I have been working in design long enough to have watched the discipline reinvent itself more than once. I go back far enough to remember designing for CD-ROMs, before websites were even a thing, fitting an entire interactive experience onto a disc and hoping nobody scratched it. Then websites arrived with a whole new set of adjustments, and I worked closely with developers, through several rounds of iteration, to get designs working across browsers that seemed to disagree with each other purely out of spite. From there I watched the field split and mature, as visual design gave way to a much deeper focus on user experience, on how something felt to use rather than just how it looked. Somewhere in between, an entire industry of design tools showed up to make the work faster, smarter, and more collaborative, and I learned a new one every time it seemed like the last one had finally, truly, promised to be the last one.
None of that prepared me for AI specifically, but it did teach me what it means to adapt. AI is the next version of that same pattern, and if there is one thing decades in this industry teaches you, it is that the designers who last are the ones who have learned, repeatedly, how to fold a new capability into the way they think and work without losing the judgment that got them there in the first place.
Output is no longer the measure
There was a time when a UI/UX designer working inside the Marketing Cloud ecosystem could measure their value almost entirely by output. How many email templates could you turn around in a sprint. How many landing page variants could you mock up before a campaign deadline. Speed and craft were the job, but that is no longer true. Generative tools can now produce a passable email layout in the time it takes to type a prompt. What they cannot do is know whether that layout actually reflects the brand, or whether the journey logic respects how a specific customer segment behaves. That gap, between what AI can produce and what a business actually needs, is where the modern Marketing Cloud designer now lives.
From producer to decision-maker
For years, the implicit deal was that a designer’s time went into production and a bit of design thinking got squeezed in around the edges. Now the ratio has flipped. Production has become cheap and fast, almost incidental. Judgment has become a scarce resource. A designer who can glance at an AI-generated template and immediately spot that it violates brand spacing rules or will clip badly in Outlook is worth far more than one who can build the same template by hand a little faster than their peers.
This does not mean designers should hand over the keys and stop paying attention. If anything, the opposite is true. When an AI tool suggests a layout, a content block, or a journey path, and a designer approves it without scrutiny, mistakes move downstream fast. In the Marketing Cloud world, an ignored accessibility flag or a broken render in a major email client does not stay a small problem. It reaches thousands of subscribers before anyone notices, and by then the damage to trust and deliverability is already done. The designer’s real job now is to be the last check before something goes live, not the first hand that builds it.
Discovery gets your full attention back
Discovery work used to mean sitting through stakeholder interviews and workshops, typing furiously to capture every requirement and every pain point that might matter later. That kind of split attention, half listening and half transcribing, was never ideal, and AI has solved it. Meeting assistants now handle transcription and can summarize a session, pull out personas, and flag journey stages or edge cases that might otherwise get missed.
What that frees up is attention. Instead of worrying about capturing everything, a designer can actually listen and understand. They can watch how a stakeholder describes a customer pain point and ask the follow-up questions that AI has no way of knowing to ask.
Wireframing and journey mapping as a starting point, not a finish line
Building wireframes for emails or a multi-step Journey Builder flow, then iterating through rounds of feedback, used to eat entire days before anything close to final even existed. AI has turned that first draft into something closer to instant. A designer can describe the goal (a re-engagement campaign for lapsed subscribers, for example) and get several structural directions back within seconds.
The value here is that a designer no longer starts from a blank page. They start from a set of rough directions they can critique and shape around what they actually know about the audience and the brand.That reaction is design work.
Visual and content design becomes an editing discipline
The same logic applies to building out templates and dynamic content blocks. AI can now draft multiple template variants and suggest personalization rules based on the segment data available in Marketing Cloud, often producing more options in a few minutes than a designer could reasonably review in a day. The real work becomes choosing well: trimming what does not fit and tightening what does.
Recognizing which of five AI-generated subject lines matches the brand voice, or which content block variant will resonate with a specific segment based on past campaign performance, requires exactly the kind of contextual assessment that generative tools do not have. A colleague put it well recently when they pointed out that AI cannot have taste. It can produce plenty of technically correct options, but it has no instinct for which one is actually right for this brand and this audience. That is still entirely human, and it is quickly becoming the more valuable half of the design process.
Quality assurance gets a second set of eyes that never gets tired
Checking contrast ratios and rendering behavior across a long list of email clients has always been tedious and easy to get wrong under deadline pressure. AI tools are well suited to exactly this kind of repetitive, rule-based checking. They can flag accessibility issues and likely rendering problems before a campaign ever reaches a tool like Litmus or Email on Acid, catching mistakes earlier and with more consistency than a tired designer running through a checklist at the end of a long day.
This changes the designer’s role from the one running every check manually to the one deciding what those flagged issues mean for this campaign and this brand. For example, a flagged contrast issue might be a real problem, or it might be an intentional design choice that needs a second opinion rather than an automatic fix.
What still belongs to the designer
None of this points toward doing less design work. It points toward doing less of the repetitive part and more of the part that actually required a person in the first place. AI can generate a layout or a journey structure, but no matter how many prompts you throw at it, it cannot handle brand politics or understand why a customer experience needs to feel a certain way rather than function correctly. That understanding still belongs entirely to the human in the room.
The designers who will do well in this next stretch are the ones who stop measuring their value by how much they can produce and start measuring it by how well they can judge what gets produced, whether by them or by a tool. Building that instinct takes the same thing it always has: attention, curiosity, patience, and a genuine understanding of the people on the other end of every campaign.
