Case study — Hotel Peaqplus City's tough year

December 31, 4:20 p.m., Hotel Peaqplus City. The last guests are checking in for the New Year’s Eve package; champagne glasses line up at reception. Daniel sits in the office, putting together the year-end report for Adam and the owner. At the top of the page, a single number: annual RevPAR closed at 105.0% of budget — 81.76 EUR against the planned 77.85 (RevPAR — revenue per available room).

From the outside, this is a boringly good year. From the inside, it is not. In April, at the low point, the full-year projection stood below 97% of budget, and the pessimistic scenario showed 94 — and Daniel honestly could not have said where the missing part would come back from. The year was assembled from four entirely different stories: in spring, most of the largest corporate partner’s volume fell away; in summer, record demand very nearly became a trap; in autumn, the hotel’s long-standing autumn MICE base disappeared (MICE — meetings, incentives, conferences, exhibitions); and in Advent, the team translated the year’s accumulated lessons into pricing rules.

This lesson — the closing lesson of the training — tells that year, season by season. You will learn no new tool in it. Instead you will see what no earlier lesson could show on its own: how the toolkit of 66 lessons assembles into one continuous year — where the spring decision log shortens the Advent debate, the summer experiment calibrates the autumn promotion, and the morning pickup routine (lesson 50) holds the whole thing together.

January — the frame everything is measured against

The year begins with the budget. Hotel Peaqplus City has 80 rooms — annual capacity is 80 × 365 = 29,200 room nights. The quarterly plan agreed at the December budget debate (lesson 60 covered the framework for arguing with numbers):

Quarter	Capacity (room nights)	Budget RevPAR	Budget room revenue
Q1 (Jan–Mar, 90 days)	7,200	58.00 EUR	417,600 EUR
Q2 (Apr–Jun, 91 days)	7,280	82.00 EUR	596,960 EUR
Q3 (Jul–Sep, 92 days)	7,360	95.00 EUR	699,200 EUR
Q4 (Oct–Dec, 92 days)	7,360	76.00 EUR	559,360 EUR
Year	29,200	77.85 EUR	2,273,120 EUR

A quick check: 417,600 + 596,960 + 699,200 + 559,360 = 2,273,120 EUR, and 2,273,120 / 29,200 = 77.85 EUR — the frame closes.

The mix behind the plan is the usual one: roughly 65/35 leisure/business, and on the channel side broadly 50% Booking.com, 20% direct, 15% Expedia and other OTAs (OTA — online travel agency), 15% corporate — the team has known the differences in per-channel net values by heart since lesson 43’s net ADR table (ADR — average daily rate). The corporate leg’s annual plan is ~3,500 room nights at an average contract rate of 88 EUR — and a single account, a regional shared-services centre, provides ~40% of that volume, ~1,400 room nights a year. That concentration came up at the budget debate as a risk. In January it was still just a row in a table.

Q1, for the record, went by without incident: 58.50 EUR RevPAR, 100.9% of budget. A routine quarter, routine meetings. The year truly began in mid-March.

Spring — the corporate leg buckles

The signal arrived before the letter

From the last days of February, Daniel’s morning pickup routine (lesson 50’s 15-minute checklist) keeps pointing to the same place: the same-point shortfall of April–May weekdays grows week by week — against last year’s point at the same distance, the midweek dates’ gap of 4-5 percentage points opens first to 6-7 pp, then to 9. The Insight Engine flags outlier deviations on the same dates. Then on March 12, a Monday morning, something appears in the pickup report’s date rows that is no longer a slowdown but an event: −34 room nights of negative pickup on April weekdays, tied to a single corporate account — the big partner has cancelled part of its April block bookings.

Four days later, on Friday, the official letter arrives too: the partner’s parent company has announced a cost freeze and an office consolidation; their second- and third-quarter travel budgets are cut in half — and the letter does not promise it will stop there. So the machine did not know what was happening — but weeks earlier it had shown where to look. By the time the letter came, Daniel was not experiencing surprise, but confirmation.

