RevPARGenius methodology and owner guides
Reviewed by Mark Joefer, Hotel Market Intelligence Researcher · 13 October 2026 · 9 min read
Knowing how to choose a hotel comp set takes about 15 minutes on the OTA. The set you build decides every dollar gap you will read afterwards, so it is worth doing carefully once.
This guide is for GMs and owners of independent properties who want a comp set they can defend. It covers matched product, how many properties to include, what a complete 90-night calendar looks like, and when to refresh the list.
To choose a hotel comp set, shortlist 8 to 12 properties a guest would compare with yours on the OTA, matched on product and location, and drop outliers. A complete calendar has 63 or more of 90 nights priced, and each night needs 15 or more priced before we publish a market middle. We do not publish figures for sets of fewer than 15 properties.
What makes a good hotel comp set?
A comp set is the short list of properties a guest would genuinely weigh against yours for the same stay. Four tests decide whether a property belongs on it:
- Matched product: similar room type, star level, size and facilities. A motel unit and a five-star suite are not competing for the same booking.
- Same guest: the same trip purpose (leisure, business, families) and similar location within the town.
- Bookable on the same basis: the property shows a public rate on the OTA for 1 night, 2 adults, 1 room on most nights, so its rates can be compared on the same search basis as yours.
- Not an outlier: no property that only opens at weekends, sells only packages or sits far outside the range of everything else.
Position in dollars (how far your rate sits from the market middle) only means something with matched comps. Behaviour, meaning how often a property’s rate differs from the night before, does not need matched comps: it shows how actively the hotels around you price, whatever their product.
How do you pick a comp set from OTA rates in 15 minutes?
Set a timer. This is a first pass you can refine later, not a consulting project.
- Minutes 0–3: search your own town. On a major OTA, search a Tuesday about four weeks out, 1 night, 2 adults, 1 room. Filter to your property type and star level. Sort by distance from your property.
- Minutes 3–7: shortlist 12 to 15 candidates. Take the properties a guest would see on the same map view as you, with a similar review score and room type. Skip anything with no rate on that night.
- Minutes 7–10: check a Saturday. Run the same search for the Saturday of that week. Drop any candidate that is missing on both nights or whose cheapest room is a different product (a dorm bed, a three-bedroom villa).
- Minutes 10–13: remove outliers. If one property sits far above or below the rest on both nights, ask whether a guest would really compare it with you. If not, drop it.
- Minutes 13–15: write the list down. Aim for 8 to 12 properties. Note the date you built it and why each one is there.
The cheapest offer on the OTA is a useful common basis, but it hides refund terms, board and room type. Our guide to why the cheapest OTA rate can mislead lists what to check before you trust it.
For each property, note one line on why it is in the set: "same street, same room type" or "where our weekend guests compare us". When you review the set in three months, the reasons tell you which ones still apply.
How many hotels should a comp set have?
Enough that 5 or more are priced on almost every night, because a market middle is only published only when 15 or more are priced. Properties close out, set minimum stays and drop out of the search, so a set of exactly 5 will often fall short. Eight to 12 leaves headroom.
A smaller set can still work for your own use, but every closed-out hotel moves it closer to the 5-priced line. Small sets have a second limit: with only a handful of hotels, a published middle comes close to showing individual hotels' prices. That is why we do not publish figures for sets of fewer than 15 properties, even though we use smaller sets as working tools.
What does a complete comp-set calendar look like?
A complete calendar has 63 or more of the next 90 nights priced. Anything built on fewer is a preview basis and should say so. The nights that tend to go missing are the busiest ones, where hotels close out or add minimum stays, which are exactly the nights you most want to see. Our market middle explainer covers the 5-priced rule in detail.
Once the calendar is complete, start with behaviour. It is the signal a GM can check in their own extranet without arguing about which hotels belong in the set. In the Auckland CBD set (18 hotels, captured 9 October 2026), the median hotel’s rate differs from the night before on 72% of nights, the median hotel’s rates range across 240% of its own minimum, and the median weekend premium is 5.4%. Weekend premium per hotel is its own weekend (Fri–Sat) rate ÷ its own weekday (Sun–Thu) rate − 1, and the figure quoted is the median of those premiums.
