pricing, search, and ot as
DESCRIPTION
TRANSCRIPT
1
Revenue Management – Pricing, Search and OTAs
Chris K [email protected]
Two Hotelies in trouble
Bill and Ted are suspected of a crime committed by two persons. They are being questioned by authorities in two separate rooms.
Each is being encouraged to cooperate (confess). There g g p ( )is very little evidence so if neither confess they will get off w/ small fine.
Two Hotelies in trouble
T: Small FineB: Small Fine
T: Long PrisonB: Free
Ted
Don’tConfess
T: FreeB: Long Prison
T: Short PrisonB: Short Prison
Ted
Bill
Don’t Confess ConfessConfess
Confess
Likely outcome?
T: Small FineB: Small Fine
T: Long PrisonB: Free
Ted
Don’tConfess
T: FreeB: Long Prison
T: Short PrisonB: Short Prison
Ted
Bill
Don’t Confess ConfessConfess
Confess
2
Price Cut/War!
Price Cut/War!
Ted
Hold
Ted
Bill
Hold Cut
Cut
Price Cut/War!
T: Moderate ProfitB: Moderate Profit
T: No MoneyB: Big Profit
Ted
Hold
T: Big ProfitB: No Money
T: Tiny ProfitB: Tiny Profit
Ted
Bill
Hold Cut
Cut
What is the result?
HP vs DellPampers vs HuggiesMarboroMarboroEtc…
’92 fare wars
3
Fare Wars
’92 a lot of variance in fares, customer’s buying two round trips to avoid S/SOAirlines w/ lots of capacity LF ~60%AA announces ‘value’ faresDelta, UA followTWA undercutsNWA 2-for-1AA 50% offRecord load factors, -20% in $$
AA, drops value fares, chairman“…we are more victims than villains – victims of our
dumbest competitor… the business is driven entirely by the behavior of our competitors….each airline y pdoing what’s best for itself versus the industry”
Industry Characteristics & PWs
SupplyCostCapacity UtilizationProduct Perishability
DemandPrice sensitivity of demandEfficient of shoppingProduct Perishability
Product Differentiation
Efficient of shoppingBrand loyaltyGrowth rate
Price Customization
4
“If I have 2000 customers on a given route and 400 different prices, I am obviously
Price Customization
pshort 1600 prices.”
-Robert L. CrandallFormer CEO of American Airlines
380
Room Response CurveNumber of rooms Sales Response Curve
B
varia
ble
unit
cost
0.00.0
10 390Variable Unit Cost
A C
Pric
e be
low
v
Sales Price
380
Room Response CurveSales Volume Sales Response Curve
190 riabl
e un
it co
st
B
0.00.0 10 390
A200
The Maximum Profit Rectangle forSingle Price (ADEF)Pr
ice
belo
w v
ar
C
D
F
E
Sales Volume
The Maximum Profit Rectangle forSingle Price
X Money Left on the Table;willing to pay more but priced too cheap so people
Passed Up Profit because reservationprice under 200
190
varia
ble
unit
cost B
(25%)
380
16
0.00.0 10 390200
Y(25%)
paid the cheaper rate; called consumer surplus.
Pric
e be
low
v
A C
50%
5
380
Room Response CurveSales Volume Sales Response Curve
X1B
254The Maximum Profit R t l far
iable
uni
t cos
t
0.00.0 10 390
Y2
Y1
263137C
Rectangle forPrice 1
The Maximum Profit Rectangle forPrice 2A
127
127
Pric
e be
low
va
Differential Pricing
Tapping segments with different ‘willingness to pay’Different ‘products’ offered to leisure versus business travelersPrevent diversion by setting restricitionsPrevent diversion by setting restricitions
Fences to Manage Segments
Differentiate ProductsPurchase FencesValue-added
Communicate Product Differentiation
Product-line Sort As A Way to Build Fences
Develop a product line and have customers sort themselves among the various offerings based on their preference (e.g., room with view)Can have vertical differentiation (good, better, best)(g )
appliances
6
“Potential” Fences
Rule Type Advanced Requirement
Refundability Changeability Must Stay
Advance Purchase
3- Day Non refundable No Changes WE
Advance Reservation
7-Day Partially refundable (% refund or fixed $)
Change to dates of stay, but not number of rooms
WDReservation (% refund or fixed $) but not number of rooms
14- Day Fully refundable Changes, but pay fee, must still meet rules
21-Day Full changes, non-refundable
30-Day Full changes allowed
Biggest Mistakes in Price Customization
Companies aim mostly for the low-price triangle (discounting), but not for the high-price triangle.
