The Evaluation Gap
Every Director and Head of Design posting you'll read this month lists the same competencies: strategic thinking, cross-functional collaboration, team leadership, visual consistency. Functionally identical across companies that have nothing else in common. These criteria tell you what HR approved. The interview panel is filtering for something else entirely: judgment fit to a specific company moment. Whether you can name the design problem before the org has cleanly named it, build a mandate rather than execute a brief, move product and engineering and legal without owning any of them.
Two forces make this gap wider right now than it's been.
AI-native expectations have leaked across every tier. Ramp's posting says "building tools and agents." Gusto calls itself "an AI-native design organization." Maven's Head of Design posting uses the phrase "AI-native" to describe a maternal healthcare role. A payroll company and a therapy-access platform now test for the same AI-mediated design judgment as OpenAI. Prepare using each tier's old playbook and you will be surprised in the room.
Player-coach altitude is the norm at Director and Head level everywhere, not just growth-stage. Headway says it explicitly. Kikoff says "this is not a hands-off role." OpenAI asks for comfort "from design details to system architecture." The altitude expectation is universal. The vocabulary each tier uses to describe it is not.
Your Alibaba case registers as enterprise-scale trust architecture in one room and big-company baggage in another. Same background. Different evaluation frame. The playbooks teach you to read which room you're walking into.
How Frontier AI Companies Actually Screen Design Leaders
Frontier AI companies screen design leaders differently from enterprise or growth-stage companies. The first conversation is the gate, and the evaluator is listening for whether you treat model behavior as a design surface with its own properties you've formed opinions about. Your essay already names the problem they're hiring someone to solve. This playbook maps the three unstated filters operating across OpenAI, Anthropic, DeepMind, and the rest of the tier, the org-chart reality behind the postings, and the specific framings that will get you cut before you ever open your portfolio.
Builder Instinct Is the Screen
Several of these searches enter final-round scheduling the week after the Fourth. You have four working days to position outreach. The single screen across this tier — Ramp, Gusto, Headway, Kikoff, Hiive — is builder instinct versus process dependence: can you ship quality at their speed without the infrastructure you're used to? They share that filter but differ in workflow shape, which changes which of your three portfolio pillars leads. This is your signal map for the cluster, organized by what each company is actually evaluating and what version of your story to tell.
Enterprise Platform Tier Playbook — How Salesforce, Atlassian, Adobe, DocuSign, Intuit, and Microsoft AI Actually Screen Design Leaders
Salesforce, Atlassian, Adobe, DocuSign, Intuit, Microsoft AI. Six companies, six different interview loops, one decoded evaluation criterion none of them will name: can you create coherence across organizational complexity without the authority to mandate it? Every portfolio review, behavioral round, and systems exercise is a different angle on that question. This playbook maps the frame to your background, prioritizes which companies deserve your outreach before the holiday weekend, and delivers an honest structural read most candidates never get: where design actually sits at the executive table, and what that means for what you're walking into.
Tier Playbook — Healthcare and Vertical SaaS
Maven Clinic, Amae Health, Ontra, Aurora Solar, Front. Five companies running the same unstated screen on every design leadership candidate: does this person treat domain constraints — clinical protocols, eligibility logic, contract hierarchies, provider burden — as the design material, or as friction between them and the work they actually want to do? Most candidates from consumer or platform backgrounds fail it before the portfolio deck loads. The playbook maps your published cases to each company's operational vocabulary, ranks all five by positioning confidence and timing, and flags the mandate risks worth verifying before you commit.