Reference · Cheat Sheet 09
Signaling & Screening
One rule: Hidden information? If you know more, signal — take a costly, hard-to-fake action that only a high-quality actor can rationally afford. If you know less, screen — design a menu that makes the other side self-select and reveal their type through their choice.
The four-step decision
- Who holds the private information? You → informed side (signal). The other party → uninformed side (screen).
- Signal credibility test — single-crossing. Ask: is this action costlier for a low-quality type than a high-quality type? If yes → credible signal, it separates the types. If anyone can copy it cheaply → cheap talk, worthless.
- Design the screen. Offer a menu with options where each type finds a different option optimal. Don't ask — let them choose. The choice reveals the information you couldn't extract by asking directly.
- Watch for adverse selection. If you can't signal or screen, expect bad types to crowd out good ones. The symptom: you attract exactly the high-cost, low-value customers.
Signal vs. screen — side-by-side
Signaling (you know more)
- Board certification / fellowship
- Published surgical outcomes (GTR rates, KPS, complications)
- Money-back guarantee on SaaS
- Warranty on a product
- Free trial on genuinely good features
- Talks at prestigious conferences
Screening (you know less)
- Free / Pro / Expert SaaS tiers
- Insurance deductible options
- Hardcover then paperback release
- Early-bird vs. full course price
- Probationary period for a new hire
- Tiered-access landing pages
The signal credibility test (worked)
Before offering any signal, run this check
| Action | Cost to high type | Cost to low type | Credible? |
| Money-back guarantee | Low (few returns on good product) | Ruinous (constant returns) | Yes |
| Board certification | Manageable (genuinely capable) | Very high or impossible | Yes |
| Published outcomes | Low (strong numbers) | High (exposes poor results) | Yes |
| "Best quality!" banner | Near zero | Near zero | No — cheap talk |
| Verbal "trust me" | Zero | Zero | No — cheap talk |
Why SaaS tiers are a screening menu
Free / Pro / Expert — each tier extracts a type through self-selection
| Tier | Who self-selects | Information revealed |
| Free | Casual, exploring, low commitment | Low willingness-to-pay; early funnel |
| Pro ($29) | Regular user, clear time savings | Moderate WTP; real engagement |
| Expert ($59) | Power user, busy practice or teaching | High WTP; high volume; reveal by clicking |
A single mid-price forces you to guess average WTP and leaves money on the table from both ends. The tiered menu captures more total surplus because each type pays closer to their true value.
Adverse selection warning: If you can't signal or screen effectively, you get the wrong mix. A surgeon who can't credibly signal quality competes only on price — and attracts the price-sensitive, high-complexity patients who generate the most costs. A SaaS without tiers prices for the average user and loses both the casual user (overpriced) and the power user (underpriced). Signaling and screening are the cure.