Goal:Price customization should not bring the average price down!
Fencing is not effectiveCustomer with high willingness to pay slip into low price categories
LEAKAGE
Price cutsWithout perfect fences rate cuts ‘leak’ more demand than they ‘tap’
Lessons from air travel
Post 2000Growth of low-fare airline, with unrestricted faresPrice matching by ‘legacy’ carriersIncreased consumer search
Movement to ‘simplified’ fares
7
Contemplating a price action?
Questions to ask?
How much must occupancy increase to profit from a price decrease?
Unilateral actionMatchMatch
How much can occupancy decline before a price increase becomes unprofitable?
Unilateral actionMatch or not match
Calculate the minimum sales volume necessaryfor the volume effect to balance the price effect.
Price
P1
P2
ΔP A
Contribution margin (CM)CM = P – VC
Breakeven ANALYSIS
Service/Rooms
Demand
Variable Cost
P2
Q2Q1
ΔQ
B A = CM lost B= CM gained
8
(P-C)Q=Original Profit(P+ΔP-C)(Q +Δ Q)=New after decrease(P-C)Q=(P+ΔP-C)(Q +Δ Q)PQ-CQ=PQ+ΔPQ-CQ+PΔQ+ΔPΔQ-CΔQ
BE ANALYSIS ΔP – assumed –ve herei.e. price cut
PQ CQ PQ ΔPQ CQ PΔQ ΔPΔQ CΔQΔQ (P-C+ΔP)=-QΔPΔQ/Q=-ΔP/(P-C+ΔP)
- ΔP
CM + ΔP %BE = X 100
• Breakeven (BE) – Minimum change in sales volumeor occupancy to offset a price change
• Percent Breakeven (%BE) – Minimum percent change in sales volume or occupancy to offset a
BE ANALYSIS
change in sales volume or occupancy to offset aprice change
%BE = ΔQ / Q X 100- ΔP
CM + ΔP %BE = X 100
Suppose a hotel is considering a $25 per room night price increase from its present price of $150 and its variable cost per room night is $15.
Room night decrease for the property to breakeven?CM = P – VC = $150 - $15 = $135
-$25- ΔP
BE Example
$25
$135 + $25=
Percent Breakeven = -15.6%
Price increase must not cause more than a -15.6% loss in volume for the hotel to break even!
Percent Breakeven =CM + ΔP
x 100 x 100
MARKET – PRICE REACTION
Hotels are part of a competitive set
Constantly evaluating matching price actions by competitors:competitors:
What is the minimum potential occupancy loss that justifies matching a competitor’s price cut?
What is the minimum potential occupancy gain that justifies not matching a competitor’s price increase?
9
PRICE REACTIONCompetitor drops price ΔPAssume we will loose some volume
How much? Are we better off losing volume or losing margin?
If we follow lost margin ΔP/CMIf we follow - lost margin= ΔP/CMIf we don’t follow lost sales ΔQBE= ΔQ/Q= ΔP/CM
BE = Δ P
CM
Suppose a competitor lowers price by $10 andcurrent price is $100.
or %BE = %Δ P
%CM
Variable cost is $20.CM = $100 – $20 = $80
%BE = %Δ P
%CM =
$10 / $100
$80 / $100 X 100 = 12.5%
If the property loses more than 12.5% of room nights sold, it will take a contribution loss!