This is not a fast-moving crisis — it is a regime change

The first and most important diagnostic decision: what are we dealing with? Lesson 62’s fast-moving crisis — when an external shock switches the market off within days — is a different genre: there, the cause of demand disappears temporarily, and you protect your position for the rebound. Here, a structural piece of demand disappears for good: the partner is not postponing the trips; it is eliminating them. This is the regime change we met in lesson 51 — the moment when the historical patterns that every forecast is built on stop being valid.

From this follows the first technical task, before any strategic debate: a manual override of the Smart Forecast across the affected date range — the Layer 3 correction, whose purpose is exactly this (lessons 38 and 55). The machine layers learned from last year’s corporate presence; if the correction is skipped, the system spends months optimising for a midweek demand that no longer exists, and every piece of downstream logic — rate recommendations, restrictions, reports — is built on phantom data. Daniel corrects the Q2–Q3 weekdays date by date, recording the reasoning.

The live meeting — lesson 47’s frame under fire

Daniel rewrites the Monday revenue meeting’s routine agenda on Sunday evening and asks for preparation — exactly as in lesson 47’s November corporate cancellation wave. The March 19 meeting runs down the five steps:

DATA. The partner’s annual plan was 1,400 room nights; Q1 was still delivered on plan (~350 nights). The halved budget alone would imply a ~700-night year — but the April blocks already cancelled cut deeper than the halving, the office consolidation shrinks the travelling population itself, and the letter promises nothing for Q4. The sales manager’s conservative planning number for the realistic annual total: ~550 nights — so the expected loss is −850 room nights. And there is a second circle: two smaller corporate accounts from the same sector are signalling a similar, milder pullback — a further −250 nights combined. Total corporate exposure for the year: −1,100 room nights × 88 EUR = −96,800 EUR.

DIAGNOSIS. A partner-side, structural cause — it is not unhappy with the hotel, nor with the city. But it is not isolated noise either: a sector-level cost cycle that can spread. The other corporate accounts’ pace is still within the usual band — from now on, the meeting tracks it weekly.

OPTIONS. Replacement can stand on two legs: new corporate acquisition (slow but durable) and a stronger leisure mix on the freed-up midweek capacity (faster, but at a lower rate level). The third option — a house-level rate cut — comes off the table in light of the diagnosis: the lost demand was contracted and never watched the public BAR (BAR — best available rate); a rate cut would cheapen the demand that remains and would not fill the hole.

DECISION + FOLLOW-UP. The sales manager is tasked with acquiring two new, smaller corporate accounts — with the displacement discipline that lessons 40–41 fixed: a group or volume deal is not “needed at any price” but a calculated decision about what it displaces. Esther takes the leisure line: targeted midweek promotions (lesson 46’s closed-group logic — not a public rate cut) and activating the direct list (lesson 49). Every action goes into the Revenue Meeting module with an owner, a deadline and a metric — and this record will pay dividends in the autumn.

Spring’s balance sheet

By year end, the replacement strategy had delivered this — the conservative March estimate proved roughly right:

Item	Room nights	Avg. rate	Value
Corporate loss (main partner −850, two smaller accounts −250)	−1,100	88 EUR	−96,800 EUR
Replacement: two new, smaller corporate accounts	+300	84 EUR	+25,200 EUR
Replacement: stronger leisure mix (targeted promo + direct)	+360	74 EUR	+26,640 EUR
Net effect	−440	—	−44,960 EUR

A quick check: 300 + 360 = 660 nights replaced, 60% of the loss; 25,200 + 26,640 = 51,840 EUR, 53.6% of the lost value; 96,800 − 51,840 = 44,960 EUR.

The table carries two honest lessons. First: the replacement is 60% in nights but only 54% in value — replacing 88 EUR contracted volume from a 74–84 EUR mixed portfolio is only possible at a discount, and that must be said out loud, not smudged over. Second: the hole could not be refilled completely — as lesson 47’s November case also showed: whoever promises 100% replacement is either destroying rate or preparing a disappointment.

