From Gut Feel to Data: The Quant Revolution in Hotel Operations Decision-Making

---

1. The Story

Mr. Chen runs a 100-room business hotel in Chengdu. 15 years in the industry.

He had a nickname — "Chen the Oracle" — because his instincts were usually right. From experience, he could forecast next month's occupancy. From feel, he could judge channel effectiveness. From intuition, he knew which employee to promote.

But last year, his "instincts started failing."

OTA rules changed — previously Ctrip rankings relied on volume, now they weigh comprehensive scores and repeat purchase rates. Customer mix changed — previously primarily business FITs, now conference groups and tour groups each make up half. Staff demographics changed — post-95 employees don't follow orders the way post-80s did, management approaches need adjusting.

Chen told me: "Before, my decisions based on experience were right about 80% of the time. Now they're right less than 60% of the time."

What's the problem?

Experience isn't useless — it's just that the environment experience relies on has changed.

---

2. Why Traditional Approaches Fall Short

When facing datafication trends, traditional hoteliers have three attitudes:

Attitude 1: Rejection

"Data is dead, people are alive. Hotels are a service industry — it's about people, not numbers."

Problem: This attitude is fundamentally "path dependency" — applying past success formulas to future challenges. Might work short-term, definitely fails long-term.

Attitude 2: Anxiety

"I don't understand data, but I know it matters. I need to hire data analysts, build BI systems, create data centers."

Problem: This attitude is fundamentally "technology worship" — thinking systems solve problems. Without data thinking, systems are just decoration.

Attitude 3: Wait-and-see

"I'll wait and see, wait for the industry to mature."

Problem: This attitude is fundamentally "opportunism" — by the time everyone knows how to do it, your window of opportunity is gone.

The common thread: None of these approaches treat "data thinking" as a fundamental capability to develop.

---

3. The MBCT Perspective

We ran a "data maturity assessment" on Chen's hotel:

Total: 6 out of 20 — "data infancy."

More importantly, we discovered an interesting phenomenon: Chen's hotel actually has plenty of data — PMS has data, OTA backend has data, membership system has data, even staff scheduling system has data. But this data is scattered across different systems — nobody's consolidated it for analysis.

This is the classic symptom of "has data, no analysis."

The deeper problem?

Data itself doesn't create value — integration, analysis, and decision-making create value.

Most hotels aren't lacking data. They're lacking knowledge of how to use data.

---

4. The Right Solution

Step 1: Build a Core Metrics System — The "North Star Metric"

We helped Chen establish a "hotel operational core metrics framework":

Revenue Metrics

• RevPAR (Revenue Per Available Room) = Occupancy Rate × ADR
• GOP (Gross Operating Profit) = Total Revenue − Operating Costs
• GOP Margin = GOP ÷ Total Revenue

Operational Metrics

• Occupancy Rate: Rooms occupied today ÷ Total rooms
• ADR: Total room revenue ÷ Rooms sold
• Guest source mix: Channel breakdown (OTA/Direct/Group/FIT)

Service Metrics

• Complaint rate = Complaints ÷ Total check-ins
• Housekeeping pass rate = Passed inspections ÷ Total cleanings
• Maintenance response time: Average time from report to resolution

Membership Metrics

• Repeat rate: Return guests ÷ Total guests
• Member activity: Monthly active members ÷ Total members
• Member contribution rate: Member orders ÷ Total orders

Step 2: The Three-Step Data Collection — "Collect-Clean-Use"

Step 1: Collect

Break down data silos — consolidate PMS, OTA, finance, and membership system data into one platform.

Tools: Cloud PMS (like Shiji, Xishan) have built-in data consolidation, or use BI tools (FanRuan, Power BI) for data connections.

Step 2: Clean

After data consolidation, clean it:

• Deduplication: Same order counted multiple times across systems
• Standardization: Unify definitions (e.g., how "occupancy rate" is calculated)
• Validation: Identify and handle outliers

Step 3: Use

After cleaning, build a "data dashboard" letting management see key metrics daily.

Core principle: Data must be visible, trackable, and alertable.

Step 3: Data-Driven Decision Process — "Six-Step Method"

We helped Chen build a data-driven decision workflow:

1. Collect: Yesterday's/last week's operational data
2. Compare: vs. same period last year, vs. budget, vs. competitors
3. Analyze: Find gaps, root causes, opportunities
4. Decide: Form action plans based on data
5. Execute: Implement action plans
6. Review: Assess results, enter next cycle

Step 4: Build Data Culture — "Top to Bottom"

The biggest barrier to datafication isn't technology — it's culture.

We suggested Chen start with three initiatives:

First: Review data daily Management spends 10 minutes each morning looking at the data dashboard, understanding what happened yesterday.

Second: Support claims with data When discussing issues in meetings, require "support claims with data" — not "I think."

Third: Reward data innovation Publicly recognize and reward anyone who uses data to discover an optimization opportunity.

---

5. The Emotional Value Angle

Chen told me later his biggest change: he stopped fearing data and started "having conversations with data."

Before, he felt data was "cold, impersonal." Now he realizes data is "eloquent" — it tells you where guests come from, what they like, when they'll likely return.

From MBCT's view, the most important emotional value of datafication is certainty.

When you know why guests come and why they leave, anxiety fades away.

Data gives managers a "basis for judgment" — not "the judgment result itself." The basis is certain; the judgment still requires human decision-making.

So datafication isn't "turning people into machines" — it's "giving human judgment more confidence."

---

6. Results

One year after implementation:

• Chen's decision accuracy improved from 60% to 82%
• Hotel RevPAR increased 12%
• GOP margin improved 5 percentage points
• Guest repeat rate increased 18%

More importantly, Chen's mindset shifted — he stopped saying "going with experience" and started saying "checking the data."

---

7. Key Takeaways

The core lesson: Datafication's essence is "using data to support decisions" — not "using data to replace decisions."

The old way: "trust intuition, shoot from the hip" — results are "at the mercy of fate."

MBCT's way: Build metric systems, break down data silos, let data speak.

Core principle: Data is a tool, decisions are human. The best datafication is when data helps humans make better judgments — not when it replaces human judgment.

---

Source: marvelbros.com/zh/lean