Compset and market positioning — building the rate bridge

Tuesday morning. Adam stops at Daniel’s desk, phone in hand: “Look, the Danubea has taken its September weekdays down to 89 euros. We’re at 105. This can’t stand — let’s go down to around 90 too, or everyone will book with them.”

Daniel glances at the pace report: the September weekdays are on the curve, pickup is healthy. “Adam, the Danubea cut its rate. But why should anything change here because of that? Our guests are coming, the tempo is fine. If we cut the rate now, the only certain effect is that we sell the same room cheaper — and the Danubea may well raise back in two days.”

This scene plays out weekly in every hotel, and the wrong answer — reflexive rate-following — is one of the industry’s biggest value destroyers. This lesson is about replacing the reflex with a system: how to calibrate the compset (building on lesson 14’s foundations), how to read the index metrics (MPI, ARI, RGI), and how to build a rate bridge — a pre-agreed rate-position ruleset against which a competitor move is a signal, not a command.

The compset — who do we measure against?

The compset (competitive set) is the circle of hotels we measure our own performance against, because in the guest’s mind we are each other’s alternatives; in practice the measurement circle is 4-6 hotels. In lesson 14 we built the selection logic in detail — as a quick recap, the criteria:

• Location: the same demand pool (city centre, the catchment of the same attractions/offices).
• Category and product level: similar star rating, condition, service scope.
• Size: broadly similar capacity (a 20-room boutique and a 300-room conference house live in different markets).
• Guest mix: similar leisure/business ratio and source markets.
• Rate band: overlapping price range — the guest actually compares the rates.

From lesson 14’s three-tier model (primary / secondary / aspirational), this lesson works with the primary circle: the index calculations and the rate bridge are built on it. The secondary is the “watched circle” — for example a hotel positioned above us that we aspire toward, or a new entrant still finding its place.

Hotel Peaqplus City’s primary compset is the four (fictionally named) hotels familiar from lesson 14 — with one important development: Hotel Danubea repositioned itself upward after its renovation and now consciously prices above us:

• Hotel Danubea — since the renovation the strongest product in the circle, consciously priced above us.
• Hotel Aurelia — the closest product analogue: in size, mix and condition.
• Hotel Citadel — a bigger house, simpler product, leisure-strong, typically priced below us.
• Hotel Belveden — a 70-room boutique, business-strong; on weekdays its public rate is often driven by its own corporate compression and restrictions.

The compset mistakes

The most common mistake is the aspirational compset: the hotel measures itself against where it wants to be, not where it is. If the compset is all stronger products, every index paints the hotel as weak, and the organisation either gets demoralised or — worse — underprices itself to “stay competitive” with a circle it does not actually compete with. The reverse distorts too: against a weak compset every index looks pretty, and the hotel gets complacent. The compset should be reviewed annually (renovations, rebrandings, new entrants — immediately after a major market event), but not constantly reshuffled: you cannot measure anything against a moving yardstick.

The index metrics: MPI, ARI, RGI

The three standard market-position metrics compare our own performance to the compset average. All three are read around 100: 100 means we perform exactly at the compset average — the market fair share.

MPI (Market Penetration Index) = own occupancy / compset occupancy × 100 ARI (Average Rate Index) = own ADR / compset ADR × 100 RGI (RevPAR Generation Index) = own RevPAR / compset RevPAR × 100

And the three are linked: RGI = MPI × ARI / 100 — the RevPAR position is the product of the volume position and the rate position, just as RevPAR itself is the product of occupancy and ADR.

Hotel Peaqplus City’s September indexes

The September benchmark data:

Metric	Hotel Peaqplus City	Compset average	Index
Occupancy	78%	74%	MPI = 78 / 74 × 100 = 105.4
ADR	102 EUR	108 EUR	ARI = 102 / 108 × 100 = 94.4
RevPAR	79.56 EUR	79.92 EUR	RGI = 79.56 / 79.92 × 100 = 99.5

Check: MPI × ARI / 100 = 105.4 × 94.4 / 100 ≈ 99.5 ✓

What do these three numbers say?

An RGI of 99.5 on its own would suggest: “we’re fine, we’re delivering our fair share”. But the decomposition is telling: MPI 105 + ARI 94 = the hotel sells more rooms than the market, cheaper. This is the classic volume position: we are partly “buying” the occupancy with a rate discount.