Price Elasticity
P = Current price of a goodQ = Quantity demanded at that price
ΔP = Small change in the current priceΔQ = Resulting change in quantity demandedΔQ = Resulting change in quantity demanded
Price in Change PercentageQuantity in Change PercentageElasticity =
Elasticity
PP
Δ
= Δ
Size of Price Elasticities
0 1 2 3 4 5 6
Unit elastic
Inelastic Elastic
Unit elastic: price elasticity equal to 1
• Elastic: price elasticity greater than 1
• Inelastic: price elasticity less than 1
10
Price
P2
P1
Price
P2
P1Demand
SALES CURVES and PRICE ELASTICITY
Quantity Q2 Q1
ElasticQuantity Q2 Q1
E < 1 % Q % P<
Inelastic
E > 1 % Q % P>
Demand
Price
P2
P1
Price
P2
P1
VC VC
SALES CURVES and PRICE ELASTICITY
Quantity Q2 Q1
ElasticQuantity Q2Q1
Inelastic
E > | 1 | P Contribution E < | 1 | P Contribution
If a market or market segment is price elastic (є > | 1 |),then raising price will reduce contribution. So, lowering price
(or matching a competitor’s price reduction) is the onlycontributory action!
SALES CURVES and PRICE ELASTICITY
If a market or market segment is price inelastic (є < | 1 |),then lowering price will reduce contribution. So, raising price
(or matching a competitor’s price increase) is the onlycontributory action!
ImpactPrice cuts need to be segmented to be incremental versus dilutiveAvoiding blanket discounts
Opaques (HW, PCLN, Top Secret) PackagesPackagesEmail offers TravelzooSearch Engine Marketing/PPCOTA promotion/positioning/flash offersGDS positioning Amadeus Instant Preference, Sabre Spotlight
11
OPAQUE PRICING
Priceline TutorialMedian retail pricing is provided to give customers a realistic benchmark for offers
Opaque Offer Guidance
12
• Only if the offer is accepted will the customer receive specific hotel information
• If the offer is unsuccessful, the customer is given an invitation to “try again” by changing one of their search criteria
• Customers cannot resubmit their offer by only changing their offer price
Hotwire
Lastminute.com
Travelocity
13
Expedia
Inline banners on Results page to Opaque pageNo access to results from home page
All inventory sourced through HotwireCo-branded as Hotwire Pricing, sort, content from Hotwire
Launch integrates ‘basic’ opaque product
Extending reach
Launch integrates basic opaque product No reviewsNo Bed ChoiceAmenities limitedFilters limited
50
Performance metricsImproved conversion by ~1%Star rating distribution
Averages between HW Opaque and Expedia Merchant
Booked ADRs boosted for hotelsUp 7.4% compared to Hotwire
Expedia Opaque Performance
2 2 5 3 3 5 4 4 5 5
51
2 2.5 3 3.5 4 4.5 5Hotwire Expedia Opaque Expedia Merchant
The Six Points of OpacityThe Six Points of OpacityLess Opacity = More DilutionLess Opacity = More Dilution
Opaque Transparent
Priceline Merchant HotwirePRICES
14
How they work?
TravelocityAll opaque offerings listed
Hotwire/Expedia UnpublishedOne star per zoneOne star per zone
Usually the lowest priced supplier
PricelineRandom allocation
Based on the customer’s search criteria, a list of eligible hotels is created
From this list begins the “First Look” processOne hotel is chosen at random, without regard for rates or availability
Then an availability search is done in Worldspan to see if the chosen hotel has a qualifying priceline rateIf a qualifying rate is found, the reservation is made and the process is complete
PCLN - How A Hotel Is Chosen
completeIf the chosen hotel fails, begin the “Second Look” process
Remaining hotels are ranked in order of their recent 14 day performance with priceline “First Looks” (hotel’s “Batting Average”)Then one by one, priceline rates and inventory are searched in Worldspan for each hotelAs soon as a hotel is found with a qualifying priceline rate, the reservation is made and the process is completeIf no hotel has a qualifying priceline rate, the customer will be notified that their offer could not be fulfilled
The highest qualifying rate is usually booked giving hotels more revenue
Hotels are encouraged to load multiple rate tiersProvides hotels with opportunity to accept more offers at various price points45% of bookings are at rates above the minimum tier
The Rate That Is Booked
For example: Guest offers: $100Hotel available priceline rates: $100, $88, $78Priceline will book: $88Priceline will book: $88
If $78 and $88 rates are closed out, priceline may book the $100 rate (making $0 margin) if no other partner has an available qualifying rate
DATA
15
Summary data of bids
0.2
0.25
0.3
0.35
0.4
Weekend
0
0.05
0.1
0.15
$125 $150 $175 $200 $225 $250 $275 $300 $325 $350 $375 $400 $425
Center for Hospitality Research
Setting Room Rates on Priceline: How to Optimize Expected Hotel Revenuehttp://www.hotelschool.cornell.edu/research/chr/pubs/reports/abstract-14705.htmlhttp://www.hotelschool.cornell.edu/research/chr/pubs/tools/tooldetails-14706 ht l14706.html
Making the Most of Priceline’s Name-Your-Own-Price Channelhttp://www.hotelschool.cornell.edu/research/chr/pubs/reports/abstract-15296.html
There’s an APP for that….
