U.S. patent application number 12/803395 was filed with the patent office on 2010-10-28 for conditional electronic coupon distribution method and system.
Invention is credited to Kin K. Leung, Hui Luo, Nemmara K. Shankaranarayanan.
Application Number | 20100274648 12/803395 |
Document ID | / |
Family ID | 21703587 |
Filed Date | 2010-10-28 |
United States Patent
Application |
20100274648 |
Kind Code |
A1 |
Leung; Kin K. ; et
al. |
October 28, 2010 |
Conditional electronic coupon distribution method and system
Abstract
A conditional e-coupon distribution method distributes e-coupons
predefined by sellers to mobile users only if the number of mobile
users requesting such e-coupons equals or exceeds a threshold. The
method receives a request to browse e-coupons from a mobile user.
The method receives the location of the mobile user and determines
a plurality of sellers local to the mobile electronic device and a
plurality of corresponding e-coupons available from the local
sellers. The method receives a request for a particular e-coupon
from a seller and authorizes the provision of said e-coupon to the
mobile user. At the end of a processing cycle, the mobile user
receives the requested e-coupon if all conditions, such as a period
of time and threshold, have been met. A computer-usable medium
having computer-readable program code embodied therein allows for
storage of the method.
Inventors: |
Leung; Kin K.; (Edison,
NJ) ; Luo; Hui; (Marlboro, NJ) ;
Shankaranarayanan; Nemmara K.; (Bridgewater, NJ) |
Correspondence
Address: |
AT & T Legal Department - Lazaroff
Attn: Patent Docketing, Room 2A-207, One AT & T Way
Bedminster
NJ
07921
US
|
Family ID: |
21703587 |
Appl. No.: |
12/803395 |
Filed: |
June 25, 2010 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
12214285 |
Jun 18, 2008 |
7769634 |
|
|
12803395 |
|
|
|
|
11245635 |
Oct 7, 2005 |
7418451 |
|
|
12214285 |
|
|
|
|
10002999 |
Nov 2, 2001 |
6996579 |
|
|
11245635 |
|
|
|
|
Current U.S.
Class: |
705/14.13 ;
707/802; 707/E17.005 |
Current CPC
Class: |
G06Q 30/0219 20130101;
G06Q 30/0257 20130101; G06Q 30/0207 20130101; Y10S 707/99942
20130101; G06Q 30/0247 20130101; G06Q 30/0249 20130101; Y10S
707/99948 20130101; Y10S 707/99943 20130101; G06Q 30/0261 20130101;
G06Q 30/02 20130101; G06Q 30/0223 20130101; G06Q 30/0211 20130101;
Y10S 707/99945 20130101 |
Class at
Publication: |
705/14.13 ;
707/802; 707/E17.005 |
International
Class: |
G06Q 30/00 20060101
G06Q030/00; G06F 17/30 20060101 G06F017/30 |
Claims
1. An e-coupon distribution method for distributing e-coupons
redeemable for value in the purchase of goods or services from a
seller, comprising: receiving a request from a mobile electronic
device of a user for at least one e-coupon from a seller local to
the location of the mobile electronic device; determining an
initial threshold number of user requests for the seller's e-coupon
by making a calculation based at least in part on the number of
pending requests, the seller's profit margin for the goods or
services associated with the e-coupon, the fee associated with the
distribution of the e-coupon, and estimated rates of e-coupon
redemption, so that the threshold number of distributed e-coupons
will provide the minimum estimated number of redemptions necessary
to make expected returns from redeemed e-coupons greater than the
fee associated with the distribution of the e-coupon; and sending
the requested e-coupon to the mobile electronic device if the
number of pending requests for said e-coupon is greater than or
equal to the initial threshold.
2. The e-coupon distribution method of claim 1 wherein the fee
associated with the distribution of the e-coupon is an advertising
fee charged for the e-coupon.
3. The e-coupon distribution method of claim 2 wherein said initial
threshold number is determined by dividing the advertising fee by a
product of the estimated rate of e-coupon redemption times the
seller's profit margin for the goods or services associated with
the e-coupon.
4. The e-coupon distribution method of claim 1 further comprising:
calculating an initial threshold number for requests for e-coupons
for additional sellers, determining the sellers having pending
numbers of requests for an e-coupon that are not greater than or
equal to the initial threshold calculated for them, and applying an
optimizing consolidation process to all sellers with insufficient
pending requests, including eliminating at least one such seller to
obtain increased estimated rates of e-coupon redemption,
recalculating using such increased estimated rates of e-coupon
redemption to determine an optimized threshold number, and sending
the requested e-coupon to the mobile electronic device if the
number of pending mobile requests for said e-coupon is greater than
or equal to the optimized threshold number.
5. The e-coupon distribution method of claim 4 further comprising,
when the number of pending requests for said e-coupon is determined
to be less than the optimized threshold number, the additional step
of sending a notification to the mobile user that no e-coupon for
the seller will be provided.
6. The e-coupon distribution method of claim 1 further comprising
maintaining mobile user profile data and seller profile data in
database memory; maintaining in the mobile user profile a quota of
e-coupons the user is allowed to receive; determining whether an
e-coupon sent to the mobile user has been redeemed; and upon
determining said e-coupon has been redeemed, increasing the quota
of e-coupons by one in the mobile user profile.
7. An e-coupon distribution system for distributing e-coupons
redeemable for value in the purchase of goods or services from a
seller, comprising: a server to receive a request from a mobile
electronic device of a user for at least one e-coupon from at least
one seller local to the location of the mobile electronic device;
said server determining an initial threshold number of user
requests for the seller's e-coupon by making a calculation based at
least in part on the number of pending requests, the seller's
profit margin for the goods or services associated with the
e-coupon, the fee associated with the distribution of the e-coupon,
and estimated rates of e-coupon redemption, so that the threshold
number of distributed e-coupons will provide the minimum estimated
number of redemptions necessary to make expected returns from
redeemed e-coupons greater than the fee associated with the
distribution of the e-coupon; and said server sending the requested
e-coupon to the mobile electronic device if the number of pending
requests for said e-coupon is greater than or equal to the initial
threshold.
8. The e-coupon distribution system of claim 7 wherein the fee
associated with the distribution of the e-coupon is an advertising
fee charged for the e-coupon.
9. The e-coupon distribution system of claim 8 wherein said initial
threshold is determined by dividing the advertising fee by a
product of the estimated rate of e-coupon redemption times the
seller's profit margin for the goods or services associated with
the e-coupon.
10. The e-coupon distribution system of claim 7 further comprising:
said server calculating an initial threshold number for requests
for e-coupons for additional sellers, said server determining the
sellers having pending numbers of requests for an e-coupon that are
not greater than or equal to the initial threshold calculated for
them, and said server applying an optimizing consolidation process
to all sellers with insufficient pending requests, including
eliminating at least one such seller to obtain increased estimated
rates of e-coupon redemption, recalculating using such increased
estimated rates of e-coupon redemption to determine an optimized
threshold number, and said server sending the requested e-coupon to
the mobile electronic device if the number of pending mobile
requests for said e-coupon is not greater than or equal to the
initial threshold calculated for them, but is greater than or equal
to the optimized threshold number.
11. The e-coupon distribution system of claim 10 further
comprising, when the number of pending requests for said e-coupon
is determined to be less than the optimized threshold number, the
server sending a notification to the mobile user that no e-coupon
for the seller will be provided.
