You have three published pillars. Every company on your list could plausibly hear any of them. The sorting logic most candidates use is industry: AI company gets AI work, enterprise gets scale work, regulated platform gets compliance work.
That logic breaks at Director+ because the industry label doesn't predict what the room is afraid of. OpenAI's Product Design Manager posting tests trust calibration across probabilistic outputs. Salesforce's VP Product Design posting tests platform cohesion inside what they're calling the "Agentic Enterprise." Both involve AI. They evaluate completely different things. The failure mode is showing the right evidence for the wrong fear.
Run the fear check before you write the message.
Quick-Reference Matrix
| Buyer Anxiety | Lead Proof | Supporting Proof | Hold in Reserve |
|---|---|---|---|
| Probabilistic-system trust (OpenAI, Suno) | Trust essay + Agentic Labs (Brand Pulse, Carrier IQ) | Alibaba trust architecture; Thermo Fisher consequence weight | Equinox+, Allē |
| Org operating model change (Capital One archetype) | Trust essay as operating framework + Agentic Labs as built proof | Alibaba enterprise credibility; Thermo Fisher regulated navigation | Alibaba/Red Cross/Thermo Fisher deep-dives |
| Automation under consequence (Headway, Gusto) | Red Cross (6→1, federal oversight) or Thermo Fisher ($20M margin, pharma) | Allē dual-surface; Agentic Labs automation fluency | Alibaba scale narrative |
| Design inside agentic ops (Ramp) | Agentic Labs live agents + Trust essay threshold framework | Red Cross multi-role architecture; Thermo Fisher agentic vision | Alibaba |
| Enterprise-scale consistency (Salesforce, Adobe) | Alibaba ($50B GMV, trust architecture across surfaces) | Allē dual-surface consistency; Thermo Fisher multi-partner system | Agentic Labs solo builds |
Anthropic fits probabilistic-trust but has no confirmed live design-leadership posting as of this week. Atlassian fits enterprise-scale consistency but no Director+ design role was confirmed this pass.
Adobe GenStudio straddles two fears. If the conversation centers intelligence-layer design and AI-agent interaction patterns, run Play Card 1. If it centers systems governance and cross-product consistency, run Play Card 5. Check the posting language or the interviewer's title to decide.
When the room shifts. These fears don't always arrive clean. If the questions pivot mid-conversation from one anxiety to another, the Hold in Reserve column is your pivot inventory. That column exists for exactly this reason. You aren't locked into the lead proof once you've started.
Play Card 1 — Probabilistic-System Trust
Who this is. Hiring managers and technical co-founders at AI-native companies where model outputs are non-deterministic. OpenAI, Suno. Adobe GenStudio also fits here when the role emphasizes "intelligence-layer design" where AI, data, and human creativity intersect.
What they're actually evaluating. Whether you understand AI as a material with properties. OpenAI's posting asks for trust-building and consistency across complex systems. Adobe's GenStudio posting asks for prototyping and experimenting with emerging UI patterns including agent handoff, and measuring "adoption, value, and trust." Salesforce's AI Design research role studies how people "think, interact with, and trust AI systems." The decoded question: Can this person design the communication of intent so interpretation produces the right outcome? Agent autonomy thresholds are the rarest credential at this level. Most candidates can talk about AI features. Almost nobody can articulate where agent autonomy ends and user control begins from shipped work.
Lead with. Trust Is the New Interface, then Agentic Labs. The essay is a stronger opener than any case study for this room. The five-handoff framework and the trust ladder (Watch → Verify → Delegate) give them shared vocabulary before you show a single screen. Follow immediately with Brand Pulse or Carrier IQ as built proof.
Support with. Alibaba's trust architecture. "A buyer committing six months of inventory spend on the strength of a screen" translates trust-at-scale into language this room respects. Thermo Fisher adds consequence weight.
Cut or hold. Equinox+ and Allē. Consumer product optimization reads as a different altitude here.
IC+manager frame. Tilt hard toward craft. Show the agent autonomy threshold decisions in Carrier IQ. The solo-build question will come. The essay is the leadership artifact. The agents are the proof it works.
Landmine. "These are solo projects. Can you lead a team doing this?" Reframe: "I built them solo to pressure-test the framework I'd published. Scaling that across a team is the role I'm here for."