16
“Hotel Negotiator” initial release Fall 2009
RetailListings or Retail radar – point to see nearby hotels and rates
Winning BidsShake or Select city to see recent Winning Bids
Re-designed Bid NowImproved screen layout makes it clear how to change dates, adds a “Help” option, and supports user-entered bid amounts.
Opaque RadarSee nearby areas and winning bids. Plus, both retail and opaque radars gain new zoom and filtering capabilities.
Income Comparison: OTA Hotel Prospects
Income Comparison – OTA Hotel Prospects(% breakdown of visitors to each OTA hotel section, Jan-Jun 2007)
25%
30%35%
40%45%
0%5%
10%15%
20%
<$30K $30-60K $60-100K $100K+
Expedia Prospects Orbitz Prospects Travelocity Prospects PCLN NYOP Prospects PCLN Retail Prospects
17
HTTP://BiddingForTravel.com
BiddingForTravel – The Fanatics
http://biddingfortravel.yuku.com/topic/98782/t/The-Curtain-is-Parted-More-or-Less.html
18
Search – SEO/SEM
What influences online travel purchases?
Base: Total usual online shoppersNote: What shopping for personal travel, how influential are (insert) in deciding what to purchase?Note: Reflects those respondents indicating these travel providers as being “strongly influential” or “somewhat influential” on a 3-point scaleSource: The PhoCusWright Consumer Travel Trends Survey Ninth Edition
Goal 1: Rank High When ConsumerSearches on Internet
Goal 2: Click Through to Reservation
19
Search Engine Technology
Organic and Paid Searches
Organic Results
Paid Results
Organic Results
Organic Results
Local Results
Organic and Paid Searches
Organic and Paid Searches
Paid Results
20
How do SE determine page position?
Google’s Measure of Importance of Page
Download from www.google.com
Keyword Phrases
What are people looking for?
Key to Success: The Right Keyword Phrases
How are they finding you today?
How are they finding your
competition today?
Google’s Cache will show you what keywords it’s reading on the site.
Search: New York City Midtown Hotel
Search: New York City Midtown Hotel
21
The Long Tail of Search
The Head—Branded
The Tail—Unbranded
Uses Search Engines Algorithmic Calculations
Pay to Search Engines to Rank High (Cost-per-Click)
PPC Performance
2nd price sealed bid auctionSubmit bid, pay 1 penny more than bidder cheaper than you that gets accepted
22
Keyword types
Search – “red eye from LAX”
Negative keywords
Impressions (I)Click–through rate (CTR)Cost per click (CPC)Conversion rate (CR)Conversion rate (CR)Average revenue (V)
CTRCR
CPC
BID
CPC
Expected Daily spendCTR*CPC*I
23
CTR
SPEND
BID
CPCSPEND
Expected Daily spendCTR*CPC*I
Expected Return per impressionCTR*CR*V – CTR*CPCCTR CR V CTR CPC
CR
Return/I
BID
Expected Daily spendCTR*CPC*I
Expected Return per impressionCTR*CR*V – CTR*CPCCTR CR V CTR CPC
Expected Return per booking(CTR*CR*V-CTR*CPC)/(CTR*CR)
24
Expected Return per booking – SELF FUNDING KEYWORDS
+ve
BID
O
-ve
Quality issues
Both paid and natural search are quality adjusted listsContentCTRLinks
Google is maximizing its PROFITS!
