I recently changed roles, moving from Faculty (recently acquired by Accenture) to Arva AI. I interviewed at a few places before deciding on Arva. My experiences working at Faculty; talking to a range of design teams at different firms; and now working at Arva have given me a range of insights into what’s happening in the Product Design market in 2026. They resonate with Tom Scott’s design hiring observations.
This is quite niche to app and website design in London 2026 and so may be quite narrow, however this is my experience and might be interesting to other designers working or looking for work in the industry.
Without question, AI (or more specifically, the introduction of LLMs) has changed the game. Every role I interviewed for was about AI use, both in process, and how to enable AI for end-users. I have written previously about the new design process.
Management is dead. Long live management.
Companies are looking for Individual Contributors (ICs). From global consulting and strategy firms through to startups, firms want outcomes. That means builders, doers, ICs. There are middle management roles out there, but the expectations have risen to make sure that everyone is focussed on outcomes rather than pure people management plays. Director-level managers are too expensive to not send out to clients. If you are not regularly billing for your time, talking to customers, or shipping code that affects the company’s bottom line, there is a target on your back. This can happen to anyone, at small companies or large FAANG style companies.
Brand vs Product
The roles available have also morphed. The 2 pathways that have risen to the top of the shakeup: brand designers and product designers. Brand being more marketing and focussed on creating graphic impact in digital/social channels. Product, focussed on making good-looking software that make sense, is easy to use. A blending of User Experience (UX) and User Interface (UI).
Both of these roles expect a high level of AI use. The teams I have spoken to are expecting more AI-pilled designers – pretty much all looking for designers using gen-AI to generate code. For brand designers, this means building tools to generate assets like social posts and design variations. For product designers, this means shipping frontend code and raising PRs.
Design engineer is a role that’s thrown around a bit at the moment, but I haven’t seen many open roles for it. It’s a new space and not particularly well established what their responsibilities would be.
Where has the risk gone?
Traditionally (i.e. pre-GPT, circa 2022), UX and UI designers would spend more time planning and experimenting to set an app’s direction. De-risking the slower, more expensive engineering time needed to build. It’s also harder to change direction when code has been written. With both engineering becoming faster and cheaper to build frontends this de-risking step has moved towards the other end of the building process: fixing confusing workflows and redesigning visuals of apps that already exist. The skills needed here are taste in the broadest sense, and judgement. You will often be given a prototype by someone else and asked to build on top of it. Before you might have been turning vague user suggestions / client requirements into experiments and guesses that could work, 0–>1.
This “taste” is manifesting in the expectation for designers to be more involved in product management. Being involved in product strategy and decision-making is expected, even taking on planning and direction work.
Craft
For designers, the craft bar has risen considerably. There is still a vital importance on the fundamental understanding of typography, layout and colour. However hiring teams expect you to match their level of obsession. They expect you to be able to raise their internal bar. This has arguably always been the case, however I think it has proliferated and extended to much wider places. Maybe because of the access to AI has made left little excuse for bad interfaces. Maybe people are now exposed to more good design themselves and can point at other people’s apps and say that’s what they want.
Tasks
I had a couple of tasks to do – good tasks were research, exploratory or vibe-coded based with a 20 min presentation or 10 minute video to summarise. Bad tasks were along the lines of “create a Figma prototype inventing a new AI interface paradigm in 2 hours”. I did not start that task. Another unique type of task was based around doing a mock workshop for an hour exploring a hypothetical scenario, investigating a problem and coming up with a solution. Hard to get the jib of that one.
Long live democracy
All designers have access to these tools and the democratising of generating frontend code has also raised this bar – demonstrated in people’s portfolios that all now come with videos, animations and “fun” interactions. As always, the portfolio is king. Here are some tips to help you get hired which I wrote while hiring at Faculty. Hiring managers are busy and don’t want to waste time downloading PDFs, rummaging for passwords or trying to get around Figma glitches. This is the best place to show off your skills (of shipping code) and taste of design. Showing you can connect a domain name also shows you are technically minded enough to work with engineers and not just throw Figma designs over the proverbial engineering wall.
What can I do?
I feel for you if you are stuck in application hell. In my experience, I received hardly any callbacks from LinkedIn applications, however talking to recruiters and improving my portfolio from feedback and continuous self-improvement was a key part of the process. Ask for help: ask your design friends and colleagues for feedback. What do they like. What don’t they like. What looks good. Your portfolio is your gateway to interviews so imagine yourself as the hiring manager “experiencing” your portfolio for the first time.
Look at other people’s portfolios, practice your craft, and keep going.
“Look at other people’s portfolios, practice your craft, and keep going.”
