Compression Events: How to Spot the Weeks That Will Sell Out Your Market

Sources: STR (CoStar) compression-night methodology and class-level ADR premium data, STR Weekly Insights for Super Bowl LVIII and CES 2024, CoStar reporting on SXSW 2023 and F1 Austin/Las Vegas, PriceLabs Coachella 2024 analysis, Lighthouse Oktoberfest 2025 data, Hotel Management Mardi Gras occupancy reporting, Cornell Hospitality Quarterly meta-analysis on revenue-management research, MDPI peer-reviewed demand modeling. Last reviewed: 2026-04-21.

Key takeaways

A compression event is not "an event with a crowd." It is a recurring week where market-wide demand exceeds supply sharply enough that occupancy crosses into STR's compression band (90% or 95% depending on which threshold you use) and the entire compset prices together. STR's own convention is that a "compression night" is any night with occupancy at or above 95% for the market ¹². The looser 90%+ band catches more markets and is used in some STR analyses ¹. Either way, the defining feature is not the event; it is the supply-side ceiling the event forces the market into.

The second thing to get right: compression events are calendar-anchored and repeat year-over-year. If the event does not have a verifiable 3+ year cadence on the organizer's own site, it is a one-off spike, not a compression event, and any market-research tool that lists it as recurring is over-claiming. Super Bowl LVIII rotates hosts, so it is a spike for Las Vegas in Feb 2024 and for New Orleans in Feb 2025; it is not a recurring Las Vegas compression event. CES in Las Vegas, SXSW in Austin, Oktoberfest in Munich, the U.S. Grand Prix at Circuit of The Americas: those are compression events because the city-event pair repeats on the same week, every year.

Third: "it's an event" is not enough signal to classify compression level. The same conference that sells out a secondary market (high compression) is a rounding error in Las Vegas or Orlando (low compression). Tier matters. Market size matters. Room supply matters. The section below is how to classify without PMS data.

Why compression events move markets differently than ordinary demand spikes

STR has the cleanest before/after data on Las Vegas. Super Bowl LVIII (February 2024) posted weekend occupancy of 83.7% (modest by Vegas standards) but Friday-Saturday ADR hit $747, a 227.1% year-over-year jump that drove RevPAR +239.6% ³⁴. Miami's prior Super Bowl ADR record was $563 (2020); Las Vegas beat it by 33% ³. One weekend in one city pulled U.S. hotel-industry RevPAR up 3.9% that week. Strip Las Vegas out and the remaining 24 tracked markets were down 3% RevPAR ³.

That asymmetry is what a true high-compression event looks like. Occupancy can peak at 85% in a glutted market. What breaks under compression is the rate ceiling: the market-clearing price detaches from baseline for 3-5 nights.

Second lens: the before/after tail. Las Vegas posted +129% RevPAR for the week of January 7-13, 2024 on the CES calendar shift, then crashed -19.2% RevPAR the following week as business travel emptied the market ⁵⁶. That is the compression fingerprint: a spike surrounded by a valley.

Contrast with decay. Coachella 2024 accommodation occupancy fell to a 58% two-weekend average (from 69%), ADR to $493 (from $543), RevPAR -23% on Weekend 1 ⁷. Ticket sales had softened 14-17% YoY ⁷. Coachella is still a demand event in the Coachella Valley, but in 2024 it stopped meeting the compression definition. A recurring event does not guarantee recurring compression; one of the few checks a pipeline at this evidence ceiling can do from grounded search alone is re-read prior-year press every year and notice when the "sellout" language disappears.

How to classify compression level without PMS data

The market-research pipeline cannot see your ADR lift last year. It cannot see your booking pace. It can see grounded-search signals: organizer announcements, prior-year press coverage, venue capacity, surrounding room supply. That is enough to sort events into three tiers.

High compression

City-wide sellouts. Occupancy crosses 90%+ market-wide on the event nights, ADR doubles or more against baseline, and the booking window compresses to 60+ days out for anyone who waited.

