Demand is not uniform. On any given night, some hotels in a market fill up quickly and at premium rates, while others struggle to convert lookers into bookers even at deep discounts. Understanding why — and predicting when demand will favour your property — is the core challenge of hotel revenue management.
The Market Demand Index (MDI) is the industry’s standard tool for measuring where a hotel stands relative to available market demand. But the MDI is a lagging indicator; it tells you where you have been. The forward-looking signals that drive future demand — booking pace, search volume, event calendars, macroeconomic conditions, flight data — are where the real competitive advantage lies. Propeter’s AI platform ingests and synthesises all of these signals across a 365-day forecasting horizon, enabling hotels to act on demand intelligence before it shows up in any traditional index.
Day forward forecasting horizon in Propeter’s demand engine
AI agents in Propeter’s orchestration pipeline
Average RevPAR improvement from AI demand monitoring
What Is the Market Demand Index (MDI)?
The Market Demand Index — also referred to as the Market Penetration Index (MPI) or Occupancy Index — is one of three standard performance indices tracked by hotels using competitive benchmarking data. It is calculated as:
MDI = (Hotel Occupancy % / Compset Average Occupancy %) × 100
An MDI of 100 means the hotel is capturing exactly its fair share of market demand. An MDI of 115 means the hotel achieved 15% higher occupancy than the compset average. An MDI of 88 means it under-captured demand by 12%.
The MDI is part of a trio of indices alongside the Average Rate Index (ARI) and Revenue Generation Index (RGI). Together, they provide a complete picture of competitive performance: MDI shows demand capture, ARI shows rate positioning, and RGI shows overall revenue performance. A hotel with a high MDI but low ARI is filling rooms cheaply — capturing demand but at a rate discount. A hotel with high ARI and moderate MDI may be pricing itself out of some demand while maximising rate on the bookings it does capture. The optimal balance between these indices defines the hotel’s revenue strategy.
STR Data and How MDI Is Calculated
The MDI calculation requires competitive set data — specifically, the occupancy performance of the hotels against which you benchmark. In most markets, this data is sourced from STR (now CoStar), which aggregates and anonymises performance submissions from participating hotels.
How STR Benchmarking Works
Hotels submit their daily performance data (rooms sold, rooms available, revenue) to STR, which aggregates it across predefined competitive sets and returns each hotel’s performance relative to its compset. Because the data is anonymised, individual competitor performance cannot be identified — but the compset averages provide a statistically robust benchmark.
STR reports are typically delivered monthly (in the standard STAR report), but many revenue management systems — including those integrated with Propeter — access weekly or even daily STR data feeds, enabling more frequent MDI tracking and faster response to demand shifts.
Compset Selection for MDI Accuracy
The accuracy and relevance of your MDI depends entirely on the quality of your competitive set definition. If your compset includes hotels that are not genuine competitors (different market tier, different location, different demand generators), the MDI will be misleading. Propeter’s team works with hotels during onboarding to validate compset composition, ensuring MDI benchmarks reflect true competitive dynamics.
Types of Demand Signals
The MDI tells you about past demand capture. Forward-looking demand signals tell you about future demand. There are several categories of signals that hotel revenue managers should monitor:
Transient Booking Pace
The rate at which individual (non-group) reservations are accumulating for future dates, compared to the same point in the booking window in prior periods. Accelerating pace signals strengthening demand; decelerating pace signals weakness. This is the highest-frequency, most operationally actionable demand signal available.
Group Booking Pace
For hotels with significant group segments (conference, events, weddings), group bookings often arrive many months in advance. Monitoring group pace against historical norms provides early warning of periods where the base occupancy from group business is stronger or weaker than expected, informing transient rate strategy for those periods.
Search Volume and Intent Signals
Increases in OTA search queries for your destination, or in Google Hotels visibility data, are leading indicators of future booking demand. These signals typically lead actual bookings by 5–15 days, giving revenue managers a short but usable advance window to pre-emptively adjust rates.
Event and Calendar Demand Drivers
Major events — conferences, festivals, sports, graduations — create predictable demand spikes. Monitoring event calendars 90–180 days in advance allows hotels to price these periods appropriately before demand becomes visible in booking pace. Propeter’s web intelligence layer aggregates event data from multiple sources, flagging high-impact events automatically for each market.
Macroeconomic and External Signals
Consumer confidence, corporate travel policy changes, airline capacity decisions, and exchange rate movements all affect hotel demand at a macro level. While these signals are harder to incorporate in day-to-day rate decisions, they are essential context for medium and long-term demand forecasting.
Propeter’s demand forecasting engine uses a hybrid of XGBoost gradient boosting and LSTM neural networks. XGBoost excels at capturing non-linear relationships between structured demand signals; LSTM handles temporal patterns and sequential dependencies in booking pace data. Together, they produce demand forecasts that are materially more accurate than single-method approaches.
