Customer Segmentation
Customer Segmentation: Leverage Emerging Fencing Techniques to Optimize Revenue Performance
2
Josh Belkin Sr Director, Global Retail Merchandising Hotels.com 3
Fencing enables hotels to set prices and promotions based on customer segments Fenced rates for valuable customers while still protecting public rates
Customer segments for fenced rates can be defined by several traits 1
Fenced rates
Public rates
Fenced customers Public customers
4
Purchase device
•
Booking on desktop vs. mobile device
2 Customer origin/ location
•
Domestic vs. international customers vs. local customers
3
•
Loyalty program members vs. email subscribers vs. others
4
Customer loyalty
Other • observable traits
Any other valuable segmentation traits
Fencing is important as prices and discounts are key to influence hotel buying decision of customers Factors influencing last hotel decision for hotel stayers
Price is the top factor influencing hotel booking decision
~ 60% of customers said Price influenced their hotel decision
63%
Price
61%
Location of Hotel Property
32%
Previous experience
Hotel brand
30%
Amenities
29%
Special offer/discount
29% 26%
Positive hotel reviews
23%
Hotel pictures
Hotel star rating
21%
Loyalty membership
20% 14%
Restaurants in the hotel
Special offers/discounts also has a large role in influencing decision
Recommendations from friends/family
13%
Hotel size
12%
Recent renovation
4%
Question: “What factor influenced your last hotel decision? Select all that apply” Source: Phocuswright (Dis)Loyalty and the US Leisure Traveler
5
1
Purchase device fencing: target mobile customers who book last minute
1 in 4 hotel room nights are booked on a mobile device (and increasing!)
~ 1 in 2 mobile customers book same day/day before their trip 2015 Advance booking window for domestic US travel
2015 Forecasted YoY growth in US digital travel sales
40.7% 53%
1+ days
80%
5.5% 47%
-2.2%
0-1 days
20% Desktop
Desktop/ Laptop
Mobile
Source: Hotels.com data
Mobile
Source: eMarketer, April 2014, October 2014
6
Overall
2
Customer origin fencing: target international travelers while protecting domestic pricing
Foreign exchange rates can impact demand patterns
International customers book earlier and stay longer 2015 Booking window (months) for US hotels
GBP, EUR and CAD/USD exchange rate (Index, July 2014=100)
27% 100 Domestic customers International customers
8% 6% 13% 27%
13%
3+ mo
GBP
73% 19%
90
Jul ’14
< 1 mo
58%
Source: Hotels.com data
EUR
80
42%
2-3 mo 1-2 mo
CAD
Sep ’14
Nov ’14
Jan ’15
Mar ’15
May ’15
US Hotels Growth rate (Indexed to July 2014 = 100) 2015 Length of stay (days) in US hotels Domestic customers
120
80
+1.2 days International Customers
US->US
100
1.8
Int’l -> US
60
3.0
Jul ’14
Source: Hotels.com data
Sep ’14
Source: Hotels.com data
7
Nov ’14
Jan ’15
Mar ’15
3
Customer loyalty fencing: target loyalty program members who travel and comparison shop frequently
The loyalty traveler population is vast and they travel frequently • •
Loyalty members are savvy travelers looking for best deals •
1 in 2 hotel stayers and 1 in 3 OTA shoppers are loyalty members Loyalty participation increases with trip frequency so they are more likely frequent travelers
•
US Loyalty traveler population (M) and program membership (% of segment)
Loyalty members are more likely to shop across multiple websites than non loyalty members 61% of elite program members said they’d check multiple websites to get a good deal Online shopping behavior by loyalty status “When planning travel, I always check multiple websites to make sure I am getting a good deal”
118M Non-members
55M 56M (47%)
Loyalty members
No hotel loyalty
42%
Entry level/ Mid-tier
41%
Elite
18M (33%)
Hotel Stayers OTA shoppers
36% 41% 61%
Strongly agree
Neutral/Unsure
Slightly agree
Slightly disagree
25% Strongly disagree
Source: Phocuswright (Dis)Loyalty and the US Leisure Traveler
Source: Phocuswright (Dis)Loyalty and the US Leisure Traveler
8
Linda Gulrajani, CRME VP, Revenue Strategy & Distribution Marcus Hotels & Resorts Member of HSMAI’s Revenue Management Advisory Board 9
Scenario 1: Branded Strategy (Hilton) Target Customer Segment: Past transient guests who don’t already have a reservation and did not book a negotiated rate.
