics, and maintain enough slack in the schedule so that fire-fighting doesn’t totally hose your tools-development schedule.