Interpreting Your MDI
Raw MDI numbers become meaningful when tracked over time and segmented by date type. A hotel might have an overall MDI of 102 (marginally above index) but find significant variation when looking at weekdays versus weekends, shoulder season versus peak, or event periods versus standard periods.
Trend Analysis
A declining MDI trend — even if still above 100 — is a warning signal. It means the hotel is gradually losing demand share to its compset. This can result from rate increases that have exceeded the market’s willingness to pay, declining distribution visibility on key OTA channels, product deterioration relative to improved competitors, or loyalty programme weakness. Identifying the cause requires correlating the MDI decline with changes in rate positioning (ARI), review scores, and channel mix data.
Segment-Level MDI
Some hotels track separate demand indices by segment — transient MDI, group MDI, and corporate MDI. This granularity reveals whether demand underperformance is market-wide or concentrated in a specific segment, enabling more targeted responses. For example, a hotel with strong overall MDI but weak corporate MDI may need to review its corporate rate programme rather than adjusting BAR rates.
Above and Below Index: Implications and Responses
Being above MDI index is generally positive, but context matters. A hotel that is significantly above index (MDI of 120+) may be selling out too early at rates that are too low — capturing more than its fair share of demand by underpricing relative to the market. In this scenario, the right response is to raise rates, not to celebrate the high MDI.
Being below MDI index (under 90) indicates demand underperformance. The appropriate response depends on diagnosis:
- Pricing too high relative to product: Rates need to be more competitive with compset peers offering similar value
- Distribution gap: The hotel may not be visible on key OTA channels or may have rate parity issues suppressing OTA visibility
- Product or reputation issues: Low review scores reduce conversion from lookers to bookers regardless of rate
- Loyalty programme weakness: Competitors with stronger loyalty programmes are winning repeat business that bypasses the hotel
AI Demand Monitoring and Propeter’s Demand Agents
Propeter’s 6-agent AutoGen AI orchestration pipeline includes a dedicated Demand Forecasting Agent that synthesises all available demand signals into forward occupancy forecasts for every date within the 365-day horizon. Unlike traditional rule-based systems that respond to past performance, Propeter’s forecasting layer uses XGBoost and LSTM neural networks to predict future demand based on the combination of booking pace, search intent, event overlay, historical patterns, and macroeconomic context.
When the Demand Forecasting Agent projects occupancy above or below the hotel’s target thresholds, it triggers the Rate Optimisation Agent — which evaluates the appropriate rate response and submits it through the 13-stage rate engine: Base Rate, Inventory, Rate Plan, Derived Rates, Promotion, Loyalty Discount, Voucher, Referral, Flash Deal, Stacking Resolver, Guardrails, Upsell, and Tax and Fee. The result is a rate that reflects genuine demand conditions — not just a historical pattern or a manual override.
For revenue managers, this means the MDI is no longer the primary planning tool — it becomes a retrospective validation of AI-driven decisions that were already made based on forward signals. Hotels using Propeter consistently achieve MDI performance above 100, not by discounting to fill rooms, but by anticipating demand and being priced appropriately when high-value guests are ready to book.
The shift from MDI-based reactive management to AI-driven predictive demand management is the single most impactful operational change available to hotel revenue teams today. Hotels that make this transition consistently achieve sustained RevPAR improvements of 18–25% — not through rate cuts, but through better timing of rate decisions across the full demand curve.
Frequently Asked Questions
What is the Market Demand Index (MDI) in hospitality?
The Market Demand Index (MDI) — also called the Occupancy Index or MPI (Market Penetration Index) — compares a hotel’s occupancy percentage to the average occupancy of its competitive set. An MDI above 100 means the hotel is capturing more than its fair share of market demand; below 100 means underperformance relative to the compset.
What are the main types of demand signals hotels should monitor?
Hotels should monitor transient booking pace, group booking pace, search volume trends on OTAs and Google Hotels, event calendars, flight search data, competitor occupancy signals, and historical seasonal patterns. Propeter’s demand agents aggregate all of these signal types into a unified demand score for each future date.
How is STR data used to calculate demand indices?
STR (now CoStar) aggregates anonymised performance data from hotels in a competitive set and calculates market averages for occupancy, ADR, and RevPAR. A hotel’s MDI is calculated by dividing its own occupancy by the compset average occupancy and multiplying by 100. An MDI of 110 means the hotel achieved 10% higher occupancy than the compset average.
What does it mean when a hotel’s MDI is below 100?
An MDI below 100 means the hotel is underperforming its compset on occupancy — it is not capturing its fair share of available demand. This can result from pricing that is too high relative to the product, poor OTA visibility, weak distribution strategy, or product issues. Propeter’s demand agents flag MDI underperformance and surface the likely contributing factors, helping revenue managers diagnose and address the root cause.
Stay ahead of market demand with AI
Propeter’s demand agents monitor 365 days of forward signals continuously — giving your hotel the intelligence to price right, every night.