Fence & Method: $79 rate with a calendar that had the dates with that rate available highlighted (had other rates available for all dates) Sent out one email campaign that was bookable for 7 days and covered a 6 week booking window and then sent another one when there was only one day left with the same offer.
Success: Booked $27k in revenue The second “one day left” email out performed the first email
Tracking: Through the Clairvoyix and Hilton.com 10
Scenario 2: Resort Strategy Target Customer Segment: Transient guests who have booked through a discount channel like Groupon, Travelzoo, etc. in the past that don’t have a future reservation (Very important to collect email addresses from these guests).
Fence & Method: Email 48 hour sale promoting “book now and get the lowest rates of the year” sent out 1 week prior to the Groupon/Travelzoo offer going out
Success: Booked 1,030 rooms through this offer, which decreased the next “special offer” production significantly, saving us the commission and capturing them direct
Tracking: With a special rate code
11
Scenario 3: OTA Strategy Target Customer Segment: Fenced guests (or members) on OTAs
Fence & Method: Discount for members only
Success: 30% - 40% of the OTA bookings are coming from this channel Improved placement on the site This requires that a guest creates a login, but there are no other fences that prevent them from seeing the special pricing so is this really a fenced offer?
Tracking: Reports provided by market managers
12
Kathleen Cullen, CRME SVP, Revenue & Distribution Commune Hotels & Resorts Chair of HSMAI’s Revenue Management Advisory Board 13
Scenario 1: Goal: Fill Suite Occupancy without public discounting Targeted Customer Segment: Guests with stays more than once and have paid an ADR range of $400-$2500.
Fence: Pull qualified customer list from CRM
Method: Via Email Communication or private mailing with an invitation to stay in one of the remodeled luxury suites. Tracking: Email Campaign Statistics, Rate Code Production and if bookable via a landing page on brand.com Google Analytics
14
15
16
Scenario 2: Goal: Build occupancy base outside the 45 day booking window. Customer Segment: OTA International Customers
Fence: Unique POS (Point of Sale) + include a cancel policy and/or MLOS restriction
Method: Via OTA partners using Unique POS (ie: UK, Brazil, etc.)
Tracking: OTA reports or set up a specific Rate Code to track all bookings.
17
Example: Extranet Results
18
Scenario 3: Goal: Increase Website Conversions Customer Segment: Direct Bookers
Fence: Guests that are within Shopping Cart Funnel
Method: Show LightBox to potential guest that has checked rates & avail but closes out booking engine. Email trigger to potential guest that started to make their reservation but does not complete.