And here comes lesson 43’s insight: the same RGI is not the same profit. A volume-heavy mix means more sold rooms — more housekeeping, laundry, breakfasts, i.e. more variable cost — and the lower rate typically arrives through commission-heavier channels. A rate-led structure — MPI 94.4 × ARI 105.4, our exact mirror image — would deliver the same 99.5 RGI at lower cost. So the long-term goal is not simply RGI > 100, but an index pair with a healthy structure.

The four basic positions:

	ARI < 100	ARI > 100
MPI > 100	Volume position: we fill, but below rate. Question: deliberate strategy, or underpricing?	Winning position: we sell more, at higher rates. (Rare and fragile — the compset will respond.)
MPI < 100	Red flag: we sell less, cheaper. A product, visibility or distribution problem — a rate cut is NOT the solution here.	Premium position: pricier, fewer. Question: does the gap (the ARI surplus) cover the volume shortfall? The RGI decides.

Reading STR-style benchmark data

Index calculation needs compset data — internationally this comes from the STR-style (today CoStar) benchmark reports: participating hotels submit their actuals and receive an anonymised, aggregated compset average in return (which is also why a circle needs at least 4 hotels — so nobody can reverse-engineer anyone else’s numbers). What to watch when interpreting:

• Trend, not snapshot. A single month’s RGI is noisy (one big group or one event tips it). The rolling 3-month and the YTD (year-to-date) trend are the real signal — a one-off swing only matters if it repeats.
• Without breakdown it misleads. The monthly average index can hide that we win on weekdays and lose on weekends (or vice versa). Where available, the day-type and segment breakdown is the real working surface.
• The market itself moves. RGI 103 in a collapsing market is a bad year in absolute terms; RGI 97 in a booming market can be a record year. Always put the index next to the absolute RevPAR trend.

Where there is no formal benchmark coverage (the reality in a large part of smaller markets), you work from proxies: official statistical data (guest-night statistics for the area), the compset rates and availability patterns visible on the OTAs, and the informal GM circle — in many cities a monthly data exchange runs between hotels. These are less precise, but sufficient for a directional signal.

The rate bridge — the working tool of positioning

The indexes look backward: what happened. The rate bridge looks forward: where we price ourselves relative to the compset — deliberately, at rule level, before the daily rate decisions are made.

The rate bridge is a simple ruleset: one target distance per competitor between our own BAR and their visible rate, justified by product difference, refined by day type.

Hotel Peaqplus City’s rate bridge:

Competitor	Product relation	Target rate position (own BAR vs. theirs)	Note
Hotel Danubea	Stronger product (freshly renovated)	−8 … −12%	On weekends we can close up to −5% (leisure demand, where our location is stronger)
Hotel Aurelia	Closest analogue — anchor	−3 … +3%	The bridge’s reference point; a persistent gap = investigate
Hotel Citadel	Weaker product	+8 … +15%	If the gap narrows below +8%, our own rate is probably low
Hotel Belveden	Boutique, business-strong	—	Indicative only: often corporate-compressed on weekdays — its public rate then reflects a remnant room (lesson 32)

The bridge’s value lies in discipline, in two directions:

1. It protects downward: if the Danubea dropped to 89, the bridge says: our band against them is −8…−12%, i.e. 78-82 EUR — we only go that deep if our own pace justifies it too. The bridge does not order us to follow; it shows what following would mean.
2. It reminds upward: if the Aurelia stands at 112 and we at 105, we have fallen out of the −3…+3% band on the downside (−6%) — the bridge signals that we have room to raise, which our reflexes (the “fill the house” instinct) would never notice on their own. The ARI of 94.4 is the monthly, aggregate imprint of the same underpricing.

Deliberate positioning vs. reactive rate-matching

And here the lesson comes together: we watch the competitor’s rate, but we react to our own pace. The compset rate is one input among many — the decision turns on our own demand data. In lesson 32 we saw when an undercut may be followed; the rate bridge lifts the same logic to rule level. The typical situations:

Situation	Reflex response	System response
Compset −10%, own pace on the curve	”Follow them!”	No rate move. Monitor for 72 hours: if pickup slows, reassess.
Compset −10%, own pace −8 pp behind	”Cut below them!”	Adjust to the bottom of the bridge band — but not below it. In parallel: root-cause hunt (what is dragging the market?).
A single competitor −20%, the rest hold	”Price war!”	Ignore. Likely cause: a dump after a group cancellation, or an extranet error. Typically reverts within 48-72 hours.
Compset stable, own pace −10 pp	”Cut the rate!”	This is not a compset problem. Visibility, distribution, event calendar, own content — diagnosis first, then the tool. A rate cut here treats the symptom.
Compset raises, we hold	”Great, we’re cheaper!”	Raise to hold the bridge — otherwise we build ARI erosion and slide into the “cheap hotel” position, which takes years to reprice.

