Hotel Pricing Analysis Final Report
Shwetha Bhaskaran
26 February 2018
Introduction
The hotel industry in India thrives largely due to the growth in tourism and travel. Due to the increase in tourism with rising foreign and domestic tourists, hotel sector is bound to grow. There is an emergence of budget hotels in India to cater to the majority of the population who seek affordable stay. International companies are also increasingly looking at setting up such hotels. Imbalance in increase in tourists both domestic and foreign not been supported with equal number of rooms is a latent source of opportunity for growth.
A hotel has different types of rooms on the basis of room size, location, view, décor, furnishings, amenities, etc. Therefore the front office generally has more than one room rate category depending on the types of rooms.
Overview
We are concerned with pricing startegy of hotel industry based on the evidences collected from the 42 different cities dependent on several parameters. Through this we try to establish any sort of influence of TouristDestination on Roomrents of the hotels. The data is collected from 42 cities for 8 different dates. We will do empirical study based on the dataset available with us.
Read your dataset in R and visualize the length and breadth of your dataset.
hotel <- read.csv(paste("Cities42.csv", sep=""))
View(hotel)
dim(hotel)## [1] 13232 20Create a descriptive statistics (min, max, median etc) of each variable
library(psych)
describe(hotel)## vars n mean sd median trimmed
## X 1 13232 6616.50 3819.89 6616.5 6616.50
## CityName* 2 13232 18.07 11.72 16.0 17.29
## Population 3 13232 4416836.87 4258386.00 3046163.0 4040816.22
## CityRank 4 13232 14.83 13.51 9.0 13.30
## IsMetroCity 5 13232 0.28 0.45 0.0 0.23
## IsTouristDestination 6 13232 0.70 0.46 1.0 0.75
## IsWeekend 7 13232 0.62 0.48 1.0 0.65
## IsNewYearEve 8 13232 0.12 0.33 0.0 0.03
## Date* 9 13232 14.26 2.82 14.0 14.39
## HotelName* 10 13232 841.19 488.16 827.0 841.18
## RoomRent 11 13232 5473.99 7333.12 4000.0 4383.33
## StarRating 12 13232 3.46 0.76 3.0 3.40
## Airport 13 13232 21.16 22.76 15.0 16.39
## HotelAddress* 14 13232 1202.53 582.17 1261.0 1233.25
## HotelPincode 15 13232 397430.26 259837.50 395003.0 388540.47
## HotelDescription* 16 13224 581.34 363.26 567.0 575.37
## FreeWifi 17 13232 0.93 0.26 1.0 1.00
## FreeBreakfast 18 13232 0.65 0.48 1.0 0.69
## HotelCapacity 19 13232 62.51 76.66 34.0 46.03
## HasSwimmingPool 20 13232 0.36 0.48 0.0 0.32
## mad min max range skew
## X 4904.44 1.0 13232 13231.0 0.00
## CityName* 11.86 1.0 42 41.0 0.48
## Population 3846498.95 8096.0 12442373 12434277.0 0.68
## CityRank 11.86 0.0 44 44.0 0.69
## IsMetroCity 0.00 0.0 1 1.0 0.96
## IsTouristDestination 0.00 0.0 1 1.0 -0.86
## IsWeekend 0.00 0.0 1 1.0 -0.51
## IsNewYearEve 0.00 0.0 1 1.0 2.28
## Date* 2.97 1.0 20 19.0 -1.05
## HotelName* 641.97 1.0 1670 1669.0 0.01
## RoomRent 2653.85 299.0 322500 322201.0 16.75
## StarRating 0.74 0.0 5 5.0 0.48
## Airport 11.12 0.2 124 123.8 2.73
## HotelAddress* 668.65 1.0 2108 2107.0 -0.37
## HotelPincode 257975.37 100025.0 7000157 6900132.0 9.99
## HotelDescription* 472.95 1.0 1226 1225.0 0.11
## FreeWifi 0.00 0.0 1 1.0 -3.25
## FreeBreakfast 0.00 0.0 1 1.0 -0.62
## HotelCapacity 28.17 0.0 600 600.0 2.95
## HasSwimmingPool 0.00 0.0 1 1.0 0.60
## kurtosis se
## X -1.20 33.21
## CityName* -0.88 0.10
## Population -1.08 37019.65
## CityRank -0.76 0.12
## IsMetroCity -1.08 0.00
## IsTouristDestination -1.26 0.00
## IsWeekend -1.74 0.00
## IsNewYearEve 3.18 0.00
## Date* 2.93 0.02
## HotelName* -1.25 4.24
## RoomRent 582.06 63.75
## StarRating 0.25 0.01
## Airport 7.89 0.20
## HotelAddress* -0.88 5.06
## HotelPincode 249.76 2258.86
## HotelDescription* -1.25 3.16
## FreeWifi 8.57 0.00
## FreeBreakfast -1.61 0.00
## HotelCapacity 11.39 0.67
## HasSwimmingPool -1.64 0.00Create one-way contingency tables for the categorical variables in your dataset.
mytable <- with(hotel,table(CityName))
