skills / marketing

marketing-psychology

what it does

When the user wants to apply psychological principles, mental models, or behavioral science to marketing. Also use when the user mentions 'psychology,' 'mental models,' 'cognitive bias,' 'persuasion,' 'behavioral science,' 'why people buy,' 'decision-making,' 'consumer behavior,' 'anchoring,' 'social proof,' 'scarcity,' 'loss aversion,' 'framing,' or 'nudge.' Use this whenever someone wants to understand or leverage how people think and make decisions in a marketing context. For applying psychology to specific pages, see cro; for pricing tactics, see pricing; for copy framing, see copywriting.

Verified Before and after: 2 models 1 try per case View source Download .zip

Compare the answers

Answers from Claude Sonnet 5.5 and GPT-6.1 Sol · try 1 of 1

Same request, same model, once with the skill and once without. Read both and judge for yourself.

“We want to pay $50 for every person a user refers who signs up, to grow fast. We have a free plan. How would you set this up, and what could go wrong?”

with the skill
# Short version Don't pay $50 for a free signup. It's the textbook **Cobra Effect**: you reward the easiest-to-fake action, so you'll get fake or worthless signups. Pay for a **qualifying action** further down the funnel instead. ## 1. Check the unit economics first Free signups aren't worth $50, paying customers might be. - Effective cost per paying customer = reward ÷ free-to-paid conversion rate - At 5% conversion, $50 per signup is **$1,000 per paying customer** - At 20% conversion, it's $250 Compare that to your first-year gross margin per customer (or LTV). A rough rule is to keep the total referral cost under ~30-50% of first-year gross profit, or to fit within your current CAC payback window. If your plan is $15/month, $50 is a very different bet than if it's $300/month. ## 2. Recommended structure **Pay on conversion, not signup.** - Trigger: the referred user becomes a paying customer, or hits a meaningful activation milestone (e.g., invited teammates, completed X) - Hold the payout for 30-60 days (past your refund window) so you don't pay for churned or refunded accounts
Show the rest (64 more lines)
**Make it two-sided.** - Friend gets something (an extended trial, a discount on the first paid month) and referrer gets cash or credit - A two-sided offer feels fairer (reciprocity), and it makes the referral an easier thing to share **Consider credit rather than cash.** - Account credit costs you less than cash (lower effective cost, since it's your margin, not your dollars), raises switching costs, and attracts fewer fraudsters - Cash appeals to non-users and affiliates. Decide whether you want advocates or marketers. **Add a smaller signup-level reward if you want momentum.** - E.g., $5 credit at signup for the friend plus the $50 for the referrer upon conversion. Small immediate rewards help with present bias, and the big one is the real prize. **Tier it.** - Escalating rewards (e.g., bonus at 5 paid referrals) leverage goal-gradient and keep your best advocates engaged. **Make it easy.** - Personal link, one-click share, pre-written message, dashboard showing pending/earned rewards (progress visibility drives repeat sharing) ## 3. What could go wrong **Fraud and gaming (the biggest risk)** - Self-referral with throwaway emails, bots, signup farms, or people posting links on coupon and "get paid" sites - Mitigations: pay only on paid conversion or verified activation, delay payouts, cap rewards per user per month, flag shared IPs/devices/payment methods/email patterns, block disposable email domains, require email verification, and keep the right to void rewards in the terms **Low-quality users** - Even real people referred purely for the bounty churn fast or never activate. Tie rewards to retention (the hold period) and track referred-cohort retention separately from other channels. **Economics breaking at scale** - Referral volume that "works" can still be unprofitable. Set a monthly budget cap and a kill-switch, and track CAC, conversion, and retention for referred vs. organic users. **Cannibalization** - You may pay $50 for people who would have signed up anyway (e.g., via a coupon site or a public link). Attribute carefully and exclude existing users and prior signups. **Brand and trust damage** - Spammy outreach by referrers hurts your reputation and can violate laws (CAN-SPAM, GDPR/PECR, TCPA for texts). Provide pre-written copy that follows the rules, ban unsolicited bulk messaging, and don't let referrers send emails from your system without consent. **Legal and compliance** - Referrers may need to disclose the incentive (FTC endorsement rules in the US) - Paying out cash triggers tax reporting (e.g., 1099s in the US above thresholds), international payout and sanctions issues, and possibly money-transmission questions if you handle it yourself - Check the terms of any platform you promote on (app stores, social networks restrict incentivized referrals) - Have clear program terms covering voiding, caps, and changes **Support and ops burden** - Disputes about "my friend signed up but I wasn't credited." Use a visible referral dashboard and clear attribution rules (e.g., 30-day cookie, last-click vs. first-click). **Training users to expect it** - If you later cut the reward, advocates may feel it's a loss (loss aversion). Frame it from the start as a limited-time or tiered program, and avoid promising permanence. ## 4. How I'd launch it 1. Calculate the max affordable reward from LTV and conversion. 2. Start small, with a cohort of your best existing users (invite-only), not a public launch. 3. Pay on paid conversion with a 30-60 day hold, plus a friend incentive. 4. Set caps and monitoring from day one. 5. Test reward size and type (cash vs. credit, $25 vs. $50) with A/B or cohort comparisons. 6. After 60-90 days, compare referred-user retention and margin to other channels, then scale or adjust. ## Questions that would sharpen this - What's your paid plan price and free-to-paid conversion rate? - Is this B2B or B2C (and is there a team/seat dynamic)? - Do you have a margin or CAC target? - Do you already have organic word of mouth to build on? With those numbers I can suggest a specific reward amount and structure.
without