• Super Bowl in host city. Las Vegas 2024: $747 Fri-Sat ADR, +227% YoY ³. MGM Resorts specifically achieved ADR near $1,000 ⁴.
• Formula 1 races. Las Vegas F1 debut (Nov 2023): Strip ADR hit $629 Thu-Sat, +134% YoY ⁸. Austin COTA U.S. Grand Prix: peak downtown ADR $877 with 88% occupancy at upper-tier hotels ⁹.
• Flagship tech shows on recurring dates. CES in Las Vegas (early January): +129% RevPAR for the event week ⁵.
• Religious pilgrimages on the Islamic calendar. Hajj draws ~1.8 million pilgrims to Mecca annually, managed through country-specific government quotas; 88% of arrivals stay in hotels ¹⁰. When quotas bind, this is structural compression, not market compression.
• Mardi Gras in New Orleans. Peak hotel occupancy ~90-95% during the Carnival weekend, with ADR +$100-$200 above the yearly baseline; Mardi Gras 2026 posted ADR of $225.77 (+14.4% YoY) and RevPAR of $177.42 (+31.4% YoY) ¹¹.
• Oktoberfest in Munich. Peak occupancy 92%, average advertised rate €415, roughly 153% above the annual baseline and 20% above the prior Oktoberfest ¹².

Signal set a synthesizer can read from grounded search: (1) prior-year press coverage using language like "sellout," "no walk-in availability," "record ADR"; (2) the event's own site listing official hotel partner programs (implies supply is block-booked); (3) documented booking windows stretching 60+ days ahead of event nights ¹³.

Medium compression

Clear market lift, but walk-in availability still possible by the tail of the event.

• Major recurring conferences in tier-1 cities. Dreamforce in San Francisco moves the market visibly (one week posted RevPAR +71.6% ¹⁴) but San Francisco has ~34,000 rooms in core submarkets, so even a 60,000-person conference does not fully saturate the compset.
• Second-weekend festivals, post-decay. Coachella 2024 weekend 2: ~55% occupancy, $463 ADR ⁷, still meaningfully elevated vs. April baseline in the Coachella Valley but not compression by STR's 90%+ definition.
• Large trade shows in purpose-built convention cities. JCK Las Vegas, NAB, HIMSS: they move the needle but rarely saturate Las Vegas or Orlando.

Signal set: (1) event appears repeatedly in STR or CoStar weekly recaps but is not the sole driver of the weekly number; (2) organizer publishes attendance figures in the 10k-50k range for a market with 30k+ rooms; (3) host-partner hotel programs exist but are optional rather than block-booked.

Low compression

Detectable lift, no supply squeeze.

• Small regional conferences in tier-1 cities. A 3k-person professional conference in New York. It moves the booking pace at nearby properties; it does not move the city.
• Season-long events. Broadway premieres, museum exhibitions, restaurant week. Elevated baseline demand, no specific compression night.
• Sporting events that draw day-trippers. College football home games in cities where visiting fans drive rather than fly.

Signal set: (1) no single-week ADR or occupancy spike visible in public STR weekly insights; (2) event attendance is well below the available room stock in the host market; (3) booking-window data shows no pre-event lead-time compression.

Where the market-research pipeline's evidence ceiling actually sits

The pipeline operates under the llm-synthesis-over-places-reviews-grounded-search evidence ceiling. That phrase is doing work; it is the set of claims the pipeline is allowed to make.

What the pipeline can do from grounded search alone:

• Name recurring events verified on organizer sites (not aggregator-only listings).
• Classify compression level based on prior-year press coverage describing sellouts, record ADRs, or walk-in-impossible conditions.
• Detect event-pairs (two events landing in the same week in the same market) from calendar overlap.
• Identify market-saturation cues from Places API competitor density around the host venue.
• Surface review-text signals ("couldn't find a room," "stayed way out of town," "prices tripled") as confirmation of prior-year compression.