12. The e-coupon distribution system of claim 7 further comprising
said server maintaining mobile user profile data and seller profile
data in database memory; said server maintaining in the mobile user
profile a quota of e-coupons the user is allowed to receive; said
server determining whether said e-coupon sent to the mobile user
has been redeemed; and upon determining said e-coupon has been
redeemed, said server increasing the quota of e-coupons by one in
the mobile user profile.
Description
CROSS-REFERENCE TO RELATED APPLICATIONS
[0001] This application is a continuation of Ser. No. 12/214,285,
filed Jun. 18, 2008, which issued as U.S. Pat. No. ______ on
______, which was a continuation of Ser. No. 11/245,635, filed Oct.
7, 2005, which issued as U.S. Pat. No. 7,418,451 on Aug. 26, 2008,
which was a continuation of Ser. No. 10/002,999, filed Nov. 2,
2001, entitled E-COUPON SERVICE FOR LOCATION-AWARE MOBILE COMMERCE
WHICH DETERMINES WHETHER TO SUPPLY REQUESTED E-COUPONS BASED ON THE
NUMBER OF REQUESTS RECEIVED IN A PROCESSING CYCLE, AND A THRESHOLD
NUMBER OF REQUESTS REQUIRED TO MAKE EXPECTED RETURNS FROM REDEEMED
COUPONS GREATER THAN ADVERTISING FEES, which issued as U.S. Pat.
No. 6,996,579 on Feb. 7, 2006.
FIELD OF THE INVENTION
[0002] The present invention relates to electronic coupon
distribution and, more particularly, to conditional electronic
coupon distribution to a mobile user.
BACKGROUND OF THE INVENTION
[0003] Today, there exist a few techniques for distribution of
electronic coupons (e-coupons) to potential customers. A seller of
a service or product can distribute mass email messages with
e-coupons attached to or within the message. The seller then hopes
that at least some of the potential customers who receive the
messages will redeem the e-coupons included. Additionally, a seller
can post an e-coupon on a web site, whether or not owned by the
seller, and hope that a potential customer will see the e-coupon
and choose to redeem it. Similar techniques exist for targeted
distribution of coupons to attract potential customers who live
local to a seller. For example, local restaurants or stores can
attempt to attract customers on web pages local to a particular
city. A city newspaper may have a website and these local sellers
can post e-coupons on the site with the same hope of a customer
seeing the e-coupon and choosing to redeem it.
[0004] Yet these and other techniques tend to be expensive to
sellers and highly inefficient in attracting potential customers.
The mass emails and e-coupon distribution carry a fixed cost based
upon a hope that a potential customer exists. A seller must pay the
website provider in order to put its e-coupons on the website,
without any assurance at all that the e-coupons will be requested
and redeemed by customers. Even existing location-aware wireless
targeted advertising methods have a cost to transmit an
advertisement when a mobile customer is around a particular store.
A need exists to allow for distribution of an e-coupon when a
desired profit margin for the seller has been statistically
guaranteed.
[0005] Generally, potential customers choose a particular type of
service or product and then seek to find the service or product
based upon some need, whether such need is based on cost, ease of
acquisition, location, or some other factor. Customer service can
be enhanced if an e-coupon service were provided based upon an
initial inquiry from the customer. In such an example, a seller has
a substantially increased opportunity to ensure that an e-coupon is
redeemed because the customer initiates the desire for the e-coupon
rather than a random e-coupon being sent to a customer.
[0006] However, a seller needs additional safeguards to ensure that
distributed e-coupons will be redeemed not only at a high
probability, but also at an economic advantage to the seller. A
seller wants to issue e-coupons so that a larger number of
potential customers will redeem the e-coupons, because the seller
has no economic incentive when one or a few customers actually
redeem the e-coupons. Additionally, a seller needs to ensure that
the distribution of e-coupons does not become economically
ineffective by continually distributing e-coupons that are not
redeemed.
[0007] In addition to business concerns, there are privacy concerns
and technical issues often left unresolved in wireless targeted
advertising and distribution systems. Privacy concerns have emerged
from FCC requirements that wireless service providers must be able
to identify the locations of cellular phone users making emergency
calls after Oct. 1, 2001. The concern stems from the worry that
wireless service providers will turn the FCC requirement into a
source of profit. In particular, the wireless service providers
could release the location information of mobile users to anyone
willing to pay for such information, including individuals or
others who will mishandle this information.
[0008] Technical issues also exist from the standpoint of wireless
service providers. In order for a participating seller to send a
targeted advertisement to mobile users at a right time and in a
right place, the location information of the mobile users must be
accurate and monitored closely. To do so, all mobile devices must
frequently send location data to the network, and thus cause
significant uplink signaling traffic that may eventually overflow
the network. The situation could get even worse if the location
identification methods are network-based or network-assisted
because they consume significant computing resources from the
network.
[0009] Therefore, a need exists to allow for distribution of
e-coupons to potential customers based upon an initial desire from
the customer and an assurance for a seller to make extra profits in
a statistical sense. Privacy concerns and technical issues also
need to be addressed in order to promote location-aware mobile
commerce in an optimal manner.
BRIEF SUMMARY OF THE INVENTION
[0010] The present invention provides a conditional e-coupon
distribution system for a mobile network that distributes e-coupons
predefined by sellers to mobile electronic devices of users only if
the number of mobile users requesting such e-coupons equals or
exceeds a threshold. Although the invention is primarily shown and
described in conjunction with a cellular telephone network
distributing e-coupons, it is understood that the invention is
applicable to other network types in which it is desirable to
distribute e-coupons.
[0011] In one embodiment, the invention operates in a mobile
communications system based on multiple cells. The invention
comprises a profile database, a profile proxy server, and a
commerce server. The system receives a request to browse at least
an e-coupon from a mobile user. The system receives the current
location of the mobile user from the mobile user and determines a
plurality of sellers local to the mobile user as well as a
plurality of the corresponding e-coupons for those sellers. Upon
receiving a request for a particular e-coupon from the mobile user,
the invention authorizes whether the e-coupon should be distributed
to the mobile user.
[0012] The invention operates according to a processing cycle. At
the end of a processing cycle, the mobile user receives the
requested e-coupon if all conditions, such as a period of time and
threshold, have been met. Only if the threshold number of mobile
users requesting a particular e-coupon has been met or exceeded is
an e-coupon, from a particular seller, distributed.
[0013] The threshold value is calculated at the end of the
processing cycle. The threshold value is then compared to the
number of pending requests in the profile of the particular seller
in question. Only upon determining that the number of pending
requests is greater than or equal in value to the threshold value
are the e-coupons distributed. The use of a threshold value and
authorization to distribute are two examples of how the invention
ensures a seller will make extra profits in a statistical
sense.
[0014] The invention can improve the advertising effectiveness and
cost-efficiency, because (1) e-coupons are only distributed to
mobile users who explicitly request such e-coupons; (2) e-coupons
for a participating seller are distributed only if the number of
mobile users who requested such e-coupons equals or exceeds a
threshold; (3) a participating seller is assured to make extra
profit in a statistical sense by adopting the threshold calculated
by the conditional e-coupon service provider; and (4) a
participating seller is charged an advertising fee by the
conditional e-coupon service provider only after its e-coupons are
distributed.
