The Red Cross teaser, pulled live this week at standard desktop resolution, has gotten materially better. The project brief now surfaces the GM title, national deployment scope, six-to-one system consolidation, and outcome metrics — all visible without scrolling. Structure is clean. Brief is scannable.
But the headline fires before any of it registers.
The largest text on the page is "Designing for the worst day of someone's life." That's the first thing a hiring manager's eye hits. By the time it tracks right to row 3 of the project brief where "Product Design Director / GM" lives, the category is already written: empathetic nonprofit work. The brief adds detail to a frame that's already set. It can't overwrite it.
Red Cross sits in your 0→1 high-stakes build pillar. The buyer for this case is evaluating whether you can own a complex, multi-system, regulated build end-to-end. The teaser gets roughly four seconds to make that claim. Here's how the five lenses read against what's live.
1. GM-level mandate scope
Doesn't lead. The title appears in row 3 of a six-row brief in the right column. The headline in the left column carries the humanitarian frame. Eye hits the headline first. The mandate is present but subordinated; it reads as context for the emotional story rather than the reverse. A CPO evaluating for a complex regulated build reads this as mission-driven design. Someone who cares about users in crisis, not necessarily someone who consolidated six systems under a federal deadline. Those are different hires, and the headline is sorting you into the wrong one.
Move the mandate into the headline.
Replace the current headline with:
"6 systems → 1 national platform. 6 months. Product Design Director / GM."
The humanitarian context doesn't disappear. It moves into the supporting paragraph, where it becomes texture beneath a mandate claim, evidence of the conditions the system had to perform under.
2. Humanitarian framing vs. mandate framing
They compete. Three elements carry the humanitarian frame before the mandate registers: the headline ("worst day of someone's life"), the opening of the supporting paragraph (displaced families, shelters), and the red italic styling that visually emphasizes the emotional phrase. Together they establish the emotional register as the primary signal.
The supporting paragraph should open with the mandate.
Replace the current opening with:
"Six disconnected tools, a federal compliance deadline, and three field roles that had never shared a system. I led design and delivery of a single shared record that deployed nationally within one disaster-response cycle — usable by an untrained volunteer on day one, under surge conditions and federal oversight."
The displaced-families context follows. It matters; it's proof of the difficulty the system had to absorb. But it belongs underneath the mandate, not above it.
3. Outcome metrics in the first scan
Technically visible. Row 6 of the project brief shows $847K disbursed, 6→1 system switches, and 1,689 cases. But row 6 is the last row, at the bottom of the right column. In a top-to-bottom scan, it's the final thing the eye reaches.
Pull one metric into the headline. $847K disbursed in the first deployment cycle stops a scanning eye in a way that row 6 cannot. Pair it with the system consolidation and the headline carries mandate and proof before the brief even registers.
If you adopt the headline fix from lens 1, the remaining metrics can stay in the brief; the mandate headline gives them a frame to land in. If you keep the current headline, at minimum swap rows 3 and 6 so the outcome leads. The headline is the stronger move. It changes what the metrics are evidence for.
4. Craft signal in the first screen
There isn't one. The hero contains text, a faint grid, and a large translucent "03." No product screenshot, interface state, system diagram, or design artifact appears above the fold. The first product surfaces (intake form, case dashboard, recovery-needs interface) require scrolling past a representational illustration.
This is the gap I flagged in issue #6: a design manager evaluating the teaser can assess the mandate and the outcomes but can't see evidence that you designed anything. In a hiring-committee debrief, that gap becomes a specific question: did she direct this or design it? Without a visible artifact, the panel defaults to "directed." And "directed a nonprofit project" is not what they're trying to hire for a 0→1 build.
Add one product surface to the hero. The intake form reduction, eight fields to three, is the strongest candidate. It's a visible design decision with clear before-and-after logic. It doesn't need to be large. A partial screen at the right edge of the hero, or replacing the decorative grid, shifts the signal from "directed" to "directed and designed."
5. CTA specificity
"Request the full case study" is generic. It tells the buyer more exists but gives no reason to believe the full case will answer the question the teaser raised.
Replace with:
"See how the intake form went from 8 fields to 3 — and how one record served three field roles that had never shared a system, under federal oversight."
That previews a specific design decision and names the multi-role scope the system served. A hiring manager who reads that knows the full case will show them something concrete.
Sequence
Change the headline first. Everything downstream depends on what category the headline establishes in the first scan. Add one interface artifact to the hero. Rewrite the CTA. Those three moves put the mandate where the eye lands first.
- Figma's recruiter on scanning: Reviewers may spend only minutes or seconds on an initial portfolio, which is why the Figma recruiter guidance recommends showing only the evidence most pertinent to the role rather than the candidate's full history.
- Gusto's trust vocabulary overlap: The Head of Design, Unified Service Platform posting uses "graceful failure," "human handoffs," and send/review/never-send language that maps directly to the Red Cross case's untrained-volunteer-under-surge-conditions evidence.
- OpenAI Engineering Acceleration loop: The Engineering Acceleration role names instrument, observe, evaluate, decide, iterate, and rollback as the product loop — the same causal chain the Red Cross case could demonstrate if the full case shows how the intake reduction was validated under deployment pressure.
- The assurance-case structure: NIST defines an assurance case as claims supported by systematic argumentation, evidence, and stated assumptions — a structure worth applying to each teaser, since the NIST definition clarifies why a mandate claim without visible craft evidence leaves the argument incomplete.