// Code-generation skill for applying behavioral science to Career Hub page implementation. Use only when changing CTA copy, page structure, tool interactions, or other code-backed conversion surfaces in the repo.
[HINT] Download the complete skill directory including SKILL.md and all related files
| name | marketing-psychology |
| description | Code-generation skill for applying behavioral science to Career Hub page implementation. Use only when changing CTA copy, page structure, tool interactions, or other code-backed conversion surfaces in the repo. |
| metadata | {"family":"marketing","owner":"marketing","last_reviewed":"2026-04-30T00:00:00.000Z","version":"1.0.0"} |
This is a code-generation skill. Use it only when the outcome is a shipped page or interaction change.
Apply behavioral science and mental models to Career Hub pages. This skill is the "why" behind CRO decisions — it explains the psychology that makes page-cro recommendations work.
Read nextjs-app/docs/BRAND.md for voice and framing rules. The Career Hub uses empowerment framing, not fear — psychology is applied ethically to help workers make better decisions, not to manipulate.
page-cro when auditing or optimizing pages/career-hub/roles/[slug])| Principle | Application | Implementation |
|---|---|---|
| Anchoring | Show high end of pay range first in hero | "$22/hr (up to $28 with certifications)" anchors expectations high |
| Social proof | FlexerTestimonials from that role | "Maria earned $19.50/hr on her first warehouse shift" |
| Goal-gradient | Career path visualization | Show progression: Picker → Lead → Supervisor with pay at each step |
| Loss aversion | Frame as missing opportunity | "Bartenders in Dallas are earning $18-25/hr this weekend" (not "you're losing money") |
| Specificity effect | Exact pay figures, not ranges | "$19.50/hr" is more believable than "competitive pay" |
/career-hub/tools/[slug])| Principle | Application | Implementation |
|---|---|---|
| IKEA effect | User builds something (calculates, plans, analyzes) | Calculator interaction creates ownership of the result |
| Goal-gradient | Progress indicators in multi-step tools | "Step 2 of 3" progress bar drives completion |
| Endowment effect | Personalized results feel owned | "YOUR take-home pay is $1,240/week" — the word "your" triggers ownership |
| Commitment & consistency | Small action leads to bigger one | Use calculator → see results → "Find shifts paying this rate on Indeed Flex" |
| Zero-price effect | Tools are free | "Free" triggers irrational preference — emphasize "Free calculator, no signup required" |
| Zeigarnik effect | Incomplete results create pull | "You've calculated your pay — now see how it compares to other cities" |
/career-hub/guides/[slug])| Principle | Application | Implementation |
|---|---|---|
| Reciprocity | Give genuine value before asking | 1,500 words of useful content before CTA |
| Authority bias | Expert attribution and data sources | AuthorByline + DataSourceCitation components |
| Mere exposure | Consistent Indeed Flex brand presence | Subtle Indeed Flex mentions throughout, not just in CTA |
| Confirmation bias | Align with what workers already believe | "You deserve to know your rights" (validates their worldview) |
| Pratfall effect | Honest caveats increase trust | "Rates vary by state — here's how to check yours" (admitting limitation) |
/career-hub/cities/[slug])| Principle | Application | Implementation |
|---|---|---|
| Contrast effect | Compare to nearby/similar cities | "Philadelphia pays $2/hr more than the national average for warehouse work" |
| Availability heuristic | Name specific local employers | "Amazon, Target, and Walmart all hire through Indeed Flex in Dallas" |
| Mental accounting | Frame earnings in monthly/yearly terms | "$18/hr = $2,880/month = $34,560/year" — bigger numbers feel more significant |
| Proximity bias | Emphasize local relevance | "In YOUR city" language, local cost of living, neighborhood guide |
| Principle | Application | Example |
|---|---|---|
| Activation energy | Make first step trivially easy | "Download the free app" (not "Create an account and complete your profile") |
| Loss aversion | Frame what they miss by not acting | "Workers in your area are picking up shifts this weekend" |
| Zero-price effect | Emphasize free | "Free app. No fees. You choose your shifts." |
| Default effect | Pre-fill context from page | After pay calculator: "Find shifts paying $18+/hr" (pre-filled from their calculation) |
| Hick's Law | One clear CTA per context | Don't offer "Download App" AND "Sign Up" AND "Learn More" — pick one |
| BJ Fogg model | Motivation x Ability x Prompt | High motivation (they want shifts) + High ability (free download) + Clear prompt (CTA button) |
| Model | Career Hub Application |
|---|---|
| Pareto (80/20) | 20% of pages drive 80% of traffic. Optimize those first. Which 5 tools get 80% of usage? |
| Theory of Constraints | If traffic is high but conversions are low, the bottleneck is the page, not the content pipeline |
| Jobs to Be Done | Workers don't want "a flexible staffing platform" — they want "to earn $500 this week around my class schedule" |
| First Principles | Don't copy competitor content hubs. Ask: what do OUR workers uniquely need that they can't get elsewhere? |
| Inversion | What would guarantee a worker leaves our page? Unsourced data, corporate language, buried CTA, wall of text. Prevent those. |
| Local vs Global Optima | Optimizing CTA button color (local) won't help if the value proposition is unclear (global). Fix the big things first. |
| Model | Career Hub Application |
|---|---|
| Hyperbolic discounting | Workers prefer immediate benefits. "Get paid this week" beats "Build your career over 6 months." Lead with immediate, follow with long-term. |
| Status-quo bias | Switching to a new app feels risky. Reduce friction: "Import your resume in one tap" / "Book your first shift in under 5 minutes" |
| Fundamental attribution | When a worker doesn't convert, examine the page experience (situational) before assuming they weren't interested (personal) |
| Peak-end rule | Tool completion = peak moment. Make results satisfying and immediately useful. Thank-you pages should add value, not just confirm. |
| Curse of knowledge | "W-2 employment with Indeed Flex" is obvious to us but confusing to workers. Explain: "You're employed by Indeed Flex — taxes are handled, you get benefits access." |
Particularly relevant for role pages, salary comparison tools, and financial content.