[0015] A feature that differentiates the conditional e-coupon
service from conventional advertising methods is that the
conditional e-coupon service neither pushes advertisement messages
to mobile users nor does it arbitrarily send (costly)
advertisements. Once a seller is advertised using the conditional
e-coupon service, the threshold value assures a sufficiently large
mobile user base from which a seller can make a profit in a
statistical sense after paying an advertising fee to the
conditional e-coupon service provider.
[0016] The conditional e-coupon service concept also includes a
quota system that strikes a good balance between giving mobile
users sufficient freedom not to redeem received e-coupons for
whatsoever reasons and preventing irresponsible mobile users from
sending a large volume of e-coupon requests when the mobile users
have no intention to redeem any of the received e-coupons.
[0017] In addition, the present invention can eliminate the
technical problem for wireless targeted advertising because it is
operated based on requests made by mobile users. The accurate
location information of mobile users is conveniently imbedded in
these requests, and therefore mobile users do not need to send
periodical location update messages to wireless networks. The
invention also eliminates any privacy concerns, because it does not
need to track the location of mobile users and it only supplies
participating sellers with statistical information of a group of
mobile users (such as how many mobile users have requested
e-coupons from the seller) instead of information of individual
mobile users.
[0018] In view of the present invention, it can be seen
qualitatively that a conditional e-coupon service expects to
receive welcome from all involved parties. Mobile users save money
right before they go shopping. The invention is particularly useful
to a group of mobile users planning to visit the same seller.
Participating sellers find the conditional e-coupon service a
cost-efficient targeted e-coupon distribution scheme, because
e-coupons are distributed to mobile users only if e-coupons can
bring enough estimated redeems that can make extra profits for the
seller. The conditional e-coupon service provider can earn
advertising fees from the participating sellers. Finally, the
wireless service providers may find they can expand customer base
by associating with the conditional e-coupon service provider
because both mobile users and participating sellers need wireless
data services to utilize the distribution service.
BRIEF DESCRIPTION OF THE DRAWINGS
[0019] FIG. 1 is a schematic diagram of an exemplary conditional
e-coupon distribution system in accordance with the present
invention.
[0020] FIG. 2 is a further schematic diagram of an exemplary
conditional e-coupon distribution system in accordance with the
present invention.
[0021] FIG. 3 is a schematic diagram of a particular embodiment of
a conditional e-coupon distribution system in accordance with the
present invention.
[0022] FIG. 4 is a pictorial representation of a particular
embodiment of a conditional e-coupon distribution system in
accordance with the present invention showing user movement.
[0023] FIG. 5 is a block diagram of an exemplary commerce server
that can form a part of a conditional e-coupon distribution system
in accordance with the present invention.
[0024] FIG. 6A is a block diagram of an exemplary profile database
that can form part of a conditional e-coupon distribution system in
accordance with the present invention.
[0025] FIG. 6B is a further block diagram of an exemplary profile
database that can form part of a conditional e-coupon distribution
system in accordance with the present invention.
[0026] FIG. 7 shows an exemplary sequence of steps for responding
to an e-coupon request in a conditional e-coupon distribution
system in accordance with the present invention based upon a
processing cycle.
[0027] FIG. 8 shows another exemplary sequence of steps for
responding to an e-coupon request in a conditional e-coupon
distribution system in accordance with the present invention based
upon a processing cycle.
[0028] FIG. 9 shows further an exemplary sequence of steps for
authorizing whether an e-coupon should be distributed in accordance
with the present invention.
[0029] FIG. 10 shows a conditional e-coupon profit versus a
threshold curve.
[0030] FIG. 11 shows a maximum conditional e-coupon profit versus
an advertising price curve.
[0031] FIG. 12 shows a maximum conditional e-coupon revenue versus
advertising price curve.
[0032] FIGS. 13A through 13E show an example of how a mobile user
might interact with the present invention.
DETAILED DESCRIPTION OF THE INVENTION
[0033] FIGS. 1 and 2 show a wireless mobile communication system
100 having conditional e-coupon distribution in accordance with the
present invention. In general, the system sends e-coupons to mobile
users based upon user location and conditions specified by users
and sellers. This arrangement provides an efficient mechanism for
suppliers of goods and services, i.e., sellers, to distribute
e-coupons to mobile users who wish to receive an e-coupon from a
particular seller.
[0034] In one embodiment, the mobile communication system 100
includes a plurality of cells 102a-d each served by a respective
base station 104a-d in a manner well known to one of ordinary skill
in the art. Each of the base stations 104a-d can be coupled to a
respective commerce server 106a-d for providing messaging
instructions to the base station as described in detail below. The
commerce server 106 can be connected to a profile proxy server
(PPS) 108 via a network 110, such as the Internet or intranet. A
plurality of mobile users 202a-N and sellers 204a-M (FIG. 2) can
communicate with the profile proxy server 108 via the Internet 110,
for example. The profile proxy server 108 can send the provided
information to a commerce server 106 that is local to the mobile
user for storage in a database.
[0035] FIG. 3 shows one particular embodiment of a conditional
e-coupon distribution system 300 in accordance with the present
invention that can be coupled to a General Packet Radio Service
(GPRS) network 350. The conditional e-coupon distribution system
300 includes a profile proxy server 108 coupled to a commerce
server 106 and to a profile database 107, which is also coupled to
the commerce server 106. As described below, the profile database
107 can store profile data for mobile users and sellers associated
with the mobile network. The profile proxy server 108 is coupled to
the Internet 110 via a conventional gateway 111. In one embodiment,
one profile proxy server 108 can support a plurality of commerce
servers 106 throughout the conditional e-coupon distribution
system. Further, the profile database 107 could comprise multiple
databases for storage of data profiles of mobile users and
sellers.
[0036] The GPRS network 350 includes a Serving GPRS Support Node
(SGSN) 352 coupled to a local commerce server 106 and to a Gateway
GPRS Support Node (GGSN) 354. The SGSN 352 communicates with a base
station 104 covering the local cell 102 for providing mobile
service to users 202 within the cell. The commerce server 106
provides e-coupon information to the local SGSN 352 for
transmission by the associated base station 104.
[0037] As shown in FIG. 4, mobile users 202a-N move within, into,
and out of different cells 102 within the network. Sellers 204a-M
want to contact potential users 202 that have expressed a desire
for their type of goods or services. In general, sellers 204 wish
to identify users proximate their location to maximize the
likelihood that a user will visit the seller. As described below,
the conditional e-coupon distribution system of the present
invention can transmit e-coupons that are requested by users 202
proximate a particular seller 204 provided that the profile
conditions, e.g., type of goods, threshold, time, location, are met
by the request. The e-coupon identifies the seller and allows the
mobile user to locate and/or contact the seller. Sellers specify
conditions, e.g., range, time, number of mobile users, under which
the e-coupons are distributed, as described below.
[0038] As described above in connection with FIGS. 1-3, mobile
users 202 provide profile information to a local commerce server
106 under the control of the profile proxy server 108. In one
embodiment, the mobile user 202 can provide profile information via
the Internet 110 to maximize user convenience. The profile proxy
server 108 ensures that the mobile user profile information is
stored in a profile database 107 associated with a commerce server
106 that is local to the mobile user's base location, e.g., home
address. As described in detail below, the base station 104
covering the mobile user's current location transmits e-coupon
information that is received by mobile users.