What the pipeline will not do:

• Predict your property's ADR uplift. That requires PMS data, which sits above the synthesis ceiling.
• Predict your booking-window shift. Requires booking history.
• Predict cancellation or wash patterns. Requires OTA-internal data.
• Predict compset share during the event. Requires rate-shopping data.
• Forecast lift for events with less than 3 years of cadence. Single-instance events (Super Bowl rotation, one-off political summits, Olympic cities) are named as spikes, not as recurring compression drivers.

That is the honest line. The research in Kalibri Labs' profit-driven commercial strategy product and comparable platforms layers transaction-level PMS data on top of market signals to do exactly those property-specific forecasts ¹⁵. The market-research pipeline here does not pretend to replace them. It names the events; your revenue system prices them.

Using the compression_events list in a market-research report

The synthesizer emits 0-10 events per market scan, each with:

• event_name: canonical name as it appears on the organizer site (not on an aggregator).
• month: name or narrow range. Some events slide within a month year to year (CES first full week of January, Mardi Gras tied to Easter).
• description: one line on why this event compresses this market. Not generic; market-specific.
• compression_level: high / medium / low, derived from the three-tier framework above.

Concrete operator actions (paste into your ops doc):

1. Lock rates 45-60 days out on high-compression weeks. Booking windows for compression events run 60+ days pre-event in saturated markets ¹³. If you have not set your rate before then, the market has already set it for you. Revenue-management 101, but consistently missed at independents that run weekly rate reviews instead of event-anchored reviews.
2. Open a 2-night minimum on Friday-Saturday of high-compression weekends. Prevents single-night pickups that leave a Sunday hole. For multi-day events with identifiable arrival patterns (F1 weekends: Thursday practice / Friday-Saturday peak / Sunday departure), 3-4 night minimums are standard; Montreal Grand Prix operators use 4 nights ¹⁶.
3. Use closed-to-arrival rather than rate hikes late in the window. By 30 days out the dynamic is no longer "raise rates"; it is "shape the length-of-stay mix to prevent a shoulder-night hole" ¹⁶. Closed-to-arrival on the peak night forces arrivals onto the shoulder, lengthening stays.
4. Watch for event-pair compounding. Two medium-compression events in the same week frequently produce high-compression conditions. The synthesizer flags these via event_driven_insights. The reverse is also true: adjacent events can cannibalize (the Las Vegas F1 debut pulled demand off the Austin U.S. Grand Prix that same season) ⁸⁹.
5. Re-audit classification annually. Coachella 2024 is the reference case: a reliably high-compression event downgraded to medium because ticket sales softened ⁷. Past-year sellouts are not a permanent classification.
6. Do not use the compression list as a pricing forecast. It names which weeks will compress the market. Your PMS is what tells you what your rate should actually be.

Common failure modes

Classifying a spike as a compression event. Super Bowl in a one-time host city, Olympics, papal visits. These move the market violently for a single year but do not recur. A rate strategy built around them is built on sand.

Ignoring decay. An event that compressed the market three years running can soften. Coachella 2024 is the canonical example ⁷. A market-research pipeline that refreshes annually catches this; one that caches last year's classification does not.

Treating every conference as compression-worthy. A 3,000-person conference in a 40,000-room market is a rounding error. Room-to-attendee ratio is the sanity check; pull it before you classify.

Over-weighting events the OTA calendar promotes. OTAs surface events aggressively because it drives their own booking pace. That does not mean the event compresses your market. The organizer's site and prior-year STR coverage are the signals that matter; aggregator listings are downstream noise.

Confusing seasonality with compression. Broadway season, beach season, ski season: these are demand patterns, not compression events. They move your baseline; they do not cause market-wide sellouts on specific nights.

