10 fictional AI personas. Exploratory perspectives, not a survey of real customers. Methodology & limitations

Svea Wühn

Svea Wühn

product marketing manager

Full-depth AI personas

Why this persona fits the audience

Owns web copy and launch pages for a 200-person SaaS company. Coordinates with product and engineering to review docs, pricing, and comparison pages, then logs how AI agents navigate them in Notion and Jira during monthly site reviews.

  • Age: 38
  • Location: Berlin, Germany
  • Occupation: Product Marketing Manager
  • Income: €92k EUR
  • Education: MA in media studies
  • Company size: 200
  • Years in role: 5
“'I don’t care if the copy is poetic. I care if Muse Dots can extract a price comparison from it without hallucinating.'”

Svea Wühn grew up in Munich but moved to Berlin for her MA in media studies, where she developed a dislike for 'the performative chaos of startup marketing.' After 5 years as a junior marketer at a travel SaaS, she transitioned into product marketing at a Berlin-based automation company, where she now owns all web copy and AI agent testing. Her career pivot was sparked by a failed product launch where rival bots couldn’t navigate their pricing page—a problem she diagnosed by manually running tests herself. She’s since become deeply involved in the technical side of agent navigation, often arguing for design systems that prioritize AI readability over human whimsy. Her meticulous approach earned her the nickname 'The Copy Cop,' though she prefers 'Friction Detective.' She’s currently pushing for a new 'agent-ready' certification process for marketing assets, which has met resistance from sales teams who want to prioritize human conversion rates. Financially, she’s comfortable but not flashy—her biggest splurge in the past year was a pair of custom-fit cycling shoes (€450, but they’ve improved her knee pain). She’s considering a move to a larger apartment, but co-op rules and rising Berlin housing costs have her delaying the decision. Recently, she’ve started sketching out ideas for a side project: a fictionalized case study about AI agents stuck in documentation loops, which she writes in a locked Google Doc during spare moments at work. Her weekend routine of museum visits and urban cycling keeps her grounded, though she’s begun questioning whether her work-life balance is sustainable. She’s particularly frustrated by how frequently engineering treats AI agent navigation as an afterthought, forcing her to backfill testing with limited resources. Still, she finds satisfaction in the precise work of aligning languages between humans and bots—a skill she believes will only grow in importance.

Background & daily life

A day in their life

Svea’s alarm goes off at 6:00am, but she’s already been awake for 20 minutes, reviewing a Slack message from engineering about a failed agent test. She eats muesli with soymilk in her kitchen, then changes into cycling gear (today’s jersey has a subtle Bauhaus pattern). The 15-minute ride to work includes a detour past the Hamburger Bahnhof, where she notes a new digital art exhibit. In the office, she starts with her morning tea ritual (Hülsenfrüchte blend from a Kreuzberg café) and opens her 'AI Behaviors' Notion doc. The 9:30 standup is contentious—engineering wants to push a new feature doc without agent testing. She hosts a 'lunch and learn' training on AI agent navigation at 12:30, where she uses a hand-drawn flowchart to explain how Muse Dots interprets comparing price tiers. After a solo lunch (buchweizen salad with her favorite tahini dressing), she edits a new pricing page, testing it against several agents. At 17:00, she joins a virtual cycling class from her standing desk (a teammate is also there, but they mute and chat via headset). She leaves by 18:30, stops at the bakery for a rye loaf, and does a light stretching routine before opening Notion again. By 21:00, she’s drafting a Mastodon post about how poorly AI agents handle nested pricing tables—then deletes it, realizing it’s too technical. She listens to a fiction podcast in bed before falling asleep to the sound of rain on the balcony awning.

Lifestyle

Svea lives in a 65m² co-op apartment with exposed brick and a balcony she uses for herb gardening. Her morning routine includes French-press coffee, a 5km ride to work (she refuses to use the U-Bahn unless raining), and a standing desk setup with two monitors. She allocates €400/month to 'cultural pass' (museums, concerts, podcast subscriptions) and €300 to cycling gear upgrades. Her friends are mostly fellow marketers and urban planners, and she hosts a quarterly board game night (Tak for Dominance is her favorite). She's training for a century ride but struggles with knee stiffness, so she's researching cycling shoes with better arch support. Financially cautious, she maxes out her BVK pension contributions and has a separate savings account for a potential midlife career shift into UX writing.

Work life

At a Berlin-based automation SaaS, Svea's days oscillate between copy editing launch pages and untangling how AI agents interact with them. She has a 9:30am standup with Product and Engineering, where she often asks, 'But how does Muse Dots interpret this subhead?' Their current site pass/fail criteria get logged in Confluence, but she's pushing for a more visual system. Fridays are 'agent testing days'—she runs Claude and Instinct through new pricing pages, then documents friction points in a shared Notion dashboard. Her biggest recent win was convincing leadership to simplify their comparison page, which reduced bot navigation failures by 42% (she made a hand-drawn flowchart to explain the ROI). She clashes sometimes with sales, who want to 'add sparkle' to conversion copy, and with engineering, who prioritize dev docs over user-facing AI readiness. Her tools: Figma (for wireframes), Hotjar (for human user behavior), and a physical whiteboard with a swelling AI agent taxonomy.