[0039] In general, sellers specify conditions for distributing
e-coupons. Exemplary conditions include distributing their
e-coupons only when a predetermined number of mobile users have
requested an e-coupon of the seller, at certain times of the day,
and certain days of the week. In addition, the seller profile can
contain driving directions to the seller and other e-coupons
available from the seller. The seller profile can further include
periods of time desired for e-coupon distribution, mobile user
distance range, and type of e-coupon. The seller conditions can be
stored in a seller profile database associated with a commerce
server local to the seller.
[0040] In one embodiment, a seller can manually determine the
number of mobile users within the seller's local area that have
requested the seller's e-coupon by connecting to the proxy server.
For example, a seller can communicate with a local commerce server
that provides mobile user information to the seller. The seller can
then manually distribute e-coupons to mobile users who have
requested the seller's e-coupon. In this example, a seller has the
option to allow the e-coupons to be distributed to users who have
requested the e-coupon although seller conditions, such as time of
day, range, etc., have not been satisfied. Thus the seller can
distribute the e-coupons without identifying a particular mobile
user. Therefore, user privacy can be maintained by protecting the
actual identity of the mobile user.
[0041] FIG. 5 shows an exemplary embodiment of a commerce server,
such as the commerce server 106 of FIGS. 1 and 2, that forms part
of a conditional e-coupon distribution system in accordance with
the present invention. In one embodiment, the commerce server 106
includes an instant message server 501 for generating instant
messages to a mobile user in response to an active request for
information from a mobile user. The instant message server 501 can
also include the mobile user in a group of mobile users to receive
e-coupons in response to an inquiry.
[0042] The commerce server 106 can further include a mobile user
location monitor 502 for monitoring the location of mobile network
users. As described more fully below, the mobile user's location
can be used to distribute e-coupons requested by the mobile user. A
multicast message gateway 503 delivers e-coupons to a selected
group of mobile users via a GPRS network in a multicast format.
Alternatively, the e-coupons can be distributed using conventional
Short Message Service (SMS) or Cellular Digital Packet Data (CDPD)
email services.
[0043] The commerce server 106 can further include a profile
database 504 for storing mobile user and seller profiles. Mobile
users and sellers can modify their profile information via the
profile proxy server 108 through the Internet 110.
[0044] In one embodiment, the mobile user and seller profiles are
stored on the commerce server 106 that is local to the respective
mobile user or seller. The profile proxy server 108 can contain a
user-message server index. With this arrangement, in the case where
a mobile user is not within the area served by the commerce server
containing the mobile user's profile, the profile proxy server can
be queried by the commerce server in which the mobile user is
currently located to obtain the mobile user's profile, as described
more fully below.
[0045] Referring to FIG. 6A, a block diagram is shown depicting an
exemplary profile database, such as the profile database 107 of
FIG. 3 that forms part of a conditional e-coupon distribution
system in accordance with the present invention. In one embodiment,
the profile database 107 includes first, 602, and second, 603,
tables that store all mobile user profiles 612a-e and all seller
profiles 613a-e respectively. The tables are indexed by a mobile
user identification and a seller identification respectively.
[0046] Referring to FIG. 6B, a block diagram is shown depicting an
exemplary profile database, such as the profile database 107 of
FIG. 3, further defining information found within a particular
mobile user profile 612a, and seller profile 613a. In one
embodiment, the mobile user profile 612a includes: a mobile user
identification field 622a; a mobile user instant contact address
field 622b; an e-coupon quota field 622c; and a number of effective
requests field 622d. Other fields 622e could include information
such as a percentage of e-coupon redemptions for the mobile user, a
password for changing account information, an option to allow for
automatic notification of e-coupons from a particular seller when
the mobile user is local to the seller, and a listing of past
purchases using the conditional e-coupon distribution service.
[0047] The mobile user instant contact address field 622b could
comprise a mobile phone number, a Short Message Service (SMS)
address, an instant message address, an email address, or a mobile
email address belonging to a mobile phone or other mobile device
capable of supplying location information.
[0048] The e-coupon quota field 622c comprises the number of
e-coupons the mobile user can possibly receive during a quota reset
cycle, a predefined period of time. At the first time when a mobile
user uses the conditional e-coupon system or at the beginning of a
quota reset cycle, i.e., week, month, a value is assigned to the
e-coupon quota field 622c in the mobile user profile 612a. After
the mobile user requests an e-coupon, the e-coupon quota field is
decreased by a value of one in the profile of the mobile user.
Should the mobile user not receive the desired e-coupon, or if the
mobile user receives the e-coupon and redeems it, the e-coupon
quota field is increased by a value of one.
[0049] If the e-coupon quota field reaches a value equal to zero,
the mobile user cannot request any e-coupons, from any seller,
until the next quota reset cycle, i.e., next week, month. This
safeguard is introduced in the profile of the mobile user to
prevent an irresponsible mobile user from abusing the conditional
e-coupon service, because a fundamental assumption for the
conditional e-coupon service is that a mobile user will redeem at
least one of the e-coupons received upon a request at a very high
probability. The e-coupon quota field 622c allows for mobile user
freedom in obtaining a number of e-coupons at any one time,
rewarding mobile users that redeem e-coupons, and protection of the
interests of sellers.
[0050] The number of effective requests field 622d is the number of
requests that result in receiving e-coupons. At the end of every
quota reset cycle, an e-coupon effectiveness ratio for every
customer can be estimated from the remaining quotas and the number
of effective requests. This e-coupon effectiveness ratio is used to
optimize the conditional e-coupon service operation and is
described below.
[0051] In one embodiment, the seller profile 613a includes: a
seller identification field 633a; a seller name field 633b; a
seller address field 633c; a seller billing information field 633d;
a seller instant contact address field 633e, a seller business type
field 633f; a number of allocated e-coupons field 633g; a type of
e-coupon field 633h; a number of pending requests field 633i; a
number of pending redeems field 633j, and an estimated number of
redeems field 633k. Other fields 633l could include information
such as driving directions to the seller, a password for changing
account information, and a percentage of e-coupon redemptions.
[0052] The seller instant contact address field 633e could comprise
a phone number, an instant message address, an email address, or
any other method by which the seller can be notified that an
e-coupon of the seller has been distributed. The seller business
type field 633f is described by a set of keywords, which can be
used in the case where a mobile user enters a keyword instead of a
seller name to request e-coupons.
[0053] In the seller profile, the seller also needs to predefine
one or more e-coupons. Every e-coupon is associated with some
conditions, including time (when e-coupons can be distributed),
range (where e-coupons should be distributed), and a
pre-advertising profit margin (how much profit the seller can make
from a mobile user redeeming the e-coupon). If multiple e-coupons
are predefined, only one e-coupon is ready to be issued at a given
time. That is, the time conditions are exclusively against each
other. A type of e-coupon field 633h could comprise the time
condition, the range condition, and the profit margin condition
defined above.
[0054] The time condition for an e-coupon can be a scheduled period
that remains the same throughout a defined period. For example, the
time condition could be a five-minute processing cycle that
operates consecutively, that is every five minutes. Alternatively,
the time condition could be a five-minute processing cycle that
only occurs between noon and midnight. Processing cycle times and
windows for distribution can vary and are not limited by the
examples used herein.
[0055] The range condition for an e-coupon defines the boundaries
for distribution. For example, the range may be any mobile user
requesting the e-coupon that is within the same cell as the seller.
