Whistl SpendingShield: AI-Powered Spending Protection Explained

SpendingShield is Whistl's flagship AI feature. It analyses 27 risk signals in real-time and automatically increases protection when you're at high risk of impulse spending. Here's exactly how it works.

What Is SpendingShield?

SpendingShield is Whistl's AI-powered spending protection system. Unlike static blocking apps, SpendingShield:

• Learns your spending patterns over time
• Detects high-risk periods before you spend
• Automatically increases protection when risk is high
• Adapts to your behaviour and feedback

"SpendingShield caught me at 11pm when I was about to drop $800 on a shopping binge. The app said 'this looks like an impulse—wait 24 hours.' I slept on it and cancelled the order the next morning." — Jake, 29

The 27 Risk Signals SpendingShield Analyses

SpendingShield analyses 27 different signals to assess your spending risk. These are grouped into 7 categories:

1. Temporal Signals (Time-Based)

• Time of day (late night = higher risk)
• Day of week (weekends = higher risk for many)
• Days since payday (days 1-3 after payday = higher risk)
• Time since last purchase (rapid purchases = higher risk)
• Time since last gambling/spending block trigger

2. Velocity Signals (Spending Speed)

• Amount spent in last hour
• Amount spent in last 24 hours
• Amount spent in last 7 days
• Number of transactions in last hour
• Spending velocity vs. your average

3. Location Signals

• Distance from home (unusual locations = higher risk)
• Proximity to gambling venues
• Proximity to shopping districts
• Whether location matches typical spending patterns

4. Merchant Signals

• Merchant category (gambling, shopping, dining, etc.)
• Merchant risk level (high-risk merchants flagged)
• New vs. familiar merchant (new = higher risk)
• Online vs. in-person (online = higher risk for many)

5. Amount Signals

• Transaction amount vs. your average
• Transaction amount vs. budget remaining
• Transaction amount vs. account balance
• Round number amounts (often indicate impulse)

6. Behavioural Signals

• App usage patterns (frantic usage = higher risk)
• Hesitation before purchase (detected via app interaction)
• Repeated attempts to access blocked merchants
• Time spent browsing before purchase

7. Contextual Signals

• Recent life events (if shared)
• Stress indicators (if connected to health data)
• Sleep quality (if connected to health data)
• Calendar events (high-stress periods)

How SpendingShield Makes Decisions

Step 1: Signal Collection

When you attempt a purchase or open a shopping/gambling app, SpendingShield collects all 27 signals in real-time.

Step 2: Individual Signal Scoring

Each signal is scored from 0-100 based on risk level:

Example Signal Scores:
- Time of day (11pm): 75/100 (high risk)
- Spending velocity: 80/100 (high risk)
- Merchant category: 90/100 (gambling = very high risk)
- Location: 30/100 (at home = low risk)
- Amount: 65/100 (above average = moderate-high risk)

Step 3: Weighted Composite Score

Signals are weighted based on their predictive power for YOUR behaviour. The AI learns which signals matter most for you.

Weighted Example:
- Temporal (15% weight): 75 × 0.15 = 11.25
- Velocity (20% weight): 80 × 0.20 = 16.00
- Merchant (25% weight): 90 × 0.25 = 22.50
- Location (10% weight): 30 × 0.10 = 3.00
- Amount (15% weight): 65 × 0.15 = 9.75
- Behavioural (10% weight): 70 × 0.10 = 7.00
- Contextual (5% weight): 50 × 0.05 = 2.50

COMPOSITE RISK SCORE: 72/100 (HIGH RISK)

Step 4: Threshold Comparison

The composite score is compared to thresholds:

• 0-30: Low risk → No intervention
• 31-60: Medium risk → Gentle reminder
• 61-85: High risk → Cooling-off timer
• 86-100: Critical risk → Block + partner notification

Step 5: Intervention Triggered

Based on risk level, SpendingShield triggers appropriate intervention:

SpendingShield Interventions

Low Risk (0-30)

• No intervention
• Purchase proceeds normally

Medium Risk (31-60)

• Gentle reminder notification
• "This looks like it might be an impulse purchase. Are you sure?"
• User can proceed with one tap

High Risk (61-85)

• Cooling-off timer activated (1-24 hours)
• "We've detected high-risk spending. This purchase is on hold for 24 hours."
• User can override but requires additional confirmation
• Partner may be notified (based on settings)

Critical Risk (86-100)

• Purchase blocked
• Partner immediately notified
• Requires partner approval to proceed
• Protected Floor enforced if applicable

How SpendingShield Learns and Improves

Learning from Outcomes

SpendingShield tracks what happens after interventions:

• Did you override the block?
• Did you express regret later?
• Did you thank the app for the intervention?

Over time, it learns which interventions work best for you.

Learning from Patterns

SpendingShield identifies your personal risk patterns:

• "User consistently overspends on Saturday nights"
• "User's risk increases significantly after 10pm"
• "User rarely regrets purchases under $50"

Protection is tailored to YOUR patterns, not generic rules.

Learning from Feedback

When you provide feedback ("This was helpful" or "This was unnecessary"), SpendingShield adjusts:

• Thresholds are adjusted
• Signal weights are updated
• Intervention types are optimised

SpendingShield vs. Traditional Blocking

Real User Results

Case Study: 73% Reduction in Impulse Spending

Who: Emma, 31, Sydney

Before: $2,000-3,000/month impulse spending

After SpendingShield: $500-800/month

Quote: "SpendingShield knows me better than I know myself. It catches me when I'm vulnerable. I've saved over $20,000 this year."

Case Study: Gambling Recovery

Who: Marcus, 29, Brisbane

Before: $94,000 lost to sports betting

After SpendingShield: 22 months gambling-free

Quote: "SpendingShield detected when I was near gambling venues and increased protection. It knew I was at risk before I admitted it to myself."

Privacy and SpendingShield

SpendingShield processes data on-device where possible. Your transaction data doesn't leave your phone unless you enable cloud backup.

• On-device processing: AI runs locally on your phone
• Encryption: All data encrypted at rest and in transit
• User control: You control what data is shared
• No third-party selling: Your data is never sold

Setting Up SpendingShield

1. Download Whistl from iOS App Store
2. Connect your bank accounts
3. Allow 2-4 weeks for AI to learn your patterns
4. Configure intervention preferences in Settings
5. Set up accountability partner for maximum effectiveness

Conclusion: Protection That Adapts to You

SpendingShield isn't static blocking. It's intelligent protection that learns you, adapts to you, and protects you when you need it most.

Let AI handle the pattern detection. You handle the life.

Related: AI in Personal Finance | Protected Floor Explained | Spending Blocker Apps Review