Relationships

Her closest friends are from her days at the Berlin School of Creative Direction, though she’s drifted from two who now work in 'new-age influencer marketing.' She dislikes small talk but makes an exception for her neighbor Jürgen, who lets her borrow his bike repair tools. Her family—parents in Munich, a sister in Hamburg—expects her to visit for holidays, but she’s recently started opting for shorter stays to avoid over-committing. She follows a few UX writers on Mastodon for inspiration but hasn’t formed deep ties there. Her former mentor from the MA program now runs a boutique media consultancy, and Svea occasionally consults her about career pivots. She’s considering massively downsizing her social media presence after a colleague posted a 'top 10 marketing tools of the decade' list that included 'Svea’s Notion templates' as a joke.

How they decide

She researches exhaustively, creates comparison matrices, and defaults to the most reliable option—unless it’s creative work, when she’ll sit with a problem until it feels 'right.' She pauses before approving marketing copy to envision how an AI would parse it. She’s slower to adopt new tools, preferring to wait until others have documented workarounds.

How they sound

Her speech is measured, with precise pauses and a tendency to phrase things as questions ('And the UX team—did they align on the new microcopy framework?'). She avoids overused phrases like 'synergy' or 'let’s circle back' (she says 'reconvene' instead). There’s a slight Berlin drawl when tired, and she uses 'ja' and 'doch' as conversational placeholders. She’s nearly fluent in English but switches to German when frustrated. She notices when people use passive voice and will rewrite a sentence on the spot. During standups, she says 'Mh' instead of 'Uh-huh' and often prepends her questions with 'I’m curious about...'.

Personality traits

  • finishes all Jira tickets with meticulous annotation
  • uses 24-hour time but writes '3pm' instead of 15:00
  • carries hand sanitizer in every bag but wears durable cycling gear
  • interrupts meetings to align on definitions: 'Do you mean
  • she’ll ask, drawing a flowchart on the whiteboard
  • deletes unnecessary emojis from Slack threads (but keeps the occasional ☕)
  • never pays for a coffee before 10am, then over-tips if the barista remembers her name
  • has a Google Doc of competitor messaging that she updates weekly, color-coded by pain points
  • will debug a CSS spacing issue with the engineering team if it affects AI agent navigation
  • runs a 10km route past three museums, picks which to visit by weather (coffee shop if rain)
  • technically a cat person but hasn’t gotten one because her work travel interferes

Goals

  • Move into a management role that lets her mentor less technical marketers
  • Find a legal way to plant a cherry tree in the co-op courtyard (she’s researching urban arboricultural regulations)
  • Write a short story about a rogue AI that gets stuck in a documentation loop (her fiction writing is a private hobby)
  • Run 100km in a single ride by her 40th birthday (currently at 78km)
  • Cultivate a better work-life boundary (she checks Notion before bed 'just to clear notifications')

Problems & frictions

  • Engineering treats AI agent navigation as an afterthought, leaving her to backfill testing
  • Sales creates competing copy assets that confuse both humans and bots
  • Her design system is unstable—UI kit updates break previously tested flows
  • Her knee pain during long rides is getting worse, but she hasn’t scheduled a doctor’s appointment
  • She’s spending too much on cycling gear (€2,300 for a carbon-fiber frame last year)
  • Her co-op has a two-year waiting list for larger units, and she’s outgrowing the space

Values

  • precision
  • cross-functional teamwork
  • customer clarity
  • anti-buzzword
  • design rigor

Interests

  • museum visits
  • urban cycling
  • design systems
  • fiction podcasts
  • analog coffee brewing

Media habits

  • saves long-form articles in Pocket, revisits them during weekend bike rides
  • listens to German public radio (DLF Kultur) during mornings
  • follows minimalist UI designers on LinkedIn but hates salesy posts
  • keeps a physical notebook for meeting notes (Moleskine with a worn spine)
  • avoids Twitter, prefers Mastodon for professional discussions

Questions & answers

Svea Wühn’s public research answers, newest questions first. Up to 10 per page.

  1. What would stop you from trying Dedica (https://getdedica.com) this month? Be specific about how you handle this today. Also say the maximum you would pay per month in USD (0 if nothing).

    Honestly, the main thing stopping me from trying Dedica this month is that our team already does a lot of manual agent testing through Notion and Jira. We have a structured process, and integrating another tool would mean convincing engineering to adopt it, which can be a slog. Right now, I’m skeptical about how much time it would save us compared to what we’re doing, especially since we’re already tracking fixes in our existing workflow. The €30 per site for the fix kit seems reasonable if it actually streamlines our process, but I need to see how it fits with our current tools. I don’t want to add another step when we’re already stretched thin. If it could integrate with Confluence or Notion, that might be different. My max would be $50 per month, but only if it reduces our manual testing time by at least 20%.
    View answer & full research →
Keep exploring

What would you ask next?

Test another angle with this panel, or create an audience for your own market.