Alternatively, the range may be within five miles of the seller.
Finally, the pre-advertising profit margin is a parameter from
which the conditional e-coupon system can calculate a threshold
value for the seller and therefore can determine whether the
e-coupon of the seller should be distributed to the mobile users
who requested the e-coupon within the preceding processing cycle.
The pre-advertising profit margin is the difference between the
price and the cost of the service/goods promoted in the
e-coupon.
[0056] FIG. 7, in combination with FIGS. 1-3, shows an exemplary
sequence of steps for responding to an e-coupon request in a
conditional e-coupon distribution system in accordance with the
present invention. In general, a local commerce server receives a
request for e-coupons, and upon receiving the request, the system
authorizes whether the e-coupon should be distributed. A processing
cycle begins in step 701. In step 702, it is determined whether the
system has received a request for e-coupons. Such a request could
be defined by keywords, such as "grocery store" or "fast food."
Alternatively, the request could be to browse for all e-coupons for
a particular seller by the seller name. For this example, assume
that the mobile user enters "fast food within one mile." If the
system receives a request, the location of the mobile device is
received with the request in step 703. The methods for determining
the location of a mobile device are well known in the art. An
exemplary embodiment entails the local commerce server receiving
the current location of the mobile user in connection with the
request to browse because the mobile electronic device
automatically inserts the location data as a header parameter in
the outgoing request to browse message.
[0057] After receiving the request for e-coupons, the conditional
e-coupon profile proxy server, in step 704, immediately sends a
reply message back to the mobile user, which shows, for example,
the remaining quota and estimated time to receive e-coupons (it is
possible that the mobile user may not receive any e-coupons). The
conditional e-coupon profile proxy server then passes the request
message to the conditional e-coupon commerce server, along with the
request time, the location of the mobile user, and the instant
contact address of the mobile user.
[0058] Consequently, the commerce server looks within the profile
database for all sellers that have "fast food" in their
service/goods descriptions. As in step 705, once a match is found,
the commerce server first checks the timing and range conditions
set by the seller. If these conditions permit e-coupon
distribution, the commerce server then compares the seller location
with the mobile user location and determines if the "one mile"
requirement can be satisfied. If this requirement is satisfied, the
seller is declared as a candidate seller for the request, and the
request is logged as a pending request to the seller. After all
candidate sellers are identified, a redeem probability, which
equals the e-coupon effectiveness ratio for the mobile user divided
by the number of candidate sellers, is assigned to every pending
request. The e-coupon effectiveness ratio is estimated from
statistics collected in the previous quota reset cycle in the
mobile user profile. The reason why the number of candidate sellers
serve as a denominator in the redeem probability is that all
candidate sellers are competitors against each other and, thus, if
all candidate sellers decide to issue their e-coupons, the mobile
user will probably redeem only one of them. Note that the commerce
server does not immediately decide whether e-coupons should be
distributed. Instead, it makes such decision at the end of a
processing cycle, which could be about 5 to 10 minutes. Before a
processing cycle ends, the commerce server simply logs 706 every
incoming request as described above.
[0059] In step 707, a determination is made as to whether a
processing cycle is complete. After a processing cycle is finished,
in step 708, the commerce server authorizes whether the e-coupon
should be distributed. The commerce server checks every candidate
seller profile and distributes e-coupons for a seller if there are
enough pending requests in the seller profile. This process
consists of the following three steps.
[0060] (1) The commerce server estimates the overall e-coupon
effectiveness ratio, denoted as r, for every candidate seller using
the following formula:
r = 1 N n = 1 N r n c n ##EQU00001##
where N is the number of pending requests and
r n c n ##EQU00002##
is the redeem probability for the n.sup.th pending request to the
candidate seller. That is, r.sub.n is the e-coupon effectiveness
ratio for the mobile user whose request is the n.sup.th pending
request to the candidate seller, and c.sub.n is the number of
candidate sellers for this request.
[0061] (2) The commerce server calculates the best initial
threshold .theta. (explained in next section) based on the
pre-advertising profit margin, the overall e-coupon effectiveness
ratio, and the advertising fee for every candidate seller, which
statistically guarantees every candidate seller to make maximum
profit. If N.gtoreq..theta., which means the number of pending
requests equals or exceeds the initial threshold, the commerce
server distributes e-coupons using SMS to every mobile user who has
a pending request in the candidate seller profile.
[0062] (3) The commerce server has all sellers with N<.theta. go
through an optimal consolidation process (explained in next
section). The objective hereby is to maximize the number of
candidate sellers that can finally distribute e-coupons by
selectively removing candidate sellers (thus the e-coupon
effectiveness ratio for remaining candidate sellers can be improved
due to less competition and hence their initial thresholds can be
reduced to a new optimal threshold). This ensures that the
conditional e-coupon service provider can make maximum revenue and
a maximum number of e-coupon requests from mobile users can be
satisfied.
[0063] Once it is determined that an e-coupon is to be distributed,
the commerce server generates a serial number as well as a redeem
confirmation number for the e-coupon. The e-coupon is then
distributed with the serial number included to all mobile users who
requested it, and a notification message that includes the serial
number and redeem confirmation number is sent to the corresponding
candidate seller via the seller instant contact address.
[0064] After receiving the desired e-coupons, a mobile user can go
to the issuing candidate seller and redeem the e-coupon. The mobile
user has various methods to confirm she has redeemed the e-coupon
in order to get back her quota. One method is to request a redeem
confirmation number from the candidate seller for a sale/visit,
which is reported back to the profile proxy server by the mobile
user. Another method is to contact the profile proxy server using
the mobile user's wireless device when she is physically in the
issuing candidate seller's store. In this case, the seller location
is automatically submitted to the profile proxy server, which
serves as the evidence that the mobile user was attracted to the
candidate seller by the e-coupon. The quota system discussed
earlier provides an incentive to the mobile user to report back.
There can also be other methods to provide an incentive for such
reporting, such as the provision of discounts or bonus points.
[0065] FIG. 8 shows an alternative embodiment. A local commerce
server receives a request to browse at least an e-coupon, and upon
receiving a request for a particular e-coupon, the system
authorizes whether the e-coupon should be distributed. A processing
cycle begins in step 801. In step 802, it is determined whether the
system has received a request to browse at least an e-coupon. If
the system receives a request, the location of the mobile device is
received with the request in step 803. An exemplary embodiment
entails the local commerce server receiving the current location of
the mobile user in connection with the request to browse because
the mobile electronic device automatically inserts the location
data as a header parameter in the outgoing request to browse
message. In step 804, a plurality of sellers local to the location
of the mobile user and meeting the request criteria, such as for
"grocery" or a particular seller, is determined. In addition, a
plurality of the corresponding e-coupons available for each seller
is determined. The e-coupon availability data is subsequently
provided to the mobile user in step 805.
[0066] In step 806, it is determined whether the system has
received a request for a particular e-coupon from the set of
e-coupons provided in step 805. If the request for a particular
e-coupon is received in 806, the conditional e-coupon profile proxy
server, in step 807, immediately sends a response message back to
the mobile user, which shows the remaining quota and estimated time
to receive e-coupons (it is possible that the mobile user may not
receive any e-coupons). The conditional e-coupon profile proxy
server then passes the request message to the conditional e-coupon
commerce server, along with the request time, the location of the
mobile user, and the instant contact address of the mobile user. In
step 808, the request for the e-coupon is logged within the
applicable mobile user and seller profiles. A log of this request
is created because the system does not authorize distribution of
e-coupons until the end of the processing cycle. Such a processing
cycle could last for any predetermined amount of time and could
vary among sellers, particular types of e-coupons, or even among
varying e-coupons of a same seller.
