06. Validation Experiments & Hypotheses
Testing the viability of the LocalPerks coalition model before scaling.
Validation Strategy: The "Neighborhood Density" Approach
The core risk of LocalPerks is the Chicken-and-Egg problem: Consumers won't join without businesses, and businesses won't join without consumers. Furthermore, the Coalition Settlement model (Shop A paying Shop B for redeemed points) requires high trust. Our validation strategy focuses on proving network effects in a single micro-geography before building the full platform.
1. Core Hypotheses
Hypothesis #1: The Coalition Desire
CRITICALWe believe that independent retailers (coffee, books, boutique)
Will agree to cross-honor rewards from neighboring competitors
If we provide a unified system that drives new foot traffic
We will know this is true when >60% of surveyed merchants agree to a theoretical "settlement model" where they pay for redemptions.
Hypothesis #2: Consumer Utility
CRITICALWe believe that local shoppers
Will consolidate spending to coalition members
If we offer a single app with accelerated earning velocity
We will know this is true when we see a 20% cross-pollination rate (earning at Store A, redeeming at Store B) in our manual MVP.
Hypothesis #3: Business Model
HIGH RISKWe believe that small business owners
Will pay a $29-$59 monthly subscription
If we demonstrate it is cheaper than running their own loyalty program
We will know this is true when we secure 10 Letters of Intent (LOI) with payment details on file.
Hypothesis #4: Operational Friction
MEDIUMWe believe that retail staff
Will consistently ask "Are you a LocalPerks member?"
If we make the scanning process under 3 seconds
We will know this is true when scan compliance exceeds 80% during peak hours in pilot.
2. Experiment Catalog
| Experiment | Hypothesis | Methodology | Metrics | Cost/Time |
|---|---|---|---|---|
| Exp #1: The "Paper Coalition" (Concierge MVP) | #1, #2 | Create a unified paper stamp card valid at 3 non-competing shops (Coffee, Book, Bakery) on one block. Shops stamp the same card. Full card = $5 off at any of the 3. We manually settle the cash difference weekly. | 1. Cross-redemption rate (redeemed at different shop than earned) 2. Merchant feedback on settlement |
$200 (printing) 3 Weeks |
| Exp #2: Merchant Sales "Dry Run" | #3 | Approach 20 businesses in a target neighborhood with a pitch deck and a "Letter of Intent" contract. Ask for credit card info to hold their spot for launch. | 1. Conversion to LOI 2. Objection analysis (Price vs. Complexity) |
$0 1 Week |
| Exp #3: Sidewalk Intercepts | #2 | Stand outside target coffee shops. Ask customers to show their current loyalty cards. Show a mockup of LocalPerks. Ask if they would trade 3 separate cards for this one. | 1. % of customers with >3 local cards 2. % expressing "Strong Interest" |
$100 (Gift cards) 3 Days |
| Exp #4: Fake Door Ad Campaign | #1 | LinkedIn/FB ads targeting local business owners: "Join the [City Name] Loyalty Coalition - Stop losing customers to Starbucks." Leads to waitlist. | 1. CTR 2. Cost Per Lead (CPL) 3. Landing Page Conversion |
$500 (Ads) 2 Weeks |
3. Prioritization Matrix
Logic
Priority 1: Paper MVP
High Effort, Critical Impact. It's the only way to test the "Coalition" behavior without building software. If businesses refuse to honor stamps from neighbors, the model fails.
Priority 2: Sales Run
Low Effort, High Impact. Validates B2B willingness to pay before writing code.
4. 8-Week Validation Sprint
Minimum Success Criteria
| Metric | Target |
| Pilot Cross-Redemption Rate | >15% |
| Merchant Settlement Compliance | 100% |
| Signed LOIs (Pre-Sales) | 10+ |
| Ad CAC (Business Lead) | < $50 |
Pivot Triggers
Trigger: Merchants won't pay subscription.
Pivot: Switch to "Transaction Fee Only" model (higher % on redemption), removing barrier to entry.
Trigger: Consumers collect points but don't cross-redeem.
Pivot: Abandon "Coalition" model; pivot to "White Label Loyalty" for individual shops (competing with Square/Toast but cheaper).
Trigger: Settlement is too messy/distrusted.
Pivot: Switch to "Perks Only" model (flash your card for 10% off) removing the points ledger entirely.
Experiment Documentation Standard
**Status:** [Complete/In-Progress]
**Cost:** $[Amount]
**Observation:** [What actually happened? Be specific with numbers.]
**Learning:** [Did we validate the hypothesis?]
**Decision:** [Iterate / Kill / Scale]