Play Card 2 — Org Operating Model Change Under AI
Who this is. CDOs or SVPs Design at large companies hiring someone to change how design work gets done with AI. Capital One's Director, Product Design — AI in Experience Design is the archetype: the role owns AI tooling, process evolution, and talent upskilling across the entire design function.
What they're actually evaluating. Whether you've lived the org transformation they're trying to install. Capital One's posting makes this explicit: the role owns "AI-enabled design process evolution" and "AI-centric talent upskilling across Design." Salesforce's VP posting asks for "cutting-edge agentic design and vibe-coding tools" alongside executive communication and organizational accountability. These are operating-model mandates. The buyer is navigating this transition themselves and needs someone who's already lived it. Confidence level on the evaluation pattern: high for Capital One (posting language is direct), moderate for the broader archetype (inferred from McKinsey's finding that top financial performers embed design in top-team deliberations and measure it with revenue-level rigor).
Lead with. Trust essay as an operating model document. The five-handoff framework becomes a design-practice operating model: how do you structure review, QA, and decision-making when the output is probabilistic? Follow with Agentic Labs as proof you've built the workflows, not benchmarked them.
Support with. Alibaba for enterprise org credibility (cross-functional mandate, executive alignment). Thermo Fisher for regulated navigation. Both secondary. Don't let them become the presentation.
Cut or hold. Deep case-study walkthroughs. This room doesn't need your interaction design decisions on a PDP page. They need to hear how you'd structure agentic-native design practice at scale.
IC+manager frame. Tilt entirely toward organizational thinking. They already assume you can design. What they're trying to figure out is whether you can redesign how 50 or 200 designers design.
Landmine. "Have you actually transformed a design org?" Be honest: "The essay is the transformation playbook. The agents are the proof-of-concept. What I haven't done is apply it at your scale, which is exactly why this conversation is interesting to me." Honesty here builds more trust than overreach.
Play Card 3 — Automation Under Consequence
Who this is. Design and product leaders at regulated platforms where automation errors carry real-world cost. Headway (clinical workflows, AI scheduling, transcript-powered notes). Gusto (payroll automation). Compliance-adjacent stakeholders often sit in these rooms.
What they're actually evaluating. Whether automation in your hands reduces burden without eroding accuracy or compliance confidence. Gusto's posting states this directly: the role owns "automation and AI initiatives that reduce manual work without compromising accuracy or trust" in payroll. Headway's posting layers AI scheduling and copilot features on top of insurance verification and billing. The decoded question: Will this person dissolve constraints into the product, or layer process on top? Embedding FEMA compliance into the transaction workflow so it never becomes a review gate is fundamentally different from adding a compliance review step.
Lead with. Red Cross or Thermo Fisher, depending on which constraint profile matches. Red Cross for speed-under-federal-oversight (6 systems→1, national deployment, 6 months). Thermo Fisher for consequence-weight ($20M margin, exceptions surfacing before the delivery gate).
Support with. Allē for dual-surface platform evidence. Agentic Labs for automation design fluency, positioned as "here's how I think about agent autonomy in high-stakes contexts."
Cut or hold. Alibaba's scale narrative. Right evidence, wrong fear. This room needs to know you can ship automation a caseworker trusts on day one without training.
IC+manager frame. Balanced, roughly 60/40 craft-to-leadership. Show the system-design decisions (Red Cross multi-role architecture, Thermo Fisher five-module structure) alongside the stakeholder navigation (32 warehouse interviews, 6 pharma partner adoption).
Landmine. "How fast can you actually move in our regulatory environment?" Show that your default mode dissolves constraints into the product rather than layering review gates on top. Red Cross: national deployment in six months under federal oversight. If you can ship under FEMA compliance in six months, you can ship without it in six weeks. The reverse has never been reliably true.
Play Card 4 — Design Inside Agentic Operations
Who this is. Product and design leaders at companies where agents are the operating model. Ramp's Director, Product Design posting describes designers "building tools and agents, designing memory, shipping PRs, and rethinking process as new capabilities unlock."