Self-audit checklist

• Each event on my compression list has a verifiable 3+ year cadence on the organizer's own site
• Each event is classified high / medium / low using prior-year press coverage (not just attendance figures)
• I have refreshed the classification within the last 12 months (decay check)
• I have a rate-lock milestone 45-60 days pre-event on every high event
• I have a 2-night minimum (or event-appropriate LOS) on the peak nights of every high event
• I have a closed-to-arrival plan for the peak night of every high event I run on short booking windows
• I have checked the week for event-pair compounding (two medium events = one high)
• I am not using the compression list as a pricing forecast for my specific property

Evidence ceiling

Backs claims produced under the llm-synthesis-over-places-reviews-grounded-search evidence ceiling. The compression events surfaced in a market-research report are calendar-grounded, verifiable on organizer sites, and classified from prior-year press coverage; they are not forecasts of your property's specific occupancy or ADR lift. For property-specific forecasting, layer a transaction-data system (Kalibri, Lighthouse, Amadeus Demand360, or comparable) on top of the event list. This article does not support claims about your ADR uplift, your booking-window behavior, your cancellation rate, or your compset share during the event.

How OTALift surfaces this

The market-research report emits seasonality.compression_events as a list of up to 10 recurring events with compression_level enum. Each event carries a one-line, market-specific rationale for why this event compresses this market, not a generic description. The Recurring Insight Routine verifies the pipeline does not over-claim: events must have 3+ year cadence, press coverage must describe compression-level conditions, and the classification must be refreshable from grounded search alone. Properties that want property-specific forecasting beyond the list are directed toward their revenue-management system.

Related articles

• How to Identify Your Anchor ICP. Compression events disproportionately attract specific ICPs (sports weekends, tech conferences, music festivals pull different profiles). Your anchor ICP tells you which compression events to over-index on.
• Linguistic Segmentation via Review Keywords. Prior-year review text is one of the pipeline's strongest grounded-search signals for confirming compression ("couldn't find a room," "prices tripled").
• Dynamic Pricing for Independent Hotels. How compression-event rate-locks interact with your baseline dynamic-pricing cadence.
• Pillar: The Hotel Revenue Flywheel. Compression events sit upstream of rate parity and direct-booking capture; the events fill the market, parity and direct strategy determine who keeps the margin.

Sources and methodology

---

Authored by Anya Cortez · Reviewed by Anya Cortez · Last reviewed: 2026-04-21

Anya Cortez is OTALift's hospitality researcher and writes The Labs. Every article is tied to a specific claim the OTALift reporting pipeline makes and is refreshed when the evidence underneath it moves.

Footnotes

1. CoStar / STR, "3 Trends in Hotel Compression Nights and What They Mean." Reports the 90% / 95% compression-night thresholds and notes that ADR premiums on compression nights have been rising in all six U.S. hotel classes since 2010. https://www.costar.com/article/1715670961 ↩ ↩²

2. CoStar / STR, "A Look at Hotel, Short-term Rental Compression Nights." Defines compression nights at the 95%+ occupancy threshold and compares hotel vs. short-term-rental price premiums (~39% for hotels in top 25 markets vs. ~15% for STRs on the same nights). https://www.costar.com/article/1290193533/a-look-at-hotel-short-term-rental-compression-nights ↩

3. STR, "STR Weekly Insights: 4-10 February 2024." Source for Las Vegas Super Bowl LVIII weekend: 83.7% occupancy, Fri-Sat ADR $747, ADR +227.1% YoY, RevPAR +239.6% YoY. https://str.com/data-insights-blog/str-weekly-insights-4-10-february-2024 ↩ ↩² ↩³ ↩⁴

4. Hotel Dive, "Las Vegas hotels see record-breaking ADR from Super Bowl LVIII: CoStar." Reports that actual Las Vegas Super Bowl ADR exceeded STR's pre-event forecast by 30%, with MGM Resorts reaching ~$1,000 ADR. https://www.hoteldive.com/news/las-vegas-hotels-record-adr-super-bowl/707937/ ↩ ↩²

5. STR, "STR Weekly Insights: 7-13 January 2024." Las Vegas RevPAR +129% and national-high 79.8% occupancy for CES 2024 week. https://str.com/data-insights-blog/str-weekly-insights-7-13-january-2024 ↩ ↩²