Q2 ultimately closed at 77.00 EUR RevPAR against the 82.00 budget — 93.9%. Of the 36,400 EUR quarterly shortfall (7,280 × 5.00), roughly two-thirds is the direct effect of the corporate hole; the rest is the sector’s general caution in midweek business. In April this number looked far more frightening — that is when the year projection hit its low point, and standing up the replacement machinery took two weeks from the March 12 signal. What remained of the quarter: a working replacement machine, a corrected forecast, and a decision log in which every move is recorded together with its reasoning.

Summer — when success is the trap

Record pickup, and what it hides

By late June the picture flips. The city’s summer event season — a festival and concert series running across several weekends — is stronger than usual, and behind it the region’s leisure demand is exceptional too. Daily pickup is regularly +40–50 room nights, and the July–August weekends’ same-point lead grows week by week. The Insight Engine now signals in the opposite direction from spring: it keeps flagging “unusually fast” days.

And here is the summer’s real risk, easy to miss after a defensive spring: the trap now is not the shortfall — it is the success. In lesson 42 we saw, on a single concert Saturday, what early sell-out means: the house fills weeks before the event at the normal rate, and the highest-paying, late-arriving demand no longer fits — the full house actually left money on the table. This summer the same phenomenon threatens not on one Saturday but house-wide, across a whole run of weekends: several peak Saturdays stand above 85% at T-30.

The rate-shopping picture is unambiguous: the compset (competitive set — lesson 44) is raising, by double-digit percentages on some weekends. If the market raises 12% and we raise 4, the ARI (average rate index — our own average rate against the compset average) deteriorates while occupancy looks “pretty” — this is the silent loss of underpricing: it shows up as a loss in no report at all, because revenue is coming in — just less of it than it could be.

The bold path — with discipline

The Pricing Engine’s recommendations (lesson 56) have been pointing upward on the affected dates for weeks. And here comes the test of override discipline — with the opposite sign from spring. In spring, Daniel rightly overrode the machine downward: he knew what the model did not — that the corporate base had fallen out structurally. In summer the temptation runs the other way: “let’s not raise this much, we’ll scare the guests away.” But that is not information; it is a feeling — the machine is now seeing exactly what is happening: exceptional demand. The rule lesson 56 fixed is the same in both directions: override only with a stated reason, in writing — and if the reason is just a bad feeling, it is not a reason.

The July meeting’s decision has three elements:

• A bold rate-increase path: the peak weekends’ BAR steps brought forward and raised, the upper rate steps genuinely armed — not stage décor.
• LOS rules (LOS — length of stay): MLOS-2 (minimum length of stay) on the strongest Saturdays, so one-night peak demand does not stuff the house ahead of the two- and three-night bookings that are worth more in total.
• A controlled check with lesson 61’s method: on two July weekend pairs with similar profiles, one gets the bold rate ladder, the other the moderate one. The result, three weeks later: the bold path’s pickup pace did not meaningfully slow, and the weekend ADR difference is +9 EUR — the demand was there in the upper band too. From August the bold path runs house-wide, and when Adam asks (“are we sure we haven’t overshot?”), the answer is not an opinion but the experiment’s control data.

Summer’s balance sheet

Q3 metric	Last year	This year	Change
ADR	110.00 EUR	122.05 EUR	+11.0%
Occupancy	89%	88%	−1 pp
RevPAR	97.90 EUR	107.40 EUR	+9.7%

A quick check: 122.05 × 0.88 = 107.40 EUR; last year 110.00 × 0.89 = 97.90 EUR.

This is the imprint of the classic “good” rate increase: +11% ADR while occupancy gave back just 1 percentage point. Q3 RevPAR closed at 113.1% of budget (107.40 / 95.00), a surplus of +91,264 EUR against the plan (7,360 × 12.40) — two and a half times Q2’s 36,400 EUR shortfall. By the end of summer the year projection had crept back above budget. But at the September meeting Daniel does not celebrate; he records: what made the bold path possible? Not courage — the frame. The rate-bridge band, the trigger-based steps, the override log and the control experiment. Being bold without discipline is gambling; this was not that.

