The shared condition
Every buyer on your list arrived at their company with a thesis about what design should do. They are hiring a leader to execute it. Daniela Jorge at Capital One. Gab Menachem at Amplitude. Jeremy Epling at Vanta. Jason Lee at Maven. Different titles, different domains, same buyer type.
They own design outcomes but did not build the current design org. They inherited it, or they inherited its absence. The posting they wrote reflects a thesis that formed during the gap between their arrival and the req going live. (I covered the tenure-to-posting gap as mandate signal in the Capital One battlecard. Longer gap, more formed thesis.)
What they say they evaluate is nearly identical across all four postings: AI experience, team leadership, craft depth, shipped products. Every serious candidate at this level has all four. The stated criteria do not differentiate. Price of entry, not decision surface.
What they actually evaluate diverges by the specific fear driving the hire. That fear is readable before you speak to them.
The fears
"Will this person change how we work, or just do good work?"
Enterprise buyer. Jorge came to Capital One from PayPal's CDO seat, where she ran design across 425 million customers and 34 million merchants. Before that: AT&T, eBay, Intuit, Yahoo!, Kodak, Kaiser. Two decades of operating design inside large, complex organizations that are not design-native. That trajectory predicts her evaluation lens with high precision: she screens for business-language fluency, not craft authority. Every org she has led required it.
Her public frame confirms it:
"Leader first, designer second."
She evaluates whether you understand revenue, margin, risk, legal, go-to-market. She does not need you to teach her what good design looks like.
Her posting is for a Director of AI in Experience Design. The person should be "actively rewriting" what the design discipline looks like. Look at the verb cluster: build, own, define, implement, establish, architect, scale, integrate. Infrastructure verbs. Not craft verbs.
The posting asks for AI experience. What she actually evaluates is whether you understand AI as an operating-model change that transforms how several hundred designers work. Not a product feature you shipped. Confidence: moderate-high. Inferred from posting verb cluster, career trajectory, and her published frame. No direct hiring-evaluation commentary from Jorge surfaced in public sources.
"Will this person slow down the loop?"
Growth-stage buyer. Menachem founded Loom Systems, an AIOps company that compressed the distance between "something happened in the system" and "a human does the right thing about it." ServiceNow acquired it. He then scaled ServiceNow's ITOM line to $1B+ ARR. Founder brain, not career-product-executive brain.
His Amplitude posting is the only one with an explicit "What this is not" section. He negated four candidate shapes: the executive who manages up and delegates down, the pure IC, design ops, the approval committee. Those negations are the real brief. They tell you exactly what went wrong before or what he refuses to repeat.
Posting title says Head of Product Design. What he actually evaluates is whether you can operate inside a product's signal architecture, stay in the work, and make data actionable at the point where decisions get made. No public design-evaluation commentary exists from Menachem. His public material is entirely about complex systems and real-time actionability. He does not think about design as a discipline with its own evaluation criteria. Confidence: moderate. Inferred from career pattern and posting language, not from any stated design-hiring framework.
"Does this person understand what's actually at stake?"
AI-native vertical buyer. Two sub-variants, both rooted in domain consequence.
Epling came from GitHub, where he scaled Actions from zero to millions of developers. Before that, 16 years at Microsoft on Azure. He says he "feeds off details" and wants math, priors, specific reasoning. He explicitly values craft, beauty, usability, and says product/design decisions carry a larger blast radius than engineering decisions because a wrong call "can confuse many engineers." He will test specificity. Principles-altitude answers will lose him. Confidence: high. Sourced from direct public statements in the First Round Review interview.
Lee came from Nurx, where he led product, design, and program management for patients and providers. His Maven posting covers product design, brand, research, clinical outcomes, AI-native workflows, and human judgment boundaries. He has said publicly that healthcare teams should be "maniacal" about experience quality, including trivial bugs, because patients seek care when vulnerable. His posting uses "not just aesthetics" and frames user research as "not a separate track." Both negation tells. He has seen design leaders who kept craft and research at arm's length from clinical outcomes. Confidence: moderate. Primarily inferred from posting negation language and a single public interview. Less direct evaluation commentary available than for Epling.
Across both: the postings ask for AI-native design experience. The actual screen is whether you grasp that a design error in their domain carries compliance, clinical, or security consequence that a design error at a consumer app does not. AI fluency without domain gravity reads as lightweight.
Reading the fear in real time
You will not always know which fear is dominant before the first conversation. Four tells to read it live.