[0067] In one embodiment, in step 808, to log a request received
from a mobile device, the conditional e-coupon system creates a
record in the profile of the mobile user, which contains the
request time and the IDs of all the sellers requested. Next, the
system will decrease the e-coupon quota field 622c in the mobile
user profile. The system then creates a record in the profile of
every seller, which contains the request time and the ID of the
mobile user. Finally, the system increases the estimated number of
redeems field 633k. The value of the increment could be chosen to
equal one divided by the number of sellers in the request.
Therefore, if a mobile user requested a particular e-coupon from a
particular seller name and if there were two stores of the seller
name local to the location of the mobile user, the estimated number
of redeems field 633k would be increased by 0.5 or one divided by
two sellers in the request.
[0068] In step 809, the system determines whether the processing
cycle has been completed. If the processing cycle has not been
completed, the system again returns to step 802 and determines
whether a request to browse at least an e-coupon has been received.
However, if the processing cycle has been completed, the system
authorizes whether the e-coupon should be distributed 810.
[0069] FIG. 9 shows the steps for the system to authorize whether
an e-coupon should be distributed and subsequent action by the
system. In step 901, the processing cycle has ended. In one
example, moving to step 902, the system determines whether the
number of pending requests in the seller profile is greater than or
equal to an initial threshold value. As disclosed above, the
initial threshold value is calculated at the end of the processing
cycle.
[0070] One method of calculation of the initial threshold is to
divide an advertising fee for the e-coupon by a product of the
estimated number of redeems field 633k in the seller profile times
a pre-advertising profit margin for the e-coupon. The initial
threshold value is then compared to the number of pending requests
633i in the profile of the particular seller in question, as shown
in step 902, to authorize whether the e-coupon should be
distributed.
[0071] Only upon determining that the number of pending requests is
greater than or equal in value to the initial threshold value do
the e-coupons get distributed, as shown in step 903, to mobile
users who requested the e-coupon in the preceding processing cycle.
If the system determines that the number of pending requests in the
seller profile is less than the initial threshold value, the
e-coupons are not distributed as shown in step 904. Subsequent to
step 903 or 904, the system could send a message to the mobile user
to inform the mobile user that the e-coupon could not be
distributed or the system could provide other options.
[0072] In another embodiment, in step 905, an optimal consolidation
process is applied to the sellers whose e-coupons are not
distributed in step 904. The optimal consolidation process is
discussed below. Then the system determines whether the number of
pending requests is greater than or equal in value to an optimal
threshold value as shown in step 906. If the number of pending
requests is greater than or equal to the optimal threshold, the
e-coupon is distributed to the mobile user as shown in step 903.
However, if the number of pending requests is less than the optimal
threshold, the e-coupon is not distributed as shown in step 907.
The optimal threshold is calculated in the same manner, but only
after the optimal consolidation process, discussed below, is
applied.
[0073] The essence of the conditional e-coupon service is to let
the conditional e-coupon service provider automatically negotiate a
"wholesale" deal for a group of mobile users, who may not know each
other, with a seller provided that the seller can make profit from
this "wholesale" transaction. The conditional e-coupon service
provider charges a commission (the advertising fee) from the seller
once a "wholesale" deal is offered (i.e., e-coupons are
distributed).
[0074] Conventional coupon distribution is a special case for the
conditional e-coupon service. If a seller wants e-coupons to be
distributed unconditionally, the seller could simply set the
pre-advertising profit margin to be infinity.
[0075] The following sections only discuss conceptually how the
e-coupons are distributed, and how the e-coupon distribution
notification messages are sent to sellers. A significant cost as
well as high effectiveness for wireless advertising is expected.
The conditional e-coupon service provider has to collect
advertising fees for every e-coupon distribution, because it is
very difficult to identify if the deal is made solely due to the
e-coupon. By contrast, it is relatively easy for a seller to verify
that some extra traffic is indeed brought in by e-coupons. E-coupon
distribution may be mapped to the cell broadcasting operation in
GSM networks that support cell broadcast short message service. It
may also be carried out using point-to-point short message service.
If the former method is used, the distribution cost is independent
of the number of mobile users. In this case, the conditional
e-coupon service provider may charge a flat advertising price for
each distribution. If the latter method is used, the conditional
e-coupon service provider may have to adopt an advertising price
that increases as the number of mobile users increases.
[0076] Now an optimal consolidation process will be discussed from
a mathematical model to proposed implementation. The following
models and calculations quantitatively demonstrate that the
conditional e-coupon service can be optimized for all involved
parties--sellers can make maximum profit by adopting the best
threshold values, the conditional e-coupon service provider can
make maximum revenue, and mobile users can receive a maximum number
of e-coupons (by running an optimal consolidation process in the
conditional e-coupon system.)
Mathematical Model
[0077] One mathematical model for calculation of a threshold is
described herein. A conditional e-coupon service provider will
charge an advertising price p against a seller each time an
e-coupon for the seller is distributed. Three typical advertising
price plans will be analyzed below. They are (a) a flat advertising
price plan (p is independent of N, which is also the number of
mobile users who will receive e-coupons if e-coupons are
distributed), (b) a proportional advertising price plan (p is
proportional to N), and (c) a linear advertising price plan (a
combination of a flat and a proportional advertising fee). p is
denoted as p(N).
[0078] There are N(t) pending requests at the seller during a
processing cycle [t, t+T), where N(t) is a random process defined
on {0, Z.sup.+} and T is the length of a processing cycle, such as
5 minutes or 10 minutes. It should be noted that N(t) might not
necessarily be stationary. Nonetheless, N(t) will be denoted as N
hereafter for this example, since it will be shown that the best
threshold is independent of the statistical distribution function
of N(t).
[0079] The goods or service offered by the seller to one mobile
user has a pre-advertising profit margin m, which is the difference
of the offering price minus the cost (excluding the advertising
price p). m is fixed in every processing cycle.
[0080] The estimated number of redeems is denoted as M. This number
is not used directly in the calculation. Instead, a new notation r
called the e-coupon effectiveness ratio, which is equal to MIN, is
introduced. The effectiveness ratio r essentially is a random
variable. It should be noted that r might vary significantly from
one processing cycle to another, because it is dependent of N and
many competition factors that are subject to large
fluctuations.
[0081] A threshold .theta. is defined with regard to the number of
pending requests N. That is, if N.gtoreq..theta., the conditional
e-coupon commerce server should distribute the e-coupon to the N
mobile users who sent in pending requests for the e-coupon. The
reason the threshold .theta. is not defined with regard to the
estimated number of redeems M is because N could be modeled by a
Poisson distribution, while modeling M is more difficult.
[0082] The extra revenue that the conditional e-coupon service can
generate for the seller during a processing cycle is a function of
the threshold .theta.,
f(.theta.)=mrNu(N-.theta.)
where, u(x) is the step function,
u ( x ) = { 1 for x .gtoreq. 0 0 for x < 0 ##EQU00003##
The advertising cost that is charged by the conditional e-coupon
service provider for distributing an e-coupon in a processing cycle
is also a function of the threshold .theta.,
c(.theta.)=p(N)u(N-.theta.).
