Stress-free Solo+ Creative Work by Stealing DevOps Techniques
I watched this GDC video by Seth Coster a while ago. It’s excellent for its intended purpose of helping game design teams, and I recommend you watch the whole thing if that’s relevant to you. However it also got me thinking of how to apply the principles to other kinds of creative work, especially solo.
The main point that stood out to me is the stance of “Crunch is not an option”. In game development, crunch is a period of overtime work before a deadline, bringing up 60, 80 or even 100 hour weeks — usually unpaid overtime. Seth’s stance is that crunch happens when systems fail. If you’re crunching, you need to stop work, however painful or counter-intuitive that might be, and make better systems. The alternative is staying with those same broken systems for years, until either the company or its components fail. It’s best for the long-term health of your company and employees — or in the case of solo work, yourself — to make systems that ensure you can get good work done in a reasonable amount of hours, by being efficient.
Their team has a hard limit on their work hours per week. When something is limited but necessary, it becomes valuable. Following the same attitude, if I have a project I want to complete by a certain time, I can’t slack now and work more later — that’ll take me over the hour limit. If I cut the safety net of personal crunch / cramming, and have a known quantity of hours to work with, it becomes urgent from the beginning to work consistently and use time well.
Types of Waste and Solutions
If your hours are limited, you need to make the most of them and be efficient. To do this you must eliminate waste. Waste is time spent not serving value to your customers/audience — in other words, Not Making [Art].
The section at 15:20 in the video outlines team-relevant types of waste like Motion (moving tasks between people) and Waiting (for others to finish things), and a more in-depth look at the rest too, so check there if you want a deeper look. Here I summarise the ones relevant to working solo with some examples of how the solutions could be applied to creative work.
Task Switching
When changing tasks it takes a little while to get going in the new one, so switching often leads to slow work.
Solution: Batch work
Do a bunch of one thing, instead of taking it through all the different stages then starting over. For example instead of taking one piece through the sketch > ink > flat > render pipeline, do a batch of sketches then a batch of lining etc. so your mind is not switching concerns and you can get into a groove.
Communication and social media are also prime candidates for batching, as they can be constant sources of distraction if tended to throughout the day, but are still important to do at some point. Cordoning off a specific time to get through a lot of emails or DMs or messages at once lets you avoid getting pulled into switching tasks too often.
Manual Processes
Doing things by hand that could be automated takes time away from other tasks that can’t be, and makes them vulnerable to human error.
Solution: Automation
Look for things I do manually that could be automatically handled by a system. It helps to have a developer friend or development knowledge to set up custom tool solutions for yourself, or if it’s really worth it or really easy, to pay a developer to make it for you. But if it’s a common activity with demand, chances are there’s a ready-made solution already out there.
Some quick examples of automatic processes off the top of my head:
- Formatting things for different social media platforms. You can set up Photoshop actions to properly resize and export them.
- Posting on social media platforms. Programs like Hootsuite and Buffer manage all your social media channels in one app, and you can post to all of them at the same time. They cost, but if they save you time that you then use earning, they pay for themselves. There’s also the side-benefit of not having to visit potentially distracting sites to post.
- Initial client contact/validation. You can set up forms and automatic responders or chatbots to do some heavy lifting at the top of the funnel. It’s a bit impersonal but if it’s practical information you need, the exchange acquiring that info will be pretty mechanical anyway so might as well make it automatic and save yourself some time.
- Communicating with clients on project progress. You can set up a public or invite-only Trello board linked to your to-do list to let them check on progress whenever they like.
Unfinished Projects
Using time and effort on something that never gets across the finish line is a waste of both.
Solution: Small batch delivery.
“When a project sits unfinished even for a short amount of time, the context of the rest of the world changes.” For my personal work, what changes over time is my enthusiasm for the project. The longer it languishes unfinished, the longer I have for doubts to creep in and other, fresher things that don’t reek of guilt and insecurity to occupy my attention. So it remains in the WIP folder, and all the work lies Wasted.
This is especially relevant if it’s something timely, like a new game launch or an art challenge that’s going around. If you leave it too long you’ll miss the window of relevance and what you’re producing will be less valuable.