Tracking: Booking Engine Reporting + Partner Reporting
19
Email Abandonment
20
Booking Engine Pop-up
21
May Results # of Abandoned Carts
4,819
# of Guests we were able to invite back
548
# of Reservations Recovered
26
Reservation Value
$ 502.72
Revenue
$13,070.75 *
Cost
$392.12
ROI
$33:1 * 15% of Pre-Opening Website Revenue
22
Scenario 4: Goal: Drive occupancy on select need dates Customer Segment: Local Residents
Fence: Live in specific zip codes
Method: Serve Paid Search Text and Display Advertising to shoppers with a Florida Zip Code Dedicated Email Blast to Thompson Hotel’s email database with a Florida address Load Florida Resident Rate to a Booking Engine Filter
Tracking: Paid Search Reports, Email results + Google Analytics
23
Paid Search
24
Email Blast
25
Results April – June MTD, 2015 Room Nights
161
Reservations
71
LOS
2.3
ADR
$253.13
Revenue
$40,754.85
Cost
$7,797.71
ROI
$5:1
26
2
Josh Belkin Sr Director, Global Retail Merchandising Hotels.com 3
Fencing enables hotels to set prices and promotions based on customer segments Fenced rates for valuable customers while still protecting public rates
Customer segments for fenced rates can be defined by several traits 1
Fenced rates
Public rates
Fenced customers Public customers
4
Purchase device
•
Booking on desktop vs. mobile device
2 Customer origin/ location
•
Domestic vs. international customers vs. local customers
3
•
Loyalty program members vs. email subscribers vs. others
4
Customer loyalty
Other • observable traits
Any other valuable segmentation traits
Fencing is important as prices and discounts are key to influence hotel buying decision of customers Factors influencing last hotel decision for hotel stayers
Price is the top factor influencing hotel booking decision
~ 60% of customers said Price influenced their hotel decision
63%
Price
61%
Location of Hotel Property
32%
Previous experience
Hotel brand
30%
Amenities
29%
Special offer/discount
29% 26%
Positive hotel reviews
23%
Hotel pictures
Hotel star rating
21%
Loyalty membership
20% 14%
Restaurants in the hotel
Special offers/discounts also has a large role in influencing decision
Recommendations from friends/family
13%
Hotel size
12%
Recent renovation
4%
Question: “What factor influenced your last hotel decision? Select all that apply” Source: Phocuswright (Dis)Loyalty and the US Leisure Traveler
5
1
Purchase device fencing: target mobile customers who book last minute
1 in 4 hotel room nights are booked on a mobile device (and increasing!)
~ 1 in 2 mobile customers book same day/day before their trip 2015 Advance booking window for domestic US travel
2015 Forecasted YoY growth in US digital travel sales
40.7% 53%
1+ days
80%
5.5% 47%
-2.2%
0-1 days
20% Desktop
Desktop/ Laptop
Mobile
Source: Hotels.com data
Mobile
Source: eMarketer, April 2014, October 2014
6
Overall
2
Customer origin fencing: target international travelers while protecting domestic pricing
Foreign exchange rates can impact demand patterns
International customers book earlier and stay longer 2015 Booking window (months) for US hotels
GBP, EUR and CAD/USD exchange rate (Index, July 2014=100)
27% 100 Domestic customers International customers
8% 6% 13% 27%
13%
3+ mo
GBP
73% 19%
90
Jul ’14
< 1 mo
58%
Source: Hotels.com data
EUR
80
42%
2-3 mo 1-2 mo
CAD
Sep ’14
Nov ’14
Jan ’15
Mar ’15
May ’15
US Hotels Growth rate (Indexed to July 2014 = 100) 2015 Length of stay (days) in US hotels Domestic customers
120
80
+1.2 days International Customers
US->US
100
1.8
Int’l -> US
60
3.0
Jul ’14
Source: Hotels.com data
Sep ’14
Source: Hotels.com data
7
Nov ’14
Jan ’15
Mar ’15
3
Customer loyalty fencing: target loyalty program members who travel and comparison shop frequently
The loyalty traveler population is vast and they travel frequently • •
Loyalty members are savvy travelers looking for best deals •
1 in 2 hotel stayers and 1 in 3 OTA shoppers are loyalty members Loyalty participation increases with trip frequency so they are more likely frequent travelers
•
US Loyalty traveler population (M) and program membership (% of segment)
Loyalty members are more likely to shop across multiple websites than non loyalty members 61% of elite program members said they’d check multiple websites to get a good deal Online shopping behavior by loyalty status “When planning travel, I always check multiple websites to make sure I am getting a good deal”
118M Non-members
55M 56M (47%)
Loyalty members
No hotel loyalty
42%
Entry level/ Mid-tier
41%
Elite
18M (33%)
Hotel Stayers OTA shoppers
36% 41% 61%
Strongly agree
Neutral/Unsure
Slightly agree
Slightly disagree
25% Strongly disagree
Source: Phocuswright (Dis)Loyalty and the US Leisure Traveler
Source: Phocuswright (Dis)Loyalty and the US Leisure Traveler
8
Linda Gulrajani, CRME VP, Revenue Strategy & Distribution Marcus Hotels & Resorts Member of HSMAI’s Revenue Management Advisory Board 9
Scenario 1: Branded Strategy (Hilton) Target Customer Segment: Past transient guests who don’t already have a reservation and did not book a negotiated rate.