Autumn — the conference that is never coming back

The news

On September 6, a Thursday, the sales manager walks into Daniel’s office — this time not with an enquiry but with news: the city’s big autumn conference, which for years has filled the neighbourhood’s hotels every October, is ceasing to exist in its current form — the organiser is merging it into another city’s event. For the hotel this meant a fixed block every year: 30 rooms for 3 nights, with a group dinner and room hire on top.

The exposure has to be sized with lesson 58’s TRevPAR mindset (TRevPAR — total revenue per available room: total revenue, not just rooms, per available room): the conference brought more than room nights.

Lost item	Quantity	Value
Room block (30 rooms × 3 nights)	90 room nights × 92 EUR	8,280 EUR
Related F&B + event revenue (F&B — food & beverage)	90 nights × 38 EUR	3,420 EUR
Total exposure	—	11,700 EUR

This is again a structural loss, like in spring — an annually recurring base has disappeared, not a booking. The difference: this time there is a broken-in playbook. What the team improvised in March, it executes in September.

The four steps

First: a tentative review. With lesson 63’s method, the sales manager and Daniel walk through the pending group pipeline: what sits in tentative status across the autumn weeks, with what probability, and what would move it. Looking back with the Time Machine also shows how the same period built last year — how much came from outside the conference. A long-pending corporate training group (12 rooms × 2 nights) would fall exactly on the vacated week: sales sends a refreshed offer with a deadline — the displacement calculation (lesson 40) is now trivial, since the block’s place stands empty. Two weeks later the group confirms at 85 EUR.

Second: hunting replacement MICE. Every autumn–winter MICE enquiry in the Sales module’s pipeline gets prioritised; the goal is not to plug this year’s hole at any price but to rebuild next year’s base — because the conference will be missing next year too. This is the spring lesson repeated at scale: for a structural loss, tactical replacement only treats the symptom — the real answer is base-building.

Third: working the vacated week day by day. Lesson 48’s principle — every date is its own case — becomes tangible here: Daniel goes through the vacated week day by day in the DCAL. Tuesday–Wednesday belong to the training group; Thursday has room for corporate steering (toward the accounts acquired in spring); the weekend is marked for a leisure promo; and Monday is deliberately left alone — normal pickup will most likely do the work there by itself. Four different days, four different answers — instead of the “let’s do something about the whole week” reflex.

Fourth: a targeted promo for the remaining hole. Esther builds a closed-group offer within lesson 46’s frame, aimed at the direct list and returning leisure guests — not a public rate cut that would cheapen a weekend already building on its own. The details are aligned with sales and Daniel in the Discussion Thread behind the report rows (lesson 65) — by the time the decision reaches the meeting, the threads are already combed together.

Autumn’s balance sheet

Item	Room nights	Value
Lost conference block (rooms + F&B + event)	−90	−11,700 EUR
Training group (12 rooms × 2 nights × 85 EUR)	+24	+2,040 EUR
Targeted leisure promo (26 nights × 81 EUR)	+26	+2,106 EUR
Ancillary spend of the replacement guests (estimate)	—	+1,000 EUR
Recovered value	+50	+5,146 EUR

A quick check: 24 + 26 = 50 nights, 55.6% of the 90-night hole; 2,040 + 2,106 + 1,000 = 5,146 EUR, 44.0% of the 11,700 EUR exposure.

Again the double reading, and again it must be said out loud: the replacement is 55% in nights, 44% in value — tactical tools cannot fully replace group rate plus event revenue. The difference from spring is not in the ratio but in the speed: in March, two weeks passed from the first signal to a working replacement machine — in September, from the Thursday news to the finished action list fixed at the Monday meeting, four days. The playbook does not eliminate the loss — it eliminates the hesitation.