The fifth row of the table is the most insidious: not moving is also a decision. Whoever stands still in a rising market becomes relatively cheaper — losing position while doing nothing.

The price war deserves a separate word: following a rate downward is always easy, raising it back is always hard. Two similar products bidding each other down typically ends with the same demand distributed the same way — at lower rates: market demand does not grow because the downtown hotels got cheaper; everyone just got poorer. The exception is the genuinely price-elastic, market-expanding situation (lesson 36) — but that is a conscious, calculated decision, not a defensive reflex.

Back to Tuesday morning

Daniel’s answer to Adam comes together like this: the Danubea move is a data point, not an event. Our own pace is on the curve (no demand signal), the Danubea’s near −25% cut (from 118 to 89) is isolated — the rest of the compset has not moved, likely a group dump — and by the bridge, following would mean 78-82 EUR in a month we are delivering without it. Decision: no rate move, a 72-hour monitor. Three days later the Danubea’s September rate was back at 118 — they had been flushing out the rooms of a cancelled group block, fast.

What does follow from the benchmark report: the ARI of 94.4 is persistent — the hotel has been pricing below the anchor band against the Aurelia for months. The real to-do is not following the Danubea down, but carefully raising our own BAR into the anchor band, starting with the weekends, watching the pace. The reflex would have recommended a rate cut; the system found a rate increase.

The classic traps

Trap 1: The aspirational compset

We measure ourselves against where we wish to be — every index bleeds, and the organisation either goes numb or panic-prices. The compset is the circle of alternatives living in the guest’s mind, not the owner’s ambition.

Trap 2: Overreacting to the monthly index

Behind one month’s RGI drop there can be a single group, an event shift, or a one-off spike from one compset member. Without trend (3 months, YTD) there is no diagnosis — we do not price on index points.

Trap 3: Index fetishism

The RGI is a relative measure. RGI 102 in a market falling 15% is an absolute revenue decline; and even above RGI 100 you can leave money on the table, if the whole compset underprices a strong demand period. Always put the absolute RevPAR and your own pace next to the index.

Trap 4: The compset rate as the only input

The competitor rate is the most visible data point — but without demand data the decision is blind. Rate shopping (lesson 32) tells you where they stand; pace and the booking curve (lesson 37) tell you how we are doing. A rate decision is only born at the intersection of the two.

Manually vs. Peaqplus

Manually, compset rate-watching means daily OTA browsing (4-6 hotels × several dates × room types — 30-60 minutes a day), and the index calculation requires a benchmark report or the manual stitching-together of the proxy sources — in most hotels what actually gets done is “we glance at Booking”.

Peaqplus’s competitor rate monitoring automates this daily work: the compset’s rates can be followed along the date axis, in a calendar view, so maintaining the rate bridge — the daily check of the distance between our own BAR and the competitor rates — is a matter of minutes, and the outlier moves (like the Danubea’s one-off cut) are visible immediately. Calculating market position (MPI/ARI/RGI) requires benchmark data — the Peaqplus Benchmark module provides it, with monthly MPI/ARI/RGI measured against an anonymous peer pool. For named-compset performance (exactly who is doing how well) there is still no direct data — that is where the proxy methods above do the work.

Key takeaways

• The compset is the 4-6 real alternatives living in the guest’s mind — by location, category, size, mix and rate band, reviewed annually. An aspirational compset ruins every measurement.
• MPI (volume), ARI (rate), RGI (RevPAR) — and RGI = MPI × ARI / 100. The diagnosis is not the RGI number itself but its decomposition: the same 99.5 can be healthy, or a cost-heavy volume position.
• Benchmark data is read in trend and in breakdown — we do not price on a monthly snapshot or an unsegmented average. Always put absolute RevPAR next to the index.
• The rate bridge is the working tool of positioning: a target distance per competitor, justified by product. It disciplines downward (no reflex-following), and reminds upward (it signals the rate-raising room too).
• We watch the competitor’s rate, we react to our own pace. A single outlier competitor move is to be ignored; even a compset-level move only makes sense together with our own demand data. Not moving is also a decision — in a rising market, standing still is losing position.