mytable## CityName
## Agra Ahmedabad Amritsar Bangalore
## 432 424 136 656
## Bhubaneswar Chandigarh Chennai Darjeeling
## 120 336 416 136
## Delhi Gangtok Goa Guwahati
## 2048 128 624 48
## Haridwar Hyderabad Indore Jaipur
## 48 536 160 768
## Jaisalmer Jodhpur Kanpur Kochi
## 264 224 16 608
## Kolkata Lucknow Madurai Manali
## 512 128 112 288
## Mangalore Mumbai Munnar Mysore
## 104 712 328 160
## Nainital Ooty Panchkula Pune
## 144 136 64 600
## Puri Rajkot Rishikesh Shimla
## 56 128 88 280
## Srinagar Surat Thiruvanthipuram Thrissur
## 40 80 392 32
## Udaipur Varanasi
## 456 264mytable2 <- with(hotel,table(FreeBreakfast))
mytable2## FreeBreakfast
## 0 1
## 4643 8589mytable3 <-with(hotel,table(HasSwimmingPool))
mytable3## HasSwimmingPool
## 0 1
## 8524 4708mytable4 <- with(hotel, table(StarRating))
mytable4## StarRating
## 0 1 2 2.5 3 3.2 3.3 3.4 3.5 3.6 3.7 3.8 3.9 4 4.1
## 16 8 440 632 5953 8 16 8 1752 8 24 16 32 2463 24
## 4.3 4.4 4.5 4.7 4.8 5
## 16 8 376 8 16 1408mytable5 <- with(hotel, table(IsMetroCity))
mytable5## IsMetroCity
## 0 1
## 9472 3760mytable6 <- with(hotel, table(IsTouristDestination))
mytable6## IsTouristDestination
## 0 1
## 4007 9225Create two-way contingency tables for the categorical variables in your dataset.
aggregate(hotel$RoomRent, by=list(weekend = hotel$IsWeekend, newyearseve = hotel$IsNewYearEve), mean)## weekend newyearseve x
## 1 0 0 5429.473
## 2 1 0 5320.820
## 3 0 1 8829.500
## 4 1 1 6219.655aggregate(hotel$RoomRent, by=list(Metrocity = hotel$IsMetroCity, TouristPlace = hotel$IsTouristDestination), mean)## Metrocity TouristPlace x
## 1 0 0 4006.435
## 2 1 0 4646.136
## 3 0 1 6755.728
## 4 1 1 4706.608aggregate(hotel$RoomRent, by=list(freewifi = hotel$FreeWifi, freeBreakfast = hotel$FreeBreakfast, swimmingPool = hotel$HasSwimmingPool), mean)## freewifi freeBreakfast swimmingPool x
## 1 0 0 0 3538.085
## 2 1 0 0 3148.628
## 3 0 1 0 5636.617
## 4 1 1 0 3984.457
## 5 0 0 1 7378.590
## 6 1 0 1 9530.906
## 7 0 1 1 5207.000
## 8 1 1 1 8246.284Draw a boxplot of the variables that belong to your study.
boxplot(hotel$StarRating,
xlab="Star Rating of the hotel",
main="Box plot of Star Rating of hotel",
col = "yellow",
horizontal = TRUE)
boxplot(hotel$Airport,
xlab="Distance between Hotel and closest major Airport(in km)",
main="Box plot of Airport Distance of hotel",
col = "yellow",
horizontal = TRUE)
boxplot(hotel$HotelCapacity,
xlab="Hotel Capacity",
main="Box plot of Hotel Capacity",
col = "yellow",
horizontal = TRUE)
Draw Histograms for your suitable data fields.
hist(hotel$StarRating, main = "star Rating Distribution", xlab = "Stars")
hist(hotel$Airport, main = "Distrubtion of distance to the nearest major airport", xlab = "Dist to the nearest major Airport in km")
hist(hotel$HotelCapacity, main = "distribution of capacity of hotels", xlab = "capacity of hotels")
hist(hotel$RoomRent,
main="Analysis of room rents of hotels",
xlab="Rents of room",ylab="Relative frequency",
xlim = c(0,30000),breaks=30,
col="green")
hotel$FreeWifi=factor(hotel$FreeWifi, levels=c(0,1), labels=c("No","Yes"))
plot(hotel$FreeWifi,col="yellow",main="Has Wifi?")
hotel$FreeBreakfast=factor(hotel$FreeBreakfast, levels=c(0,1), labels=c("No","Yes"))
plot(hotel$FreeBreakfast,col="blue",main="Has Free Breakfast?")
hotel$HasSwimmingPool=factor(hotel$HasSwimmingPool, levels=c(0,1), labels=c("No","Yes"))
plot(hotel$HasSwimmingPool,col="green",main="Swimming pool?")