[0083] The expectation of the extra profit that the seller can make
from the conditional e-coupon service in a processing cycle is
given by,
P ( .theta. ) = E ( f ( .theta. ) - c ( .theta. ) ) = n = .theta.
.infin. ( mrn - p ( n ) ) Pr ( N = n ) ##EQU00004##
where, Pr(N=n) is the probability of N=n. It is assumed that
Pr(N=n)>0 for any n.
[0084] The objective for a seller is to find the best threshold
.theta. such that the seller can maximize the expectation of the
extra profit from the conditional e-coupon service. This is
equivalent to maximizing the expectation of the extra conditional
e-coupon profit in any processing cycle. That is,
Max{P(.theta.)}.
[0085] The seller does not solve the maximization problem alone,
because the effective ratio r varies in every processing cycle.
Alternatively, the seller only gives the pre-advertising profit
margin m in its seller profile, from which the conditional e-coupon
controller determines the best threshold and authorizes
distribution for the seller accordingly.
[0086] The objective for the conditional e-coupon service provider
is to find the best advertising price p such that the conditional
e-coupon service provider can maximize the conditional e-coupon
revenue from all sellers, subject to the sellers all adopting best
thresholds respectively. For example, this objective can be
addressed by maximizing the conditional e-coupon revenue that the
conditional e-coupon service provider earns from one seller in
every processing cycle.
[0087] The objective for mobile users is to maximize the ratio of
the number of received e-coupons versus the number of requests,
subject to all candidate sellers adopting their best thresholds
respectively. The same optimal consolidation process done by the
conditional e-coupon service provider achieves this objective; this
maximizes the number of sellers that can eventually meet the
threshold requirement and thus distribute their e-coupons to mobile
users.
Best Threshold--Flat Advertising Price Plan
[0088] Under the flat advertising price plan p(N)=p.sub.o, the best
threshold for a seller is equal to p.sub.o/mr, no matter what kind
of statistical distribution N obeys, where x is the smallest
integer that is no smaller than x.
[0089] Let .theta..sub.m denote the best threshold and let
P(.theta..sub.m) denote the maximum conditional e-coupon profit.
That is,
P ( .theta. m ) = Max { P ( .theta. ) } = n = .theta. m .infin. (
mrn - p o ) Pr ( N = n ) ##EQU00005##
[0090] Assume .theta..sub.m=.theta.'<p.sub.o/mr, we have
P ( .theta. m ) = n = .theta. ' p 0 mr - 1 ( mrn - p 0 ) Pr ( N = n
) + n = p 0 mr .infin. ( mrn - p 0 ) Pr ( N = n ) ##EQU00006##
[0091] Because the first term in the right side of the above
equation is always negative, we have,
P ( .theta. m ) < n = p 0 mr .infin. ( mrn - p 0 ) Pr ( N = n )
= P ( p 0 mr ) ##EQU00007##
[0092] This is contradictory to the assumption that
P(.theta..sub.m) is the maximum conditional e-coupon profit, so it
has to be .theta..sub.m.gtoreq.p.sub.o/mr.
[0093] Similarly, assume .theta..sub.m=.theta.'>p.sub.o/mr, we
have
P ( .theta. m ) = .theta. ' .infin. ( mrn - p 0 ) Pr ( N = n ) <
n = p 0 mr .theta. ' - 1 ( mrn - p 0 ) Pr ( N = n ) + .theta. '
.infin. ( mrn - p 0 ) Pr ( N = n ) = P ( p 0 mr ) ##EQU00008##
[0094] This is because the first term in the second row of the
above equation is always positive. This is also contradictory to
the assumption that P(.theta..sub.m) is the maximum conditional
e-coupon profit, so it has to be that
.theta..sub.m.ltoreq.p.sub.o/mr.
[0095] Combining these results, we have .theta..sub.m=p.sub.o/mr.
It should be noted that the function form of Pr(N=n) is not needed.
That is, the best threshold .theta..sub.m does not depend on the
statistical distribution of N.
[0096] The independence between the best threshold .theta..sub.m
and the statistical distribution of N is a feature for the
conditional e-coupon service, which makes it easy for the
conditional e-coupon controller to choose the best threshold that
is valid all the time for every seller.
Best Threshold--Proportional Advertising Price Plan
[0097] Under the proportional advertising price plan p(N)=p.sub.1N,
the best threshold for a seller is equal to 1 if p.sub.1<mr, or
infinity if p.sub.1.gtoreq.mr.
[0098] Plugging p(N)=p.sub.1N into P(.theta.), we have,
P ( .theta. ) = n = .theta. .infin. ( mr - p 1 ) nPr ( N = n )
##EQU00009##
[0099] Note that n Pr(N=n).gtoreq.0 for any n. If
p.sub.1.gtoreq.mr, i.e., mr-p.sub.1.ltoreq.0, we have
P(.theta.).ltoreq.0. In this case, the maximum conditional e-coupon
profit is zero only if .theta..THETA..infin., which means the
seller cannot issue e-coupons no matter how many requests are
pending.
[0100] If p.sub.1<mr, i.e., mr-p.sub.1>0, we have
(mr-p.sub.1)nPr(N=n)>0
for every n. Clearly, Max {P(.theta.)}=P(1), which means the seller
should issue e-coupons as long as there are pending requests.
[0101] Similarly, the function form of Pr(N=n) is not needed. That
is, the best threshold .theta..sub.m doesn't depend on the
statistical distribution of N.
Best Threshold--Linear Advertising Price Plan
[0102] Under the linear advertising price plan
p(N)=p.sub.0+p.sub.1N, the best threshold for a seller is equal
to
p 0 mr - p 1 ##EQU00010##
if p.sub.1<mr, or infinity if p.sub.1>mr.
[0103] Plugging p(N)=p.sub.0+p.sub.1N into P(.theta.), we have,
P ( .theta. ) = n = .theta. .infin. ( ( mr - p 1 ) n - p 0 ) Pr ( N
= n ) ##EQU00011##
[0104] If p.sub.1.gtoreq.mr, i.e., mr-p.sub.1.ltoreq.0, we have
((mr-p.sub.1)n-p.sub.0)Pr(N=n).ltoreq.0
for every n. Hence, P(.theta.).ltoreq.0. In this case, the maximum
conditional e-coupon profit is zero when .theta..fwdarw..infin.,
which means the seller cannot issue e-coupons no matter how many
requests are pending.
[0105] If p.sub.1<mr, i.e., mr-p.sub.1>0, following the same
proving method used in the calculation based upon the Flat
Advertising Price Plan above, the best threshold is equal to
p 0 mr - p 1 . ##EQU00012##
[0106] Again, the function form of Pr(N=n) is not needed. That is,
the best threshold doesn't depend on the statistical distribution
of N.
[0107] From the above calculations, it can be seen that the best
threshold .theta..sub.m is always independent of the statistical
distribution of N. This is a nice feature, which makes it very easy
for the commerce server to choose the best threshold without a
statistical model for N.
[0108] A "conditional e-coupon profit vs. threshold" curve for a
participating seller is shown in FIG. 10, where 21 curves are
plotted from top to bottom, corresponding to the cases that N obeys
a Poisson distribution with an arriving rate .lamda. varying from
25 to 5. The flat advertising price p is 2.00, the pre-advertising
profit margin m is 2.00, and the effectiveness ratio r is 0.05.
