ESEN← Back to projects
Data · Case study

Business analysis and financial evaluation of a lodge in Chiloé

By Guille Ferveg· June 2026· Data analysis · Python

I took scattered data from a real seaside cabin business and turned it into decisions: I consolidated bookings from three sales channels and 94 reviews to measure seasonality, channel profitability and reputation, then evaluated the financial viability of a direct-channel digital strategy.

Client
Lodge & cabins · Ancud, Chiloé
Period
Jul 2021 – May 2026
My role
Data analysis and financial evaluation
Deliverable
11-section report + KPIs

01 The challenge

The business was growing strongly, but its information was scattered across three different platforms (Airbnb, Booking and direct bookings), plus disconnected reviews on Google and Tripadvisor. There was no unified view capable of answering financial questions: Which channel performs best? How dependent is the business on high season? Is digital marketing worth the investment? My job was to turn that disorder into one clean dataset and produce actionable conclusions.

02 Data and methodology

I consolidated 133 bookings at transaction level (Airbnb, Booking and direct) plus 94 reviews—all converted into comparable gross amounts. Before analyzing them, I had to clean the data: I corrected source errors such as amounts with an extra zero, incorrectly entered dates and inconsistent calculations—a step that often determines whether an analysis informs or misleads.

Using that foundation, I defined standard hospitality and financial indicators so the conclusions would be comparable and defensible:

Key criterion: I explicitly stated that “net revenue” means after commission, not final profit, since operating costs such as cleaning, heating and breakfast are still missing. Clearly stating the limitations of the data is part of an honest analysis.

03 The numbers that matter

A snapshot of the business:

CLP $29.6Mrealized gross revenue (CLP)
×8revenue growth over three years
+19%recent year-over-year growth
4.67/5reputation · 94 reviews · 86% rated 5★
125completed bookings · 287 nights
≈207%estimated ROI of the digital project

Real business figures shared with the client’s authorization. Gross amounts in CLP; net revenue is after commission and before operating costs.

04 Profitability by channel

Airbnb is the volume engine, but volume is not the same as profitability. Once average booking value, commission and cancellations are considered, the priority order changes.

Share of revenue by channel
Where revenue comes from today
  • Airbnb — 79% · CLP $23.3M · 101 bookings, no cancellations
  • Direct — 12% · CLP $3.5M · highest value per booking
  • Booking — 9% · CLP $2.7M · lower booking value and costly commission
The strategic finding: the direct channel generates an average booking value of approximately CLP $394,000 —nearly 2.5× Booking’s value (≈CLP $159,000)—with 0% commission and zero cancellations. It is the channel to develop even though it is currently the smallest. The business-wide average daily rate is about CLP $104,000 and the average booking value is about CLP $234,000.

05 Seasonality: the major risk

The business’s main vulnerability is its dependence on only a few months: 73% of revenue is generated in summer while winter contributes only 4%. January (≈CLP $10.0M) and February (≈CLP $7.7M) account for nearly 60% of annual revenue: cash flow depends on three months.

Revenue share by season
Summer (Dec–Feb) dominates; winter contributes very little
Summer
73%
Shoulder season
23%
Winter
4%

Revenue is also concentrated in the product mix: only two cabins generate nearly 71% of the total. Simultaneous seasonal and product concentration creates a double reason to diversify demand.

Revenue share, top cabins
“Ideal familias” (CLP $12.1M) and “Cancha de tenis” (CLP $8.8M) generate 71%
Ideal for families
41%
Tennis court
30%
Sea view
16%
Other
13%

06 Guest booking behavior

The booking-pattern analysis revealed clear profiles that can be used to refine pricing, availability and communication:

07 A cautious forecast, not guesswork

I tested forecasting models on the monthly series. I rejected an exponential trend model because it produced inflated projections, and selected a seasonal model with damped growth, which was more stable for short-term decisions. The result points to approximately CLP $14.0M over the next 12 months (versus CLP $11.8M in the previous 12), once again heavily concentrated in January (≈CLP $6.5M) and February (≈CLP $3.5M).

Analyst’s criterion: I presented the forecast as an indicative reference for cash-flow and campaign planning, not as a definitive prediction. With a limited monthly history, promising precision would be misleading.

08 Financial evaluation of the digital project

The underlying question was: Is it worth investing in a direct-channel digital strategy? (a proprietary website, a winter campaign and intermediary-free bookings). I evaluated it as an innovation project with an investment, expected benefits and sensitivity analysis.

✔ Technical feasibility: high ✔ Economic feasibility: high ✔ Market feasibility: high

The financial logic is strong: limited investment, permanent commission savings and a 4.67/5 reputation that supports direct acquisition. The estimated ROI is approximately 207% with payback in less than one season.

ROI sensitivity analysis
Three campaign-conversion scenarios—the project remains viable even in the pessimistic case
Pessimistic
+
Base
++
Optimistic
+++

Relative scale: even in the conservative scenario, the investment is recovered and financial risk remains controlled.

09 Recommendations

I concluded the report with an actionable plan, not just a diagnosis:

What this project demonstrates

Beyond this particular business, the case summarizes how I work with data: consolidate messy sources, clean and validate them, choose the right indicators and translate everything into decisions while clearly stating the limitations. The final deliverable was an 11-section report with monitoring KPIs and a measurable implementation plan.

PythonPandasData cleaningFinancial analysisForecastingData vizKPIs

Do you have data you cannot quite organize?

Bookings, sales, surveys or disconnected spreadsheets—I can turn them into decisions. Let’s talk.

Let’s talk →