Advent — institutionalising the lessons

The pattern the decision log reveals

In early November, at the quarterly review (run on lesson 64’s Decisions and Revenue Track logic), Daniel does not open with the usual “what happened” round but with a pattern he has mined from the decision log. Reading across the year’s recorded pricing decisions, the same delay repeats: our increases are consistently late. In spring, the leisure promo went live two weeks after the ideal launch; in summer, the first bold rate step moved ten days after the pickup signal was already there; in autumn the reaction shortened to four days — but the pattern stands: between recognition and executed decision, 8-10 days pass on average, and that slippage cost money every single time.

This is the kind of lesson that cannot be formulated without a decision log — from memory, everyone would tell their own version. In writing, with dates and outcomes, it is not an opinion but a pattern. And December — the year’s last big demand period — is the terrain where it can be fixed immediately.

Three decisions for Advent

First: recutting the December rate steps. The Advent weekends’ increase points move 2-3 weeks earlier than last year — not because demand is stronger this year, but because according to last year’s log the demand was there last year too; only the increase arrived late, and in the final weeks most of the rooms sold at the lower rate that had already settled in.

Second: a closed package for the weak weekdays. For the early-December weekday trough, Esther builds a closed-group packaged offer (room + an Advent programme element) on lesson 46’s closed-group logic: it does not touch the public rate and targets an entirely different buyer set. The real novelty, though, is not the package but the length of the debate. Last year the Advent pricing discussion with Adam dragged on for two weeks — this year it closes in twenty minutes. Not because Adam has become more lenient, but because behind lesson 60’s argumentation frame now stand two things that did not exist last year: the decision log (what we did at this time last year and what came of it — not a memory but a record) and the control data (the summer experiment showed how demand behaves on the bolder path). The debate was not a fight but a joint reading.

Third: disciplined weekend protection. MLOS rules and armed upper steps for the Advent Saturdays — backed by the rate shopper’s market picture, using the trigger logic proven in summer. The peak days now do not “fill up by themselves”; they build according to plan.

Advent’s balance sheet

The 26 days spanning the four Advent weekends (late November to Christmas Eve, 2,080 capacity nights) closed at 89.40 EUR RevPAR against last year’s 82.00 — +9.0% (89.40 / 82.00 = 1.090), which is ~15,400 EUR of surplus against last year’s Advent (2,080 × 7.40 = 15,392). Q4 as a whole: 83.60 EUR RevPAR — 110.0% of the 76.00 budget: October was pulled back near plan by the playbook, and December was carried above it by the Advent rules and a strong New Year’s week.

And more important than the number: Advent was the first period of the year in which no external event helped or hurt. There was no crisis, no record demand, no vanishing conference. Just a team that had translated the year’s lessons into rules — and the rules delivered 9%.

Year-end balance — back to December 31

This is how the number the lesson opened with comes together:

Quarter	Budget RevPAR	Actual RevPAR	Index	Revenue variance
Q1	58.00 EUR	58.50 EUR	100.9%	+3,600 EUR
Q2	82.00 EUR	77.00 EUR	93.9%	−36,400 EUR
Q3	95.00 EUR	107.40 EUR	113.1%	+91,264 EUR
Q4	76.00 EUR	83.60 EUR	110.0%	+55,936 EUR
Year	77.85 EUR	81.76 EUR	105.0%	+114,400 EUR

A quick check — the weighted derivation: actual room revenue = 7,200 × 58.50 + 7,280 × 77.00 + 7,360 × 107.40 + 7,360 × 83.60 = 421,200 + 560,560 + 790,464 + 615,296 = 2,387,520 EUR; that is 105.0% of the 2,273,120 EUR budget, and 81.76 EUR across 29,200 room nights. The variance column sums to the same: 3,600 − 36,400 + 91,264 + 55,936 = 114,400 EUR.