FIG. 10 shows that, no matter how X changes, the best threshold
.theta..sub.m corresponding to the maximum conditional e-coupon
profit P(.theta..sub.m) is always equal to 20.
Maximum Conditional E-Coupon Profit
[0109] Another property for the best threshold is the maximum
conditional e-coupon profit recited herein. When choosing the best
threshold .theta..sub.m, a participating seller can earn a positive
maximum conditional e-coupon profit P(.theta..sub.m), no matter how
expensive the advertising price p(N) is.
[0110] It should be noted that the maximum conditional e-coupon
profit P(.theta..sub.m) does depend on the statistical distribution
of N, although the best threshold .theta..sub.m does not. If N
obeys a Poisson distribution with an arriving rate .lamda., where
Pr(N) is the Poisson probability distribution function with an
arriving rate of .lamda., i.e.,
Pr ( N = n ) = .lamda. n n ! - .lamda. ##EQU00013##
[0111] Under the flat advertising price plan the maximum
conditional e-coupon profit P(.theta..sub.m) is:
P ( .theta. m ) = mr .lamda. Pr ( N .gtoreq. p 0 mr - 1 ) - p 0 Pr
( N .gtoreq. p 0 mr ) ##EQU00014##
[0112] Under the proportional advertising price plan with
p.sub.1<mr, the maximum conditional e-coupon profit is given
by:
P(.theta..sub.m)=(mr-p.sub.1).lamda.
[0113] And under the linear advertising price plan with
p.sub.1<mr, the maximum conditional e-coupon profit is given
by:
P ( .theta. m ) = ( mr - p 1 ) .lamda. Pr ( N .gtoreq. p 0 mr - p 1
) - p 0 Pr ( N .gtoreq. p 0 mr - p 1 ) ##EQU00015##
[0114] A "maximum conditional e-coupon profit vs. price" curve for
a participating seller is shown in FIG. 11, where N obeys a Poisson
distribution with an arriving rate .lamda.=35. The pre-advertising
profit margin m is 2.00 and the effectiveness ratio r is 0.05. It
can be seen that the maximum conditional e-coupon profit
P(.theta..sub.m) for the seller will always be positive, although
the profit approaches to zero rapidly as the advertising price p
increases. Pursuant to these equations, a seller will make money
under any circumstance in a statistical sense.
Upper Bound of Price
[0115] If a flat advertising price plan is adopted, there exists an
upper bound of the advertising price p.sub.o for the conditional
e-coupon service provider to charge every seller for each
distributed e-coupon. The conditional e-coupon service provider can
make the maximum conditional e-coupon revenue from a seller if the
advertising price p.sub.o is set to be the upper bound, and there
is no advantage in using a higher price. FIG. 12 demonstrates this
correlation between the maximum conditional e-coupon revenue and
the upper bound of the advertising price p.sub.o.
[0116] In FIG. 12, the line 1202 on the top shows that the ideal
conditional e-coupon revenue that a conditional e-coupon service
provider can earn from every distributed e-coupon for a seller
increases linearly as the advertising price p.sub.o increases.
However, line 1202 is the ideal case, assuming the seller chooses
an unconditional e-coupon distribution scheme. If the seller
chooses the conditional e-coupon service with the best threshold,
the best threshold increases rapidly as the advertising price
increases. This causes the probability of distributing e-coupons
for the seller to decrease rapidly, which is shown by line 1203.
Therefore, there exists the maximum conditional e-coupon revenue
for the conditional e-coupon service provider to earn from the
seller at some price point, as shown by line 1204. This advertising
price is the upper bound price, because even if the conditional
e-coupon service provider sets a higher price, he cannot earn more
conditional e-coupon revenue. The value of the upper bound
advertising price depends on the statistical distribution of N.
[0117] If a proportional advertising price plan is adopted, there
exists an upper bound for the unit advertising price p.sub.1 that
the conditional e-coupon service provider charges a seller for
every e-coupon in each e-coupon distribution. Clearly, the upper
bound is equal to mr. The conditional e-coupon service provider can
set a unit price close to the upper bound but cannot make it equal
to the upper bound, because the store cannot make profit at all in
this scenario.
[0118] If a linear advertising price plan is adopted, there exists
an upper bound for the flat advertising price p.sub.0 as well as an
upper bound for the unit advertising price p.sub.1. The upper bound
for the unit advertising price p.sub.1 is equal to mr. The upper
bound for the flat advertising price p.sub.0 is dependent of the
unit advertising price p.sub.1 as well as the distribution function
of N
[0119] In one example, a conditional e-coupon service provider sets
the advertising price below the upper bound in order to expand the
base of sellers and to compete with other conditional e-coupon
service providers.
Optimal Consolidation
[0120] At the end of every processing cycle, the commerce server
may find that some sellers have insufficient pending requests to
warrant an e-coupon distribution. This does not mean that all of
these sellers cannot issue their e-coupons. Some of them may become
eligible after others are declared hopeless by the commerce server,
because the overall e-coupon effectiveness ratio of the remaining
sellers can increase significantly due to less competition after
others quit the game. Clearly, there is an optimization
problem--the commerce server must follow an optimal procedure to
selectively remove sellers from the candidate sellers, such that
the number of remaining sellers that eventually become eligible to
issue e-coupons is maximized. This optimal consolidation process
guarantees mobile users receive a maximum number of e-coupons,
provided that all issuing stores can make maximum profit. It also
guarantees the conditional e-coupon service provider can make
maximum revenue after the advertising price is determined.
[0121] Assume that N e-coupon requests arrive at the commerce
server in a processing cycle. Each of them has an individual redeem
probability of r.sub.n, n=1, 2, . . . , N, which are estimated from
the quota system in the last quota reset cycle. They are pending at
L sellers with various competition factors d.sub.ln, where d.sub.ln
is the competition factor for the n.sup.th request pending at the
l.sup.th store. It can take either of two values. One is zero,
which means the n.sup.th request is not a pending request to the
l.sup.th store. The other is equal to one divided by c.sub.n, which
is the number of candidate sellers for the l.sup.th request. For
the l.sup.th store, there are in total N.sub.l pending requests. We
have 1.ltoreq.c.sub.n.ltoreq.L. Without loss of generality, we
assume none of the sellers has a sufficient number of pending
requests that can warrant an e-coupon distribution in the beginning
of the consolidation process. We define an eligibility distance
D.sub.l(i), l=1, 2, . . . , L, for every seller,
D l ( i ) = p l - m n = 1 N r n d ln ##EQU00016##
where i is the iteration index.
[0122] Because derivation of a globally optimal consolidation
algorithm is a difficult problem, a sub-optimal algorithm is given
below.
[0123] (1) The commerce server calculates the eligibility distance
for every remaining seller at Step i and finds the maximum
eligibility distance Max{D.sub.l(i)};
[0124] (2) If Max{D.sub.l(i)}.ltoreq.0 or none is remaining, the
commerce server stops the consolidation process and distributes
e-coupons for all remaining sellers, if there are any remaining
sellers. Otherwise, the algorithm continues.
[0125] (3) If there is only one seller associated with
Max{D.sub.l(i)}, this seller is declared hopeless and removed from
the candidate sellers. Otherwise, the commerce server randomly
chooses one of such sellers and declares it hopeless.