Juno, the CS-02 teaser has improved substantially since the last cycle. The browser tab title does real positioning work. The six-row project brief is scannable and outcome-loaded. Below the fold, the live embedded dashboard and full design-system grid are the strongest published proof you have for regulated 0→1 builds.
The first screen still isn't earning the click from the people you need to reach.
Outcome-Forward Title
Current headline: "The supply chain six partners never had."
A hiring manager at Gusto scanning open tabs at 10 PM reads this and files you under "supply chain project." She knows you built a platform. She doesn't know what you decided. The headline forms the first theory about what this case proves, and right now that theory is too generic to do any work for you. The outcomes are on the page — $20M+ margin opportunity, 100% partner adoption, 83% IRR — but they sit in the right-column brief where they read as metadata, not as the lead claim.
What matters at Director+ level is the design decision that produced those numbers. The $20M and 83% IRR support the claim. They shouldn't lead it.
Rewrite the headline to name the design mechanism.
Two options, in order of strength:
- "Review logic was in people's heads. We put it in the system." Names the design decision. Translates directly to what Gusto is hiring for: when does a human intervene, when should an output not ship.
- "Exceptions cost 3–5× more at the gate. We moved detection upstream." Leads with the structural insight that drove every design decision in the project.
Both give the hiring manager a design-thinking signal before she reads a single line of body text. The current headline gives her a deliverable.
Structural Insight on First Screen
Current summary paragraph: Establishes that no shared platform existed, six partners used spreadsheets and phone calls, exceptions surfaced at the delivery gate when remediation cost 3–5× more, one platform was adopted by all six, production in twelve months.
Solid project summary. But it doesn't name the design insight — the specific structural decision that separates this from what Accenture or an enterprise vendor would have built.
That insight exists on the page. It lives below the fold, in the closing argument and the design-system grid: exception-first architecture, approve/defer/reject controls that surface constraints before they become violations, a system that supplies evidence for review without replacing the qualified person who signs off. That's what a Director+ panel evaluates. It needs to appear before any scroll.
Add one sentence to the summary paragraph, after the remediation-cost line:
"We designed an exception-first platform: every view surfaces what's at risk before it reaches the gate, and the system supplies evidence for review without replacing the qualified person who signs off."
This names the design architecture (exception-first), signals the regulated-domain constraint (qualified-person release), and uses vocabulary — risk visibility, evidence supply, human review authority — that a hiring manager writing postings about trust, uncertainty, and human handoffs will recognize as her problem described in different domain language.
Craft Signal
The first screen has no visible design artifact. The project brief, the headline, and the summary are all text. The lower edge of the viewport hints at the dashboard section — browser chrome, the top of a phone mockup — but nothing a hiring manager would register as craft evidence before deciding whether to scroll.
Below the fold, the evidence is strong. The live embedded dashboard and full component grid covering color tokens through batch timelines, workflow states, and KPI cards would satisfy a making-depth question in most regulated-domain panels. But a hiring manager who doesn't scroll past the hero never sees any of it.
Pull one artifact above the fold. The approve/defer/reject control set from the design-system grid is the strongest candidate. Small footprint. Visually distinct from the surrounding text. Directly illustrates the structural insight you're claiming — the system makes the decision moment visible. Place it in the right column below the project brief, or as a narrow strip between the hero and the dashboard section.
If layout constraints prevent adding a component: replace the decorative "02" behind the hero content with a three-state flow — at-risk, under review, released — pulled from the batch timeline below. Same space, but now it carries a craft signal instead of serving as a visual treatment.
CTA
Current CTA: "Request the full case study" — red button, visible on first screen, links to the full CS-02 page.
Present and visible above the fold — mechanically fine. But does the first screen create an unanswered question that makes clicking feel necessary?
Right now, it doesn't. The summary paragraph delivers a complete arc — problem, cost, result. The project brief adds the margin figure and IRR. A hiring manager who reads both has a satisfying business story with no loose thread. She can close the tab.
The headline and summary changes above partially fix this. Once the first screen names exception-first architecture and approve/defer/reject logic, it creates a claim the reader hasn't yet seen executed. The mechanism is named but not shown. That gap is what the CTA should harvest.
Rewrite the CTA to name what's on the other side. "Request the full case study" is generic. Try: "See the exception-detection workflow and full design system →" This gives the reader a specific reason to click and connects the promise to the structural claim she just read.
Priority-Ordered Action List
-
Rewrite the headline. Lead with the design mechanism. Strongest option: "Review logic was in people's heads. We put it in the system."
-
Add one sentence to the summary paragraph naming exception-first architecture, risk visibility before the gate, and human review authority preserved. This is the vocabulary bridge to your target postings at Gusto, Ramp, and regulated-domain companies.
-
Pull one design artifact above the fold. The approve/defer/reject control set — small footprint, high signal density, directly illustrates the structural claim.
-
Rewrite the CTA to name what the full case contains. "See the exception-detection workflow and full design system →"
- Gusto's trust vocabulary match: Their Head of Design posting asks the designer to determine when outputs are ready to ship, need human review, or should not be sent — language that maps almost directly to the approve/defer/reject controls sitting below CS-02's fold.
- OpenAI's instrument-to-rollback loop: The Engineering Acceleration posting names instrument, observe, evaluate, decide, iterate, and rollback as the product loop, which means CS-02's weekly partner benchmarking and forecast-versus-actual comparisons could be reframed as operational evaluation evidence if the teaser surfaced them.
- Progressive transparency tension: A CHI 2026 study found that eight of twelve participants preferred on-demand or progressive disclosure over maximal process visibility, which complicates how much of the exception-first architecture to show on the teaser's first screen versus reserving it for the full case.
- Assurance case as portfolio structure: NASA describes an assurance case as a reasoned artifact connecting explicit claims, systematic argumentation, and heterogeneous evidence, a model that could organize CS-02's full case around the claim that review logic belongs in the system rather than in a person's head.