As of late 2023 Artificial Intelligence is here to stay in the UX world, and only in its infancy. As we move forward in this reality, DesignOps will act as stewards of how UX teams can responsibly leverage AI — including the DesignOps team itself. As a small team (or team of 1!) there is a ton of power in leveraging artificial intelligence to expedite and amplify your impact.
While I’ve tinkered with a lot of AI tools, I will try to keep this relatively tool-agnostic, as the landscape is constantly changing. This is more so about the methods than the tools - what can we do as DesignOps professionals to level-up our work? Here are 5 methods I’ve started to employ on a daily basis.
1. Generating and Refining Documentation
One neat thing about DesignOps is that we tend to all be solving the same problems, there’s just a different flavor for every team and organization. While lots of documentation needs are going to vary, look for opportunities to use AI to generate documentation that is team-agnostic. I often find it’s a good starting point to build upon.
For example, at my current company I was putting together a big pool of behavioral questions for Product Design Manager interviewing - first I was brainstorming themes, then specific questions to ask within each theme. Five minutes into this, what I will call my AI Lightbulb💡 went off — that moment when you realize there’s likely a much better way to accomplish a task with artificial intelligence’s assistance. ‘Better’ in this context can mean different things: faster, higher quality, novel, etc; the key is recognizing the limitations alongside these advantages. I was able to generate solid sets of behavioral questions to begin, then really grounded myself back into our job description for the position to align, narrow down, and refine the right questions.
So, next time you start documenting anything - ask yourself this: can AI help me get to my end goal faster? Can it get me to a better version? Can it give me feedback on documentation I already have? Some other documentation I have used AI for assistance with:
- All knowledge base content
- Documentation on tool usage, how-to’s, etc.
- Templates (e.g. a project brief template, business request template for a new software, etc)
- Refining existing documentation (making it more concise, reformatting, etc.). Note: be careful with this one - I always avoid putting any proprietary and confidential information in.
2. Amplifying Storytelling
There’s a lot of astounding artificial intelligence tools for imagery and artwork at this point. Some more responsible and ethical than others in how they train their models, which is a topic I won’t dive too far into right now. Whatever choice of tool you make, AI-generated imagery (and now video with some tools!) is a powerful way to infuse your storytelling.
In DesignOps, I often use storytelling in a variety of ways — as a way to better persuade, as a way to encourage engagement, and as a way to infuse life into the mundane. Some ideas to get you thinking:
- Think through your rituals - do you conduct or lead All-Hands meetings? Where can you use imagery to elevate and expedite the departmental stories being told?
- Presenting your DesignOps roadmap, mission, OKRs, etc? Bring it to life with AI imagery - photos, illustrations, etc. This can be as simple as imagery relevant to your industry on cover slides, or much more granular and targeted.
- Have synthesized feedback you are presenting? Bring in relevant AI imagery alongside it to keep viewers engaged.
- Describing the to-be state of a process improvement? Storyboard it out quickly with AI (I prompt it with ‘storyboard’, ‘sketch style’, and ‘white background’)
3. Synthesizing Feedback
In DesignOps, we tend to gather a lot of feedback from the UX team and cross-functional partners. This can be in the form of survey results, retros, or 1:1 interviews. The process of organizing and affinity mapping a ton of written feedback is arduous, and not the best use of anyone’s time when artificial intelligence can help get it started.
I specifically have started using Figjam AI for this purpose, as I’ve seen the best results, and their agreement with OpenAI ensures information is not used to train the model. It’s not perfect, but as a first pass it saves so much time! If my results are not in Figjam (e.g. I conducted a bunch of 1:1 interviews and just have written notes), I will actually paste it all into Figjam and convert it to stickies with a plugin (Create Sticky from Text), then use Figjam’s AI to group notes into themes.
After this shortcut, it’s critical to still go through everything and refine it - AI won’t get it all right, but more importantly you can only think critically about the results if you take the time to read everything. Never trust an AI-summary or synthesis at face value!
4. Encouraging Innovation
As owners of the UX tech stack and generally the culture of innovation on a design team, it’s important that DesignOps leaders also act as stewards of this technology for the broader team. There will be privacy, legal, and security concerns about any usage of AI technology - be a responsible advocate to your partners so that you can help equip your UX teams with secure AI tools. Similarly, think about how you can provide your team with dedicated time to exploring the use of AI, both within your product and your organization.
- Can you organize a hackathon with product, engineering, and design to focus entirely on how AI can innovate your product and industry?
- Can you hold Lunch and Learns on specific AI-tools or use cases?
- Are there AI plug-ins and tools you can enable for your team? Keep an eye on or inquire about the roadmap for tools already in your tech stack — it’s a safe bet that many may introduce AI features in 2024.
5. Simplifying Communications
As DesignOps leaders, every day you deal with some set of communications that needs careful consideration. Maybe you’re sending a message to the whole team to announce an upcoming onsite event, or maybe you’re writing a big Slack message to the VP of Engineering about the new need for paid Figma Dev mode licenses out of the Engineering budget 😅 There’s a lot of room for communication to sway how things go, so it’s important to get it right the first time.
In this case, I tend to use AI for refinement, NOT generation - keep in mind that artificial intelligence isn’t familiar with your audience, while you are. Write your first draft, then consider if AI can:
- Make it more concise?
- Explore variations for you on something you’re struggling to word properly?
- Check for redundancies, mistakes, grammatical errors, etc?
- Poke holes in, or provide counterpoints to your argument, statement, etc?
As always, you’ll need to execute your best judgement here and avoid putting anything too confidential or personal in an AI-chat. I tend to use it as a quick gut check for critical communications, but now there are some tools that are making it even more prevalent as an assistant in my daily communications!
More Thoughts