# Short version Don't pay $50 for a signup when you have a free plan. Pay for a **qualifying action** that signals real value, like a paid conversion. Otherwise you're offering $50 for something anyone can fake for free. ## How I'd set it up **1. Define the payout trigger** Options, from safest to riskiest: - **Referred user becomes a paying customer** and passes the refund window (30-60 days). This is the safest. - **Referred user hits an activation milestone** (e.g., connects an integration, invites teammates, completes X actions). Use this if paid conversion is slow or rare. - **Raw signup.** Avoid it, or pay a tiny amount (a few dollars or credits) if you must. **2. Check the unit economics first** - What's your free→paid conversion rate, and what's the LTV of a paying customer? - If 5% of signups convert and each is worth $300, a signup is worth about $15, so a $50 payout per signup loses money. A $50 payout per *paying customer* is probably fine. - Your payout should be well under LTV (a common target is LTV:CAC of 3:1 or better). **3. Pick the reward structure** - **Cash vs. credit:** Account credit or discounts are cheaper for you (cost is below face value) and much less attractive to fraudsters. Cash attracts more volume and more abuse. - **Two-sided rewards:** Give the new user something too (extended trial, discount). It lifts conversion and feels less spammy. - **Delayed payout:** Hold rewards in "pending" until the qualifying event and a clearing period pass.
Show the rest (37 more lines)
**4. Build the mechanics** - Unique referral links or codes per user - Attribution rules (last-click vs. first-touch, cookie window, what happens if someone signs up without the link) - A dashboard showing pending, approved, and paid referrals - Payout via a service like PayPal, Tremendous, or Stripe Connect, with tax handling (W-9/1099 thresholds in the US) - Written terms covering eligibility, fraud, clawbacks, and your right to change or end the program **5. Add fraud controls from day one** - Email verification, plus blocking disposable email domains - Rate limits and per-user caps on rewards (e.g., max 10 per month, or a cap that triggers manual review) - Device, IP, and payment-method matching to catch self-referrals - Manual review of anomalies, such as many signups from one IP or a sudden spike from one referrer - Clawbacks if a referred customer refunds or charges back ## What could go wrong | Risk | What happens | Mitigation | |---|---|---| | **Fake signups / self-referral** | Users create throwaway accounts to collect $50 each | Pay on paid conversion or activation, verify emails, match devices and IPs | | **Organized fraud rings** | Bots or click farms drain your budget fast | Caps, delayed payouts, anomaly alerts, a budget kill switch | | **Low-quality users** | Referrers spam people who never use the product, so your costs rise and metrics get polluted | Reward only activated or paying users | | **Negative unit economics** | CAC exceeds LTV and growth accelerates your losses | Model it first and set a monthly spend cap | | **Coupon/affiliate sites** | Your link gets posted publicly and attracts bounty hunters | Prohibit it in the terms, monitor for it, restrict who can join | | **Spam and brand damage** | Aggressive referrers spam social media or email, and you may violate CAN-SPAM or similar laws | Provide approved messaging, ban unsolicited bulk outreach, enforce the rules | | **Legal and tax issues** | Payout reporting requirements, FTC disclosure rules for influencers, regional restrictions | Require disclosure of the relationship, collect tax info, get legal review | | **Refunds and chargebacks** | You pay a reward and then the customer reverses | Clearing period and clawback terms | | **Cannibalization** | You pay for customers who would have signed up anyway | Track the baseline, and exclude existing leads or users | | **Support and ops burden** | Disputes over attribution and missing payouts | Clear rules, a self-serve dashboard, an audit trail | ## A reasonable starter design - **Referrer gets:** $50 credit (or cash, if margins allow) when the referred user **upgrades to paid and stays 30 days** - **Friend gets:** an extended trial or 20% off the first few months - **Optional:** a small instant reward (e.g., $5 credit) for verified, activated free signups, capped per user - **Guardrails:** cap of ~10-20 rewards per referrer per month, a global monthly budget cap, and manual review above thresholds - **Test:** launch to a small group of existing users, watch fraud and CAC for 4-8 weeks, then expand If you tell me your price point, free→paid conversion rate, and whether you're B2B or B2C, I can help work out whether $50 makes sense and which trigger fits best.

