💰 Tiered Pricing Optimization for a Storage-as-a-Service Product

How Data Science Can Add $5.6M–$6.9M to Annual Revenue

This analysis uses regression discontinuity and price elasticity modeling to find where we're leaving money on the table—and how to capture it.

Note: company details have been removed, and the charts on this page are drawn from simulated data for illustration.

Executive Summary: What We Found

  • $6.2M/year lost to "threshold gaming" - customers staying at 99 TB to avoid 100 TB pricing jump
  • 40% of customers are bunching just below tier thresholds to save money
  • Healthcare, Finance & Government are price-insensitive (elasticity -0.3 to -0.6) and can absorb 6–10% increases
  • Retail customers need 8% discount to prevent churn (elasticity: -2.0)
  • Quick win: Move Tier C/D boundary from 100 TB → 120 TB = +$1.5M/year net

📊 At-A-Glance: The Opportunity

$5.6–6.9M
Annual Revenue Opportunity
2,000
Customers Analyzed
3-6
Months to Implement
6
Industry Segments

🎯 Part 1: The Threshold Gaming Problem

In plain English: Our customers are smart. They know that going from 99 TB (Tier C) to 101 TB (Tier D) can raise their monthly bill by about 70% (~$6,500 → ~$11,000). So what do they do? They stop at 99 TB—even when they actually need more storage.

The Evidence: Customer "Bunching"

💡 What This Chart Tells Us

See those spikes at 19 TB, 49 TB, 99 TB, and 249 TB? Those are NOT natural. Companies don't naturally need exactly 99 TB. They're gaming the system to stay just under the threshold and avoid the price jump.

The cost: 482 customers at 90-99 TB who probably need 110-120 TB = $6.2M/year in lost revenue.

Comparing Customers Near the 100 TB Threshold

Let's use regression discontinuity - comparing customers just below vs. just above 100 TB:

💡 What This Means

Customers at 95-99 TB (Tier C) pay ~$6,500/month and are happy (8% churn rate).
Customers at 100-105 TB (Tier D) pay ~$11,000/month and churn MORE (15% churn rate).

The problem: The price jump at 100 TB causes sticker shock. Customers either stop growing their storage or leave entirely.

✅ Recommended Action #1: Smooth the Tier C/D Transition

Option A: Move threshold from 100 TB → 120 TB

  • Customers at 95-99 TB can now grow to 110 TB without price shock
  • Expected revenue gain: +$2.3M/year
  • Risk: Existing Tier D customers (100-120 TB) pay slightly less: -$800K/year
  • Net impact: +$1.5M/year

Option B: Add new Tier C+ (75-150 TB at $55/TB)

  • Creates smoother price curve - no cliff
  • Expected revenue gain: +$2.8M/year
  • Risk: More complexity in pricing communication

🎢 Part 2: Not All Customers Are Equal

In plain English: Healthcare customers care most about reliability and compliance, so price matters less to them. Retail customers are highly price-sensitive and would likely move to a public cloud provider after a 5% increase. One price for every segment leaves money on the table.

Price Sensitivity by Industry (Elasticity)

💡 How to Read This Chart

Elasticity = "If I raise price 10%, how much demand do I lose?"

  • Government (-0.3): Raise prices 10% → Lose 3% of customers = +7% revenue
  • Healthcare (-0.4): Raise prices 10% → Lose 4% of customers = +6% revenue
  • Retail (-2.0): Raise prices 10% → Lose 20% of customers = -12% revenue

Translation: Healthcare/Gov will pay more. Retail needs discounts.

What Happens If We Adjust Prices by Segment?

Industry Current Customers Elasticity Recommended Action Expected Revenue Impact
Healthcare 527 -0.4 (Inelastic) Raise prices +8% +$1.4M/year
Financial Services 366 -0.6 (Inelastic) Raise prices +6% +$1.4M/year
Government 105 -0.3 (Very Inelastic) Raise prices +10% +$580K/year
Manufacturing 408 -1.2 (Moderate) Keep current pricing $0
Technology 298 -0.8 (Moderate) Keep current pricing $0
Retail 296 -2.0 (Elastic) Offer -8% discount +$760K/year (volume gain outweighs the discount)

✅ Recommended Action #2: Segment-Specific Pricing

Implementation:

  1. Raise prices 8% for Healthcare, 6% for Financial Services, 10% for Government
  2. Keep Manufacturing and Technology at current prices
  3. Offer 8% "volume commitment discount" to Retail (positioned as loyalty benefit)

Net revenue impact: +$1.4M + $1.4M + $580K + $760K = +$4.1M/year

⚠️ Risk: Price Discrimination Concerns

Pricing by industry alone raises fairness and legal concerns. Tie each price difference to the value or commitment the customer receives instead:

  • Healthcare/Finance: "Premium compliance and security tier" (+8-10%)
  • Retail: "High-volume commitment discount" (-8%)
  • Everyone else: Standard pricing (no change)

🎮 Part 3: Try It Yourself - Revenue Simulator

Interactive tool: Adjust the sliders below to see how price changes affect revenue in real-time.