They ask "what would you have done differently?" then probe the judgment call, not the process. The question itself is common. What matters is what they do with your answer. If they push on the tradeoff, the thing you sacrificed, the bet you made with incomplete information: they are evaluating you as a product peer. Menachem and Epling pattern. Lead with judgment calls, tradeoffs, things you got wrong.
They ask "how would you have managed the team through it?" then probe the org decision, not the outcome. Common question, different follow-up. They want the structure you chose, the hiring sequence, how you reorganized when the mandate shifted. They are evaluating you as an org leader. Jorge pattern. Lead with operating-model decisions.
They raise a domain constraint unprompted in the first ten minutes. They are screening for domain seriousness, not domain expertise. Lee asks about clinical validation. Epling asks about compliance workflows. They want to know you understand why this domain is harder. You do not need to be the expert. You need to demonstrate that the stakes register differently for you.
They spend the first five minutes on your current role, not your portfolio. They are evaluating currency. This is where TinyFish does its work: 3 products shipped in 3 months, enterprise AI agents in production, daily work with agent traces, auditability, governance. Do not cite TinyFish as design portfolio proof. Cite it as evidence you are inside the problem they are hiring for right now.
What wins and what loses
Enterprise buyer (Jorge archetype). Lead with the Trust essay and Agentic Labs. You have built the agentic workflows she is trying to install across a several-hundred-person org. Support with Alibaba for enterprise credibility at scale. Do not lead with product design cases. She is hiring someone to rewrite how designers work, not to do design work. I covered this positioning in the Capital One battlecard.
Vulnerability: The enterprise design director archetype (Salesforce, Adobe pedigree) outperforms you on org size. Counter: They have managed designers. You have built the thing she needs her designers to adopt.
Growth-stage buyer (Menachem archetype). Lead with TinyFish velocity as speed proof, then map the Trust essay's commitment-point framework directly to Amplitude's product. Amplitude's Wave Opportunities feature is the moment where a team decides whether to act on an agent-surfaced signal. That is your commitment point, named in your vocabulary before Amplitude named it in theirs. I covered this vocabulary convergence as the strongest possible outreach hook. Do not lead with team-building. He explicitly negated it.
Vulnerability: The AI-native design leader archetype (Anthropic, OpenAI pedigree) outperforms you on AI brand recognition. Counter: They have designed AI interfaces. You have built and shipped enterprise AI products from the product side. You understand the signal loop, not just the surface it renders to.
AI-native vertical buyer (Epling/Lee archetype). Match the domain constraint. For Vanta: lead with Thermo Fisher. Regulated, compliance-adjacent, exception-first design, 0→1 function build. For Maven: lead with Thermo Fisher and Red Cross. High-stakes, vulnerable users, multi-stakeholder platforms where a design error has human consequences. Support with the Trust essay's Decision Gate handoff in both cases. Epling will ask for the math. Have it. Lee will ask about the patient. Know why that question is different from asking about the user.
Vulnerability: The IC-track senior candidate (Staff+ from FAANG) outperforms you on pure craft depth. Counter: They have not built functions, operated in regulated contexts, or shipped products where the design surface carries clinical or compliance consequence.
The default before you know anything
When you encounter a new-mandate buyer and have not yet identified which fear is dominant:
- Open with TinyFish as currency proof. One sentence. You are building enterprise AI agents in production right now.
- Name their company's commitment point. The specific moment where a human decides whether to trust what the system produced. Every company on your list has one. Find it before you write the outreach.
- Hold your portfolio in reserve until you read the tell. The first question they ask reveals which case to lead with. Do not pre-commit.
The new-mandate buyer has already decided what they need. Your job is to figure out what that is before they tell you, because by the time they tell you, you are already being compared to someone who figured it out faster.
- Gusto just posted fresh: A Head of Design, Unified Service Platform role appeared on Greenhouse July 22 with explicit language about trust, graceful failure, AI-human handoffs, and "outputs that should not be sent" — the clearest commitment-point posting language on the board right now.
- Epling on design's blast radius: His First Round Review interview is the most detailed public window into how a CPO-as-design-buyer thinks about product/design decision weight, quality bars, and why he demands specificity over principles.
- Vanta's Agent for Risk launch: Vanta's June 2026 announcement introduced the Trust Graph spanning 400+ integrations and 1,400+ continuous tests — the product surface the Head of Design will own, and the strongest domain-specific commitment point to name in outreach.
- Maven's title tension matters: Maven's leadership page now lists Jason Lee as Chief Product & Operating Officer, while the VP of Design posting still reports to "Chief Product Officer" — a scope expansion that may change the buyer's available attention and the design leader's operating latitude.