6. STR, "STR Weekly Insights: 14-20 January 2024." Post-CES Las Vegas RevPAR -19.2%, occupancy -9.9 pp, ADR -6.8%, the "valley after the spike" pattern. https://www.hotelnewsresource.com/article129964.html ↩

7. PriceLabs, "Coachella vs. Stagecoach Impact on Palm Springs Hospitality (2024)." Coachella Weekend 1 2024 occupancy 61% (vs. 72% in 2023), ADR $523 (vs. $580), RevPAR -23% YoY. Weekend 2: 55% / $463. Ticket-sale softening of 14-17% YoY. https://hello.pricelabs.co/analyzing-coachella-stagecoach-2024/ ↩ ↩² ↩³ ↩⁴ ↩⁵

8. CoStar, "Las Vegas' F1 Debut Accelerates Hotel Performance on the Strip." Strip Thu-Sat ADR $629, +134% YoY for the Formula 1 Las Vegas Grand Prix debut weekend. https://www.costar.com/article/216221282/las-vegas-f1-debut-accelerates-hotel-performance-on-the-strip ↩ ↩²

9. CoStar, "Austin's Formula 1 Grand Prix: Hotels Underperform Amid High Fashion and Celebrity Attendance." Austin COTA U.S. Grand Prix: peak Saturday downtown upper-tier ADR $877 (vs. $913 prior year), 88% average downtown upper-tier occupancy (vs. 94%), with supply growth softening the compression year over year. https://www.costar.com/article/751467996/austins-formula-1-grand-prix-hotels-underperform-amid-high-fashion-and-celebrity-attendance ↩ ↩²

10. Future Market Insights, "Saudi Arabia Hajj Tourism Industry: Global Market Analysis Report." Documents the ~1.8-2.0 million-pilgrim annual Hajj attendance, country-level quota system, 88% hotel-vs-apartment share. https://www.futuremarketinsights.com/reports/hajj-tourism-in-saudi-arabia ↩

11. Hotel Management, "Mardi Gras reigns over New Orleans hotel occupancy" and Travel And Tour World, "New Orleans and Las Vegas Are Saving America's Hotel Industry in 2026." Source for Mardi Gras peak 90-95% occupancy, ADR +$100-$200 above annual baseline, and 2026 ADR $225.77 (+14.4% YoY), RevPAR $177.42 (+31.4% YoY). https://www.hotelmanagement.net/operate/mardi-gras-reigns-over-new-orleans-hotels-occupancy ↩

12. Lighthouse, "Are German hoteliers capitalizing on the surge in demand for Oktoberfest 2025?" Munich peak occupancy 92%, average advertised rate €415, ~153% above the annual baseline and ~20% above the prior Oktoberfest. https://www.mylighthouse.com/resources/blog/oktoberfest-hotel-demand ↩

13. STR, "Window of Opportunity: Analyzing Shifts in Booking Trends." Documents 55-day average booking window post-pandemic (down from 70-75 days pre-pandemic) and event-driven lead-time patterns. https://str.com/data-insights-blog/window-opportunity-analyzing-shifts-booking-trends ↩ ↩²

14. CRE Daily via CoStar reporting, "SF Hotel Bookings Up 70% on Convention Surge." San Francisco hotels hosting Dreamforce saw RevPAR +71.6% for the conference week. https://www.credaily.com/briefs/sf-hotel-bookings-up-70-on-convention-surge/ ↩

15. Kalibri Labs, "Hummingbird Commercial | Hotel Commercial Strategy & ProfitMix Insights." Representative of transaction-data-driven compression forecasting that layers on top of the event-naming tier. https://www.kalibrilabs.com/profit-driven-commercial-strategy ↩

16. Hotel News Resource, "Revenue Management Strategies for High-Compression Events." Source for event-anchored LOS, closed-to-arrival, and rate-lock sequencing; includes the 4-night Montreal Grand Prix minimum and 90-day / 50%-inventory LOS re-evaluation rule. https://www.hotelnewsresource.com/article140268.html ↩ ↩²