What is Google Quality Score?Quality Score for Google and the search network is a dynamic metric assigned to each of your keywords. It's calculated using a variety of factors and measures how relevant your keyword is to your ad group and to a user's search query. The higher a keyword's Quality Score, the lower its minimum bid and the better its ad position.
The components of Quality Score vary depending on whether it's calculating minimum bid or ad position:Quality Score for minimum bid is determined by a keyword's clickthroughQuality Score for minimum bid is determined by a keyword s clickthrough rate (CTR) on Google, the relevance of the keyword to its ad group, your landing page quality, your account's historical performance, and other relevance factors. Quality Score for ad position is determined by a keyword's clickthrough rate (CTR) on Google, the relevance of the keyword and ad to the search term, your account's historical performance, and other relevance factors.
Landing PagesLanding Pages are also a factor in Quality Score
Load TimeKeyword Rich ContentOriginal ContentSending the Right AdGroup to the Right Landing Page.Sending the Right AdGroup to the Right Landing Page.
If you have “Wedding” related keywords, you should consider sending them to a “Wedding” page on your site to improve relevance and Quality Score
25
Strategic Link Building
Why Link Building? Because it works…
Check on Your Competitors
www.linkpopularity.com www.compete.comwww.marketleap.com
Who’s Linking To You?
26
Different Search Engines View Links Differently
Facilitating The Reservation - Conversion
The Booking Experience on Your Website
4 Screens to Book 1 Reservation
The Booking Experience via OneScreen
27
Case Study – St. James Hotel
Best Practices in Search Engine Marketing and Optimization: The Case of the St. James Hotel
http://www.hotelschool.cornell.edu/research/chr/pubs/reports/abstract-15320.html
Search, OTAs and online booking: The Billboard Effect
Do OTAs impact non-OTA reservation volume?
Experimental study with JHM Hotels facilitated by Expedia
Four JHM properties3 Branded3 Branded1 Independent
3 month period, cycled properties on and off Expedia (7-11 days per cycle)
40 days on Expedia40 days off
For all arrival dates
Do OTAs impact non-OTA reservation volume?
“Data”Reservations made during the experimental period
d b h i hi d f h d i dStay dates both within and after the study period
Removed any reservations through ExpediaCompare (non-Expedia) reservations during the on and off treatments
28
OTA Implications – Creating Visibility
OTA Impact on non-OTA reservations
Property Non-OTA Volume Increase
Branded 1 7 5% 9 Brand family properties withinBranded 1 7.5%Branded 2 9.1%Branded 3 14.1%Independent 26%
3 Brand family properties 20 miles
9 Brand family properties within 15 miles
≈
OTA Implications – Creating Visibility
OTA Impact on non-OTA reservations/rate
Property Non-OTA Volume Increase
ADR Increase
Branded 1 7 5% 3 9%Branded 1 7.5% 3.9%Branded 2 9.1% 0.8%Branded 3 14.1% 0.3%Independent 26% 0.8%
ADR across several stay dates (in and beyond 3 month study period)
ADR increase controlling for DOW, DBA, LOS
Value Implications
OTA demand acquisition ‘costs’ spread over all impacted demand
e.g. 10% reservations through OTABillboard Effect~20%
20% of the remaining originates/impacted by OTA60% supplier direct - impacts 10% (50*1.2=60) 90% total - impacts 15% (75*1.2=90)
OTA impacted volume = 10% + (10% to 15%)Acquisition costs are less than ½ originally assumed Lower the OTA share, further decrease costs
Billboard Effect I
Probably ~ 20% lift in non-OTA reservations created through marketing effect of the OTA
depending on OTA volume results in reduction in ‘fees’ by factor of 2-4(or more)
Limitations Only 4 (mid scale) properties3 month sample window
29
Part II - Online consumer behavior
Online consumer panel (~2 million)All domain level internet traffic 2 months during each of 08,09 and 10
All upstream traffic of IHG.com bookingsSearch @ Google, Bing, YahooTravel site – OTA, Meta Search ….60 days prior to booking
Online consumer behavior74.7% of consumers visit OTA prior to booking at supplier.com82.5% perform a search
65% do both31% OTA 1st, 29% same day, 40% search 1st y
1/2 of searches are URL related2/3rds are branded
only 10.3% direct to supplier.com (no search or OTA)
Travel Site/Search Distributions
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150
Rel
ativ
e fr
eque
ncy
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150
Number of site visits
0
0.1
0.2
0.3
0.4
0.5
0.6
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150
Number of searches
Rel
ativ
e fr
eque
ncy
OTA site behavior – the first page or bust?
