Selling a quarter nobody wants to travel in
Hostels empty out between July and September. The obvious answer is to discount, which trades margin for a problem it never actually solves. TERM 0 sells a different thing entirely: the two-to-four week gap between a college admission and the first day of class.
- Brief
- Drive occupancy in the lean quarter without discounting
- My role
- Problem structuring, primary research, modelling, deck
- Format
- One-day case competition
- Outcome
- First place
The quarter is not a pricing problem
This was the whole case, and it sits in two measures side by side. Between the peak quarter and the lean one the nightly rate barely moves. Occupancy falls off a cliff.
Nightly rate
Holds. Barely moves at all.
Occupancy
Collapses. Well over half of it gone.
So the revenue collapse is empty beds, not cheap beds, and a discount attacks the one variable that was never the problem. It also gives away margin on the guests who would have booked anyway. Meanwhile the beds themselves are already paid for, so a point of occupancy recovered at the standing rate is close to pure contribution.
That reframes the brief entirely. The question is not what to charge. It is who could possibly want to travel in July, August and September — and whether enough of them exist to move a quarter.
Absolute occupancy, rate and revenue figures come from the goSTOPS competition brief and are withheld here. The relationship between them is the finding; the specific numbers are theirs to publish, not mine.
Seven segments, three gates
Before scoring anything I ran each candidate segment through three yes-or-no questions. A weighted score can be argued with. Already in our base cannot.
| Segment | Locked to the quarter | Not already ours | Big enough | Verdict |
|---|---|---|---|---|
Pre-college transitioners The admission-to-orientation gap is fixed by CUET, JoSAA and NEET dates. goSTOPS has no share of it today. | pass | pass | pass | Finalist |
Corporate campus joiners Joining dates cluster June to August. The long-stay business touches only a sliver of it. | pass | pass | pass | Finalist |
Post-attempt repeaters The result-to-coaching dead window is real and dated, across a few lakh candidates a year. | pass | pass | pass | Finalist |
International backpackers Already in the base. The lean quarter is measured with this segment counted, so it cannot close its own gap. | pass | fail | pass | Excluded |
Himalayan pass travellers Leh already sells out in these months. There is no unmet demand to create. | pass | fail | not run | Excluded |
Monsoon-seeking travellers Self-selecting and small, with no mass channel to reach it. A pilot, not a quarter. | pass | pass | fail | Excluded |
Remote workers No calendar constraint at all. The control behaved as expected, which is the point of running it. | fail | not run | not run | Excluded |
Remote workers were carried through deliberately as a control. They fail the first gate outright, which is what a control is for.
Then I tried to break my own answer
Three finalists went to a weighted score. The obvious objection to any weighted score is that the author chose the weights, so I re-ran it under two other schemes, including one built to favour the criterion my answer is weakest on.
| Weighting scheme | Pre-college | Corporate joiners | Repeaters |
|---|---|---|---|
| Derived, 30/25/20/15/10 | 4.6 | 3.2 | 3.3 |
| Equal weights, 20 each | 4.4 | 3.0 | 3.0 |
| Feasibility first, 10/15/20/25/30 | 4.2 | 2.8 | 2.7 |
Under the scheme least kind to the answer, the gap is still 1.4 points. The conclusion does not rest on the weights, which is the only thing that makes a weighted score worth showing anyone.
Eighty conversations, five decisions
Sixty incoming first-years and twenty parents across nine cities, with twelve longer interviews. Each question was there to settle something specific.
How much free time, and what is planned?
71% report six or more free weeks. 64% had nothing planned.
The window is real, unstructured, and exists only in these months.
What worries you about the first month?
68% coded as not knowing anyone, or walking in alone.
The need is belonging, not boredom. Sell people, not places.
Have you travelled without family?
58% never have. Verified operator mattered to 74%, group visibility to 61%.
The parent is the gatekeeper. Build trust rather than persuasion.
Who pays, and how much?
82% a parent pays. 66% sit in the 5,000 to 8,000 band, median 6,800.
Price at 6,499, inside willingness. Never cut the bed rate.
Where do you hear about things?
79% in the batch WhatsApp or Telegram group. 71% act on a senior's word.
Distribution is peer-led. The batch captain is the channel.
Convenience sample recruited through incoming-batch groups. Directionally indicative, not statistically representative, and the deck said so too.
What it is worth
Sized bottom-up from enrolment data down to a deliberately conservative year-one capture.
Students entering undergraduate education each year
Within the top-500 college catchment
Outstation, over 18, in an incoming-batch group, near a property
Willing and able at 6,499
Year-one capture at 10%
Per student, three nights
The bed is already paid for. Contribution on an otherwise empty bed is close to pure margin, which is why this beats a discount that fills the same bed for less.
₹3.9 Cr
incremental revenue in year one
18,000
bed-nights filled, 7.8 occupancy points
12.4×
return on a ₹20 lakh programme
What I would do differently
The sample is the weakest part. Eighty people recruited through batch groups is a convenience sample, and it over-represents students who are already social enough to join one. That biases the exact behaviour the whole plan depends on. With more than a day I would have sampled outside those groups specifically to test it.
The admission calendar is load-bearing and it moves. In 2024 a CUET delay pushed Delhi University classes to late August, which happened to widen this gap rather than close it. That was luck, not resilience. A real version needs a plan for the year the window narrows instead of widening.
I would also pressure-test the ten percent year-one capture harder. It came from a channel assumption about batch captains that the research supports directionally but has not proven at scale.