| Principle | Application |
|---|---|
| Anchoring | Show the high end of the range first: "$25/hr for experienced workers, starting at $16/hr" |
| Framing | "$18/hr" vs "$720/week" vs "$37,440/year" — use the frame that feels most impactful for context |
| Rule of 100 | For hourly rates under $100 (always true for Career Hub), percentage increases feel larger: "earn 15% more with a certification" |
| Mental accounting | "$18/hr" → "That's $144/day" → "That covers your weekly groceries in one shift." Connect earnings to tangible outcomes. |
| Contrast effect | Show pay progression: "$16/hr entry → $20/hr with 1 year → $25/hr as lead" — each step feels achievable relative to the last |
| Specificity effect | "$18.50/hr" is more credible than "$18-19/hr" which is more credible than "competitive pay" |
| Challenge | Models to Apply | Career Hub Fix |
|---|---|---|
| Low conversions on tool pages | IKEA effect, Goal-gradient, Commitment & consistency | Add progress bars, personalize results, CTA uses result context |
| Workers don't trust our data | Authority bias, Pratfall effect, Social proof | Add BLS citations, acknowledge limitations, add worker testimonials |
| High bounce on role pages | Anchoring, Activation energy, Hick's Law | Lead with high pay range, simplify CTA, one action per section |
| Workers don't finish calculators | Zeigarnik effect, Goal-gradient, Zero-price effect | Show "Step 2 of 3", "Almost there!", emphasize "free" |
| CTA feels pushy | Reciprocity, Commitment & consistency, Zero-price effect | Give value first, frame as free and no-commitment, use value-first copy |
| Pay data feels unbelievable | Specificity effect, Authority bias, Contrast effect | Use exact BLS figures, cite source inline, compare to regional/national averages |
| Workers stuck in comparison mode | Paradox of choice, Default effect, Decoy effect | Highlight "most popular" role/city, limit comparison tables to 3-5 options |
| Content feels generic | Curse of knowledge, Confirmation bias, Jobs to Be Done | Use worker language (from worker-research), validate their worldview, frame around their job-to-be-done |
The Career Hub uses psychology to help workers make better decisions, not to manipulate.
When applying psychology to a page audit:
## Psychology Audit: [Page URL]
### Current Psychological Profile
- Primary motivation lever: [what's currently used]
- Missing principles: [what's absent that should be present]
### Recommendations
#### 1. [Principle Name]
Location: [where on page]
Current: "[current copy or design]"
Recommended: "[improved version]"
Expected impact: [why this works psychologically]
#### 2. ...
### CTA Psychology Check
- Activation energy: High / Medium / Low
- Value framing: Present / Missing
- Social proof proximity: Near CTA / Far from CTA / Missing
- Risk reducers: Present / Missing
| Resource | Path |
|---|---|
| Brand voice (empowerment framing rules) | nextjs-app/docs/BRAND.md |
| Conversion goals and CTA hierarchy | nextjs-app/docs/MARKETING_CONTEXT.md |
| Page-type CRO checks | .agents/skills/marketing/page-cro/SKILL.md |
| Persona definitions (motivation triggers) | nextjs-app/src/features/career-hub/personas/data/persona-hubs/* |
| Skill | Relationship |
|---|---|
page-cro | This skill provides the "why" behind page-cro's "what" — use together |
copy-editor | Sweep 6 (Emotion) and Sweep 7 (Zero Risk) draw from these principles |
content-writer | CTA guidelines and benefit framing informed by these models |
frontend-design | Visual hierarchy and interaction design should reinforce psychological principles |
worker-research | Research validates which psychological triggers resonate with actual workers |