Draw suitable plot for your data fields.
library(car)##
## Attaching package: 'car'## The following object is masked from 'package:psych':
##
## logitscatterplotMatrix(formula = ~ RoomRent + IsWeekend + IsNewYearEve, data = hotel, pch = 16)## Warning in smoother(x, y, col = col[2], log.x = FALSE, log.y = FALSE,
## spread = spread, : could not fit smooth
#Create a correlation matrix.
cor(hotel$RoomRent, hotel[,c("IsWeekend","IsNewYearEve")])## IsWeekend IsNewYearEve
## [1,] 0.004580134 0.03849123Visualize your correlation matrix using corrgram.
library(corrgram)
corrgram(hotel, order=TRUE, lower.panel=panel.shade,
upper.panel=panel.pie, text.panel=panel.txt,
main="Hotel Pricing Corrgram")
#Create a scatter plot matrix for your data set.
scatterplotMatrix(formula = ~ RoomRent + StarRating +Airport+ IsWeekend + IsNewYearEve ,
data = hotel ,
main="Scatter plot matrix of external factors")## Warning in smoother(x, y, col = col[2], log.x = FALSE, log.y = FALSE,
## spread = spread, : could not fit smooth
## Warning in smoother(x, y, col = col[2], log.x = FALSE, log.y = FALSE,
## spread = spread, : could not fit smooth
## Warning in smoother(x, y, col = col[2], log.x = FALSE, log.y = FALSE,
## spread = spread, : could not fit smooth
scatterplotMatrix(formula = ~ RoomRent + FreeWifi + FreeBreakfast + HotelCapacity + HasSwimmingPool,
data = hotel ,
main="Scatter plot matrix of internal factors")## Warning in smoother(x, y, col = col[2], log.x = FALSE, log.y = FALSE,
## spread = spread, : could not fit positive part of the spread
T-tests
Null Hypothesis:Their is no Difference between the Room Rent on new year’s eve and on other days
t.test(RoomRent~IsMetroCity,data=hotel)##
## Welch Two Sample t-test
##
## data: RoomRent by IsMetroCity
## t = 10.721, df = 13224, p-value < 2.2e-16
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
## 888.0308 1285.4102
## sample estimates:
## mean in group 0 mean in group 1
## 5782.794 4696.073Since p-value<0.05, we reject the null hypothesis.Therefore , we can conclude that the room rents of Hotels in non-metro cities is more than that of metro cities.
Null Hypothesis:There is no significant difference between the Room Rent of Hotels on normal Eve and New Year’s Eve.
t.test(RoomRent~IsNewYearEve,data=hotel)##
## Welch Two Sample t-test
##
## data: RoomRent by IsNewYearEve
## t = -4.1793, df = 2065, p-value = 3.046e-05
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
## -1256.5297 -453.9099
## sample estimates:
## mean in group 0 mean in group 1
## 5367.606 6222.826Since p-value<0.05, we reject the null hypothesis. Therefore, we can say that the room rents of Hotels on normal days are cheaper than that on New Year’s Eve.
Regression Analysis
In order to test Hypothesis , we proposed the following model:
RoomRent= b0 + b1StarRating + b2Airport + b3FreeWifi + b4FreeBreakfast + b5* HotelCapacity + b6HasSwimmingPool + b7IsMetroCity + b8IsNewYearEve + b9IsTouristDestination + error
Linearregression <- lm(RoomRent ~ StarRating + Airport + FreeWifi + FreeBreakfast + HotelCapacity + HasSwimmingPool+IsMetroCity+IsNewYearEve
+IsTouristDestination, data = hotel)
summary(Linearregression)##
## Call:
## lm(formula = RoomRent ~ StarRating + Airport + FreeWifi + FreeBreakfast +
## HotelCapacity + HasSwimmingPool + IsMetroCity + IsNewYearEve +
## IsTouristDestination, data = hotel)
##
## Residuals:
## Min 1Q Median 3Q Max
## -11696 -2375 -701 1063 309539
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) -8906.418 405.396 -21.970 < 2e-16 ***
## StarRating 3564.570 110.489 32.262 < 2e-16 ***
## Airport 11.265 2.710 4.157 3.24e-05 ***
## FreeWifiYes 485.597 224.134 2.167 0.0303 *
## FreeBreakfastYes 182.992 123.296 1.484 0.1378
## HotelCapacity -10.990 1.026 -10.714 < 2e-16 ***
## HasSwimmingPoolYes 2227.069 159.327 13.978 < 2e-16 ***
## IsMetroCity -1548.328 138.527 -11.177 < 2e-16 ***
## IsNewYearEve 844.123 174.046 4.850 1.25e-06 ***
## IsTouristDestination 2113.725 134.336 15.735 < 2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 6607 on 13222 degrees of freedom
## Multiple R-squared: 0.1887, Adjusted R-squared: 0.1881
## F-statistic: 341.7 on 9 and 13222 DF, p-value: < 2.2e-16Result
The starRating and Tourist Destination coefficient is positive which will increase the price of the hotels. SwimmingPool facility the internal feature will positively affect the hotel prices. NewYearEve will certainly make the Tourist Destination much more hot favourite which will further esclate the prices.
Conclusion
Factors like StarRating, IsMetroCity, IsTouristDestination and IsNewYearEve affect RoomRent much much significantly. Factors like FreeWifi and FreeBreakfast have a very little effect on Roomrent as their confidence interval is closer to zero. Factors like Distance from Airport and Hotel Capacity are not statistically significant for RoomRent as their confidence interval includes zero.