Pages per visit
Minutes per visit
Number of visits
Average behavior per booking (supplier.com)
visit visit visitsOTAs 7.44 4.67 11.6
30
OTA site behavior – the first page or bust?
Pages per visit
Minutes per visit
Number of visits
Average behavior per booking (supplier.com)
visit visit visitsAll OTAs 7.44 4.67 11.6Expedia 7.47 4.78 7.5
74.4% of OTA visits are to Expedia
OTA site behavior – by brand/scale
Pages per visit
Minutes per visit
Number of visits % Reservations
Candlewood Suites 9.1 5.5 6.2 5.9Crowne Plaza Hotels 9.1 5.4 13.9 9.0
Average behavior per booking (supplier.com)
Crowne Plaza Hotels 9.1 5.4 13.9 9.0Holiday Inn 7.7 4.4 11.4 80.1Staybridge Suites 8.1 4.7 9.9 3.9Hotel Indigo 7.6 4.3 23.7 0.6Inter-Continental Hotels 5.9 3.4 28.6 0.6
Channel Mix
Panel reservations at Expedia.com as wellIHG.com : Expedia.com reservations ~10:1
IHG.com% Reservations
Expedia.com% Reservations
Candlewood Suites 5 9 5 7Candlewood Suites 5.9 5.7Crowne Plaza Hotels 9.0 13.8Holiday Inn 80.1 73.2Staybridge Suites 3.9 1.6Hotel Indigo 0.6 0Inter-Continental Hotels 0.6 5.7
Billboard Part II
% IHG.com Ratio IHG.com/Expedia Reservations
Visit Expedia Expedia Only OTA All Impacted Expedia Only
61 8% 21 5% 8 7 3 061.8% 21.5% 8.7 3.0
31
Billboard Part II
Ratio IHG.com/Expedia Reservations
All Impacted Expedia Only
Candlewood Suites 7.4 2.6Crowne Plaza Hotels 5.8 1.5Holiday Inn 9.5 3.4Staybridge Suites 20 9Hotel IndigoInter-Continental Hotels 1 0
∞ ∞
Billboard Part II
% IHG.com Ratio IHG.com/Expedia Reservations
Visit Expedia Expedia Only OTA All Impacted Expedia Only
61 8% 21 5% 8 7 3 0
~3+ reservations @ IHG.com (impacted by visibility) for each @ Expedia
Similar to JHM commission reductionsIgnores non-IHG.com impact
61.8% 21.5% 8.7 3.0
SummaryView OTA as any other marketing expense
Part of the demand funnelVisibility at OTA increases non-OTA reservation volume s.t. OTA margins are on order of ¼ (or less) of actual transactional fees
The Billboard Effect: Online Travel Agent Impact on Non-OTA Reservation Volume
http://www.hotelschool.cornell.edu/research/chr/pubs/reports/abstract-15139.html
Email and Flash Offers
TravelzooSniqueAway/Jetsetter/Expedia ASAP
32
Email Blasts
33
SniqueAway (Jetsetter)
34
Travel Agent Targeted Advertising
Galileo Headlines
Generate Up to 3 Times More Sales with Preferred Placement
Why Not Be Here Tomorrow!
Preferred Placement WorksResearch shows that agents are up to 3.5 times more likely to select hotels that appear at or near the top of hotel displays.
2004 Travel Agent Media Study
Your Hotel is Here Today.