Recommendation: Raise prices by 8–12% across all listing tiers.
Analysis of 358 active Airbnb listings in Albany, NY reveals that demand is inelastic — higher-priced listings attract more bookings, not fewer. A 10% price increase is projected to grow revenue by ~12.5% across all tiers, with positive expected outcomes even in the worst-case simulation scenario. The data supports price increases, particularly for Premium and Luxury listings where absolute revenue gains are highest.
Should short-term rental hosts raise, lower, or hold prices?
The conventional assumption is that higher prices reduce demand. This analysis tests that assumption empirically using occupancy and booking frequency data from Albany's Airbnb market. The answer has direct implications for pricing strategy: if demand is inelastic, hosts leave money on the table by pricing conservatively.
Dataset: Inside Airbnb — Albany, NY (February 2026 scrape, 453 raw listings)
Clean sample: 358 active listings with valid occupancy and review data
The dataset's price column was unpopulated. Guest capacity (accommodates)
is used as a price tier proxy — the primary structural driver of nightly
rate on Airbnb. Listings are segmented into four tiers:
| Tier | Guest Capacity | Listings | Estimated Nightly Rate |
|---|---|---|---|
| Budget | 1–2 guests | 179 | ~$80/night |
| Mid | 3–4 guests | 98 | ~$140/night |
| Premium | 5–6 guests | 45 | ~$220/night |
| Luxury | 7+ guests | 36 | ~$350/night |
Elasticity method: Log-log linear regression of annual occupancy on guest capacity. The coefficient β estimates how demand responds to a change in price tier.
Scenario method: 1,000 Monte Carlo simulations per scenario, sampling from the elasticity confidence interval. Results report mean expected revenue and 95% confidence intervals.
Elasticity (β): 0.2294
95% CI: [0.0513, 0.4076]
p-value: 0.0120
R²: 0.0176
The elasticity coefficient is positive and statistically significant (p = 0.012). In Albany's Airbnb market, larger and higher-priced listings attract more annual bookings, not fewer.
A 1% increase in listing size (price tier) is associated with a 0.23% increase in annual occupancy. Demand does not collapse when prices rise — it grows with them.
This is classified as inelastic: price increases grow revenue because volume holds up or increases alongside the price change.
| Tier | N | Avg Occupancy (nights/yr) | Avg Reviews/Month |
|---|---|---|---|
| Budget (1–2p) | 179 | 89.1 | 1.66 |
| Mid (3–4p) | 98 | 102.9 | 2.28 |
| Premium (5–6p) | 45 | 117.6 | 2.26 |
| Luxury (7+p) | 36 | 131.8 | 2.51 |
Luxury listings average 131.8 nights booked per year — 48% more than Budget listings at 89.1 nights. This pattern holds across both occupancy and reviews per month (booking frequency signal).
Premium and Luxury hosts do not trade off higher prices against lower volume. They command higher rates and fill more nights — a compounding revenue advantage over Budget tier hosts.
Monte Carlo scenario modeling (1,000 simulations):
| Tier | Base Revenue | +10% Revenue | Change | Downside Risk (5th pct) |
|---|---|---|---|---|
| Budget (1–2p) | $7,126 | $8,020 | +12.5% | +11.0% |
| Mid (3–4p) | $14,410 | $16,212 | +12.5% | +10.7% |
| Premium (5–6p) | $25,872 | $29,105 | +12.5% | +10.9% |
| Luxury (7+p) | $46,122 | $51,893 | +12.5% | +10.9% |
Two results stand out:
1. Every tier shows positive revenue under a price increase. In elastic markets, revenue falls when prices rise. Albany's inelastic demand means hosts can raise prices without losing revenue gains to volume drops.
2. Even the downside risk is positive. The 5th percentile outcome — the worst 5% of simulations — still shows +10.7% to +11.0% revenue growth. Even if the elasticity estimate understates price sensitivity, a 10% price increase is still expected to grow revenue.
Lowering prices by 10% produces the inverse: revenue declines ~12% across all tiers, with downside risk reaching -13.4%.
Raise prices by 8–12% across all listing tiers.
| Tier | Annual Revenue Gain (+10%) | Priority |
|---|---|---|
| Luxury (7+p) | +$5,771 | 1 — highest absolute gain |
| Premium (5–6p) | +$3,233 | 2 — strong inelastic demand |
| Mid (3–4p) | +$1,802 | 3 — consistent positive signal |
| Budget (1–2p) | +$894 | 4 — smallest gain, still positive |
Implementation:
- Apply increases gradually: 4–6% immediately, remainder over 60 days
- Monitor occupancy weekly after each increase
- If occupancy drops more than 15% within 60 days, revert to prior pricing
- Re-run analysis quarterly as market conditions shift
- Price proxy: Guest capacity is a structural proxy for price, not actual nightly rate. Revenue estimates use market-rate assumptions (~$80–$350/night by tier), not observed prices.
- Causal inference: Regression estimates association, not causation. Location, amenities, and host quality confound the occupancy-capacity relationship.
- Geography: Results reflect Albany, NY only. Elasticity patterns differ across markets.
- Seasonality: Analysis uses annual occupancy aggregates. Seasonal pricing strategy is not modeled.
pricing-analytics-case-study/
├── README.md ← This case study
├── analysis.ipynb ← Jupyter notebook: EDA → regression → scenarios
├── data/
│ └── airbnb_listings.csv ← Inside Airbnb, Albany NY, Feb 2026
└── output/
├── 01_eda.png ← Distribution and relationship charts
├── 02_elasticity.png ← Log-log regression with 95% CI
├── 03_cohort_analysis.png ← Occupancy and booking frequency by tier
└── 04_scenario_analysis.png ← Revenue under three pricing scenarios
Python · pandas · numpy · scipy · matplotlib · seaborn · Jupyter
Data source: Inside Airbnb — publicly available.