The video says “To avoid this, break projects down into the smallest possible deliverable stages, so the project is always in some kind of usable state, even if that state isn’t fully ideal.” I translate that as working iteratively on projects and even individual pieces, having it presentable at many stages. For example having the sketch presentable enough to post, even if I have to do a little extra cleanup, rather than just quick shorthand that only make sense to me. Then if I get to the stage where I want to move on, it’s in a state where I can without wasting all my effort.
Unnecessary Tasks
Solution: Don’t?
Deciding you don’t need to do something before starting is a great way of saving unnecessary effort, and frees up time for things that do matter.
How do you decide what matters? “[Does] that feature really push the vision of the game in a way players will care about?” To me this sounds like making a conscious effort to avoid putting unnecessary detail into things the viewer won’t even notice. Does the piece really need that complex an environment, or would it have the same impact with something easier but just as well executed? Is this 5th iteration of the arm that much different to the first? Do you really need to detail the hair any more or is the hint of smaller shapes enough? Having a clear idea of your intent before starting sounds like the best way to answer these questions.
I can also see it on a macro scale, and take it to mean evaluating whether the projects I’m tackling truly get me closer to my goals, or are simply distractions masquerading as progress.
HEROICS
Heroics are “when unreasonable acts are required to deliver a result”, and are a sign that all your processes have broken down. As Seth says, they are the worst kind of waste because they cause almost every other type, and also perpetuate themselves. For me, this is staying up late to finish work. In the context of any other job, this is an unreasonable request, yet I still do it on the regular because it’s on my own time.
The only way to stop heroics is to make the decision they have to stop. Halt production and put in place systems and processes to make sure they don’t happen. Refuse to crunch.
Setting Up Workflow Management
The processes Seth set up for his company are outlined in the video, so I will present my own version I came up with after watching it. I like to call it a Task Funnel. It’s quite similar to Agile as it takes all the things you have on at the moment, breaks them down into small pieces then feeds them out in manageable chunks with the most important things coming first. It’s substantial enough that I think it deserves its own article, but the basic process is:
- Set some high-level goals you’re working towards.
- List everything you are doing or want to do.
- Prioritise based on how they further your goals.
- Split your higher-priority objectives into smaller tasks.
- Pick a number of tasks you could reasonably achieve in 2 weeks.
- Arrange those tasks into 16-hour “sprints” to complete over 2 days.
- Work through them, taking a little time to review and adjust between 16-hour sprints.
- Review how the 2-week period went. Did you achieve your goals? What would you change for next time? Do any of your priorities need to change?
- Repeat from step 5.
I use Quire as pen-and-paper would get unwieldy fairly quickly, and it’s free, but there are a lot of other project management software options out there if that one doesn’t sit right with you. The funnel takes a bit of effort to set up but the time and stress you save in the long run is definitely worth it. It also serves as an archive so you don’t forget projects you want to work on.
More translated tips
Here are some more tips from later in the video translated to be useful for solo art work.
Check production workflow for bottlenecks, and see if there are any ways to increase throughput. If you make a bunch of good sketches, but only produce one finished piece out of them a month, perhaps more time should be spent on later stages to get things moving through the whole pipeline faster. Or maybe you avoid those stages because you don’t enjoy them, in which case you could work to change that, adjust your process to avoid it or make it less painful, or hire/team up with someone to take care of it for you.
If there’s a part of the process that hurts, do it more not less, so you can learn about it and fix the issues. When you find things to fix, start simple to get a solution working immediately. You can refine it later.
This sounds like good vague advice so apply it as you will, but it makes me think of deficiencies in fundamentals. Lean into things you shy away from, get simple fixes to get them “good enough” and implement them immediately.
Amplify feedback loops. Seek feedback wherever you can to better serve your audience/customers/clients, and ideally automate the collection process to make it painless and smooth. Feedback is a great way to identify problems, but much less reliable for solutions, so use your best judgement. Asking for feedback after you’ve completed a job, or during if it’s long term, could bring up little things to improve in the future. See next:
Always be learning, and “create a work culture” (or personal attitude) where that is accepted. Always be open to new things, and look for new ways to refine processes to make things faster and easier for yourself, so you can spend the most time possible Making Art.
Signing off
There’s a lot here, but hopefully some things have stood out and will stick with you, whatever your path is, to help make your work Stress-Free. Whether that’s keeping to limited work hours to make them more valuable, minimising waste however you can, or making your workflow more efficient with a task funnel or eliminating bottlenecks, there are better processes out there for you.
Best of luck on your journey.