The table’s most important reading is not the total but the asymmetry: the 105% is not “every quarter slightly above budget” — it is the weighted sum of a deep Q2 and an exceptional Q3–Q4. Not a single quarter of the year was “average”. And that is exactly why handling any one quarter well in isolation would not have been enough: the spring damage control limited the loss (−36,400, not close to −60,000), the summer discipline extracted the surplus (+91,264, not +50,000), the autumn playbook reacted fast, and the Advent institutionalisation made the whole thing repeatable.

In the closing report, Daniel condenses what made the 105% possible into four lines — and those four lines are also the training’s four big lessons, projected onto the four episodes:

1. The early signal — the machine’s job. The spring corporate loss was shown by the same-point gap weeks before the official letter, and by the negative pickup days before it; the summer underpricing risk by the Insight Engine’s “unusually fast” flags. The machine does not know what is happening — but it shows early and reliably where to look. Anyone who checks the numbers only monthly would have lived through every episode weeks more expensively.
2. The fast diagnosis — the human’s job. Fast-moving crisis or regime change? Isolated loss or systemic weakening? Distorted base or genuine shortfall? At every turning point of the year, the diagnosis decided which tool to reach for — and no system makes the diagnosis for you.
3. The disciplined frames — the playbooks’ job. The bold summer pricing was not courage but rule-following: the rate-bridge band, trigger steps, the override log, the control experiment. The autumn speed was not improvisation but a rerun of the spring playbook. The frame does not take the decision away — it takes away the scrambling.
4. The recorded decision — the learning’s job. The decision log produced Advent’s pattern recognition (“our increases are consistently late”), and the decision log shortened the Advent debate from two weeks to twenty minutes. Recorded decision + recorded outcome = organisational knowledge; without it, every year starts from scratch.

The end of the training — and the beginning of the work

The first lessons started here: what RevPAR is, and why watching occupancy alone is not enough. Then came the tools: segments, the booking curve, the forecast, displacement, channel economics, the rate bridge. Then the methods: the morning routine, leading the meeting, day-level strategy, promotional discipline. Finally the system level: the division of labour between machine and human, experimentation, crisis management, the life story of decisions.

This year — Hotel Peaqplus City’s tough year — was placed at the end of the training because this is the format reality comes in. Reality does not arrive lesson by lesson. It throws the lost partner, the record summer, the vanishing conference and the owner’s questions at you all at once — and it does not tell you in advance which week brings which. What the 67 lessons can give you is not a recipe for every situation but something better: a way of thinking — position and speed, diagnosis before action, frames against stress, recorded decisions for learning — and a toolkit that turns that way of thinking into a daily 15-minute practice.

Revenue management is not a list of lessons. It is a daily practice. What you have learned here is worth something if — and only if — tomorrow morning, on your own hotel’s numbers, on your own pickup report, you start doing it.

Key takeaways

• The year’s result is decided by how the extremes are handled, not by the average: the 105% year was the weighted sum of a 93.9% Q2 and a 113.1% Q3 — the goal is not evenness, but limiting the bad period’s loss and fully harvesting the good period’s surplus.
• In a structural loss (a regime change), the first step is a manual forecast correction — without it, every downstream system works on phantom demand. And replacement is never 100%: in spring 60% in nights / 54% in value, in autumn 55% / 44% — saying this out loud is honesty; smudging it is self-deception.
• Success is as much a risk as crisis: in strong demand, underpricing is the silent loss — rate-bridge discipline, earlier rate steps, LOS rules and a controlled experiment together delivered +11% ADR at just −1 pp of occupancy.
• The machine signals, the human diagnoses, the frame executes: the early signal (pickup report, same-point, Insight Engine) runs weeks ahead of the official news; the diagnosis (fast-moving crisis or regime change? isolated or systemic?) picks the tool; the playbook shortens the time from diagnosis to execution — from two weeks to four days.
• The recorded decision is the cheapest learning tool: the pattern (“our increases are consistently late”) can only be read out of a decision log, and the log + control data pair turns a two-week debate into a twenty-minute joint reading. That is how the organisation learns — not just the individual.