Healthcare Segment: How much can we raise prices?

Current: 527 customers, avg $4,800/month, elasticity -0.4

+8% price increase

📊 Impact at +8% Price Increase

Customer loss: ~17 customers churn (3.2%)

Revenue change: +$1.4M/year

Why it works: Healthcare customers value uptime and compliance over cost. Small churn, big revenue gain.

Retail Segment: Should we offer discounts?

Current: 296 customers, avg $3,200/month, elasticity -2.0 (price sensitive!)

-8% price change

📊 Impact at -8% Discount

Customer gain: +47 customers (16% growth)

Revenue change: +$0.8M/year

Alternative: If we raise +5%, we lose ~30 customers and -$0.6M/year

Strategy: Offer "volume commitment discount" to retain price-sensitive retail customers.

Tier C/D Threshold: Where should we draw the line?

Current: Tier C ends at 100 TB, Tier D starts at 100 TB

Tier D starts at 120 TB

📊 Impact of Moving to 120 TB

Customers affected: 482 currently at 90-99 TB can now grow without price shock

Expected growth: ~200 customers expand storage by 15-20 TB

Revenue gain: +$2.3M/year

Trade-off: ~150 customers currently in Tier D (100-120 TB) get lower rate: -$800K/year

Net impact: +$1.5M/year

🚀 Summary: A $5.6M–$6.9M Revenue Plan

Combined Strategy

Initiative What We Do Annual Revenue Impact Implementation Time
1. Fix Tier Thresholds (Option A) Move Tier C/D boundary to 120 TB +$1.5M 2-3 months
2. Segment Pricing Raise Healthcare +8%, Finance +6%, Gov +10% +$3.4M 3-4 months
3. Retail Retention Offer 8% "commitment discount" to Retail +$760K 1-2 months
1b. Add Tier C+ (Option B, instead of 1) New tier at 75-150 TB smooths curve +$2.8M (replaces +$1.5M) 4-6 months
TOTAL ANNUAL IMPACT +$5.6M (Option A) / +$6.9M (Option B) 6 months

Next Steps: 90-Day Pilot

✅ Phase 1: Pilot (Months 1-3)

  1. Week 1-2: Legal/compliance review of segment-specific pricing approach
  2. Week 3-4: Select pilot customers (50 Healthcare, 30 Retail, 50 near 100 TB threshold)
  3. Month 2: Implement pricing changes for pilot group
  4. Month 3: Measure results: churn rate, expansion rate, revenue change

✅ Phase 2: Rollout (Months 4-6)

  1. Month 4: Analyze pilot results, adjust approach if needed
  2. Month 5: Create sales communication materials and train account teams
  3. Month 6: Full rollout to all 2,000 customers
  4. Ongoing: Monitor monthly - track revenue, churn, customer satisfaction

💡 Why This Works: The Science Behind It

These recommendations aren't guesses—they're based on two proven economic techniques:

  • Regression Discontinuity: We compare customers just below vs. just above pricing thresholds. Since they're nearly identical except for price, any difference is caused by the pricing itself.
  • Price Elasticity Modeling: We measure historical response to price changes. Healthcare customers barely react (elasticity -0.4). Retail customers are 5x more sensitive (elasticity -2.0).

This is the same approach used by: Amazon (dynamic pricing), Netflix (tier optimization), and airlines (segment-based pricing). It's proven to work.

⚠️ Risks to Monitor

  • Customer backlash: If word spreads that industries pay different prices, could cause trust issues. Mitigation: Package as "compliance tier" vs "standard tier" vs "volume discount."
  • Competitor response: AWS/Azure might undercut us. Mitigation: Emphasize on-premises security and compliance advantages.
  • Internal complexity: Sales teams need training on new tiers. Mitigation: Create clear playbooks and tools.

Possible Follow-up Work

Natural extensions of this analysis:

Bottom line: Data science isn't about fancy algorithms—it's about finding money we're leaving on the table and building systems to capture it. This analysis shows one example. There are dozens more opportunities like this hiding in our data.