What they're actually evaluating. Whether you can operate where the design surface itself is agentic. Salesforce's Design Systems posting frames this as building "component metadata schemas, semantic ontologies, structural specifications, guardrails, governance models, and trust criteria for AI-driven UI composition." Old product boundaries between internal tooling, external product, and agent behavior have collapsed. Read Ramp's posting language: "building tools and agents, designing memory, shipping PRs, and rethinking process as new capabilities unlock." That verb list is the signal. Building and shipping before process.
Lead with. Agentic Labs live agents. Brand Pulse, Carrier IQ. These prove you've built and shipped in the material they work in daily. Follow with the Trust essay's decision-gate framework, specifically the autonomy threshold.
Support with. Red Cross multi-role architecture (one record, multiple role-appropriate views, permission-aware visibility). This pattern maps directly to agentic operations where different agents and humans see different slices of the same system.
Cut or hold. Alibaba. Enterprise pacing reads as a different operating speed for this room. Keep for Q&A only.
IC+manager frame. Tilt toward craft. Ramp wants someone who ships PRs and builds tools. The management layer comes through the Trust essay's framework, which proves you can systematize judgment at scale.
Landmine. "You've been Head of Product at an enterprise company. Can you work at our speed?" Change what they're measuring: "Speed without consequence is one skill. Speed under constraint is rarer. I've shipped a national platform under federal oversight in six months and solo-built three live AI agents. I don't import process that slows you down."
Play Card 5 — Enterprise-Scale Consistency
Who this is. Design leadership at platform companies where experience quality must be repeatable across products, teams, and governance layers. Salesforce (VP Product Design, Informatica, reporting to SVP of Salesforce UX & Product Design). Adobe GenStudio when the emphasis is systems thinking across micro-interactions and enterprise governance. Atlassian fits the archetype but has no confirmed live Director+ design posting this pass.
What they're actually evaluating. Whether you can make quality repeatable without being in every room. Salesforce's VP posting asks for "modular, reusable design patterns that support flexibility, extensibility, and long-term platform cohesion." Their Design Systems posting frames this as evolving from a component library into AI infrastructure for safe composition at runtime. The decoded question: Can this person build the system that produces consistency?
Lead with. Alibaba. Full arc: named the structural gap at $50B GMV, built the research case, secured executive mandate, led three cross-functional sprints across homepage, search, and PDP. Consistency built into the trust architecture so it holds across surfaces.
Support with. Allē for dual-surface consistency (consumer and provider redesigned from a single strategic premise). Thermo Fisher for multi-partner system coherence.
Cut or hold. Agentic Labs solo builds. In this room, solo work triggers the team-leadership concern before you've established credibility. If AI comes up, pivot to Alibaba's agentic sourcing vision (the case closes with it) or the Trust essay's framework.
IC+manager frame. Tilt toward leadership and organizational influence. Show the mandate-building at Alibaba: research case → executive buy-in → cross-functional execution. Craft depth surfaces through the design decisions within each sprint.
Landmine. "Your recent work is solo AI projects and a product role. Are you still a design leader?" Lead with Alibaba's team-led outcomes. Then: "The product role sharpened how I think about commercial outcomes. Deliberate detour. The AI work proves I stepped deeper into the material that's reshaping design." Pivot to the Trust essay.
H2 headcount unlocks mid-July. Inboxes will be light through the long weekend. Run the fear check before you write the message Monday morning. The proof is the same every time. The sequence is what changes.
- Suno's Head of Product Design: Posted April 21 on Ashby, this $5.4B-valuation creative-AI role spans Product Design, Design Engineering, and UX Research, making it a clean Play Card 1 target worth prioritizing this cycle.
- Salesforce's design-systems-as-AI-infrastructure move: Their Director of Product Management, Design Systems posting reframes Lightning from component library to semantic ontology with guardrails and trust criteria for AI-driven UI composition, which signals where enterprise-scale consistency is heading across the sector.
- Adobe GenStudio's dual-fear profile: The Director of Product Design posting explicitly asks for "intelligence-layer design" and agent-handoff prototyping alongside enterprise governance, so check whether the interviewer's title signals Play Card 1 or Play Card 5 before choosing your lead.
- NIST's expanding AI trust framework: The AI Risk Management Framework now includes a Generative AI Profile and a 2026 concept note for critical infrastructure, giving you shared regulatory vocabulary for any Play Card 3 room where compliance stakeholders are present.