Sets differ widely on this measure, so read your own against the hotels around you rather than against a single benchmark. If your own rate differs from the night before on far fewer nights than the set around you, that is the first thing to look at, before any dollar gap.
Dollar position comes second and needs matched comps. Revenue-share indexes such as MPI and RGI depend on the same choice of comps; our guide to MPI and RGI explains how a poorly matched set distorts them.
How often should you refresh a hotel comp set?
Review it every quarter and after any obvious change in the town. Triggers to act on:
- A property in the set renovates, rebrands, changes star level or closes.
- A new property opens within your map view with a similar product.
- One member has no public rate on more than a third of nights for a month. It may have moved to direct-only selling or long-stay contracts.
- Your own product changes, for example a new room type or refurbished rooms.
Keep the old list when you change it. If your gaps shift right after a comp-set change, you want to know whether the market moved or the set did.
What to do this week
- Run the 15-minute process above for one Tuesday and one Saturday about four weeks out.
- Write a one-line reason beside every property you keep.
- Count how many of the set have a public rate on each of your next 10 key nights. If any night falls below 5, add a property.
- Check your own rate in your extranet: on how many of the last 30 nights did it differ from the night before? Compare that with the behaviour of the hotels in your own set.
- Put a quarterly review in the calendar.
Our take
A good comp set is short, matched and written down. Eight to 12 properties gives enough room for 5 to be priced on almost every night.
Lead with behaviour, which needs no argument about the set, then use the matched set for dollar position. In the Auckland CBD set of 18 hotels, the median hotel’s rate differed from the night before on 72% of nights. How often your own rate changes, against the hotels around you, is a useful first question.
Frequently asked questions
How many hotels should be in a comp set?
Eight to 12 is a practical range. We only publish a market middle when at least 15 are priced on a night, and some properties close out or drop from the search on busy nights, so a larger set keeps most nights valid.
What is a complete comp-set calendar?
A calendar where 63 or more of the next 90 nights are priced. A calendar with fewer priced nights is a preview basis and any figures from it should be labelled that way.
Should I include luxury hotels in my comp set?
Only if guests really compare them with your property for the same stay. Properties with a very different product distort dollar comparisons, though they can still be useful as a behaviour reference.
How often should I update my hotel comp set?
Review it every quarter, and immediately when a property in the set renovates, rebrands or closes, a similar new property opens nearby, or your own product changes.
Can I use OTA search results to choose my comp set?
Yes. Search your own town for 1 night, 2 adults, 1 room on a weekday and a weekend night, shortlist properties a guest would see beside yours with a similar product and review score, then remove outliers.
The next step is to put your own dates beside the market middle for the same nights, with the capture date on every number.
Free Hotel Market Intelligence
Build your comp set and see the calendar
Try RevPARGeniusSee all insights
Data sources: RevPARGenius behaviour capture for the Auckland CBD set (18 hotels, captured 9 October 2026, 90-night window); public OTA search basis (1 night, 2 adults, 1 room, cheapest offer; refund and board unknown). Analysis run October 2026. Last reviewed October 2026. RevPARGenius is a hotel market intelligence platform. It is not affiliated with any online travel agency or hotel chain.
Research methodology and disclosures: Market figures in this article are calculated by RevPARGenius from room rates publicly displayed on online travel agency (OTA) channels, on one search basis (1 night, 2 adults, 1 room, cheapest offer shown). We publish a market figure only where at least 15 properties were priced. The figures are advertised asking prices. They do not show occupancy, bookings, negotiated corporate rates or revenue achieved. Worked examples marked as synthetic use invented data. This article is general information, not pricing advice: each hotel sets its own rates independently. RevPARGenius and CloseMode are commercially affiliated platforms.