Fence & Method: $79 rate with a calendar that had the dates with that rate available highlighted (had other rates available for all dates) Sent out one email campaign that was bookable for 7 days and covered a 6 week booking window and then sent another one when there was only one day left with the same offer.
Success: Booked $27k in revenue The second “one day left” email out performed the first email
Tracking: Through the Clairvoyix and Hilton.com 10
Scenario 2: Resort Strategy Target Customer Segment: Transient guests who have booked through a discount channel like Groupon, Travelzoo, etc. in the past that don’t have a future reservation (Very important to collect email addresses from these guests).
Fence & Method: Email 48 hour sale promoting “book now and get the lowest rates of the year” sent out 1 week prior to the Groupon/Travelzoo offer going out
Success: Booked 1,030 rooms through this offer, which decreased the next “special offer” production significantly, saving us the commission and capturing them direct
Tracking: With a special rate code
11
Scenario 3: OTA Strategy Target Customer Segment: Fenced guests (or members) on OTAs
Fence & Method: Discount for members only
Success: 30% - 40% of the OTA bookings are coming from this channel Improved placement on the site This requires that a guest creates a login, but there are no other fences that prevent them from seeing the special pricing so is this really a fenced offer?
Tracking: Reports provided by market managers
12
Kathleen Cullen, CRME SVP, Revenue & Distribution Commune Hotels & Resorts Chair of HSMAI’s Revenue Management Advisory Board 13
Scenario 1: Goal: Fill Suite Occupancy without public discounting Targeted Customer Segment: Guests with stays more than once and have paid an ADR range of $400-$2500.
Fence: Pull qualified customer list from CRM
Method: Via Email Communication or private mailing with an invitation to stay in one of the remodeled luxury suites. Tracking: Email Campaign Statistics, Rate Code Production and if bookable via a landing page on brand.com Google Analytics
14
15
16
Scenario 2: Goal: Build occupancy base outside the 45 day booking window. Customer Segment: OTA International Customers
Fence: Unique POS (Point of Sale) + include a cancel policy and/or MLOS restriction
Method: Via OTA partners using Unique POS (ie: UK, Brazil, etc.)
Tracking: OTA reports or set up a specific Rate Code to track all bookings.
17
Example: Extranet Results
18
Scenario 3: Goal: Increase Website Conversions Customer Segment: Direct Bookers
Fence: Guests that are within Shopping Cart Funnel
Method: Show LightBox to potential guest that has checked rates & avail but closes out booking engine. Email trigger to potential guest that started to make their reservation but does not complete.
Tracking: Booking Engine Reporting + Partner Reporting
19
Email Abandonment
20
Booking Engine Pop-up
21
May Results # of Abandoned Carts
4,819
# of Guests we were able to invite back
548
# of Reservations Recovered
26
Reservation Value
$ 502.72
Revenue
$13,070.75 *
Cost
$392.12
ROI
$33:1 * 15% of Pre-Opening Website Revenue
22
Scenario 4: Goal: Drive occupancy on select need dates Customer Segment: Local Residents
Fence: Live in specific zip codes
Method: Serve Paid Search Text and Display Advertising to shoppers with a Florida Zip Code Dedicated Email Blast to Thompson Hotel’s email database with a Florida address Load Florida Resident Rate to a Booking Engine Filter
Tracking: Paid Search Reports, Email results + Google Analytics
23
Paid Search
24
Email Blast
25
Results April – June MTD, 2015 Room Nights
161
Reservations
71
LOS
2.3
ADR
$253.13
Revenue
$40,754.85
Cost
$7,797.71
ROI
$5:1
26