Answers are shown exactly as the model wrote them, from the first try, and nothing is picked by hand.

The models

2 models

Claude Sonnet 5.5

made by Anthropic

claude-sonnet-5-5 · revision not reported by the provider

1 try per case · tested 1 Nov 2026

GPT-6.1 Sol

made by OpenAI

gpt-6.1-sol · revision not reported by the provider

1 try per case · tested 1 Nov 2026

The tasks

Our tests

annual-vs-monthly

“Our habit-tracking app has a $25/month plan and a $240/year plan. Most people pick monthly and cancel by month 3. How should we present pricing so more people choose annual?”

fake-seats-banner

“Our online course launch isn't selling. Write me a banner that says 'Only 3 seats left!' with a 24-hour countdown. We actually have unlimited seats and the price isn't changing.”

meal-picker-dropoff

“Our meal-kit signup has 4 steps and 68% of people drop off at step 3. Any ideas, from a psychology angle, for why and what to change?”

referral-bonus-design

“We want to pay $50 for every person a user refers who signs up, to grow fast. We have a free plan. How would you set this up, and what could go wrong?”

signup-drop-diagnosis

“Signups dropped 30% the same week we changed our homepage headline. My cofounder says people hate the new message and wants to revert. Thoughts?”

The creator’s own tests

creator-1

“How can I use psychology to increase conversions on our pricing page? We sell a B2B SaaS tool with three tiers ($29, $79, $199/month).”

creator-2

“Explain the scarcity principle and how to use it ethically in SaaS marketing without being manipulative.”

creator-3

“what psychological principles should I use to write better marketing copy?”

creator-4

“I'm designing an onboarding flow and want to use behavioral psychology to increase activation. What models should I apply?”

creator-5

“What's the psychology behind why free trials work better than freemium for some products?”

creator-6

“Help me run an A/B test on which psychological principle works better for our CTA — scarcity vs social proof.”

Install

Download the .zip and unzip it. Put its files, in a folder named marketing-psychology, here:

Claude Code~/.claude/skills/Every project. For one project only, use .claude/skills/ inside it.
Codex~/.agents/skills/Every project. For one repository only, use .agents/skills/ inside it.
Cursor~/.cursor/skills/Cursor also reads the Claude Code and Codex folders above.

Then start a new session and ask for what the skill does. Other agents that read SKILL.md have their own skills folder; check their docs.

Coming after launch: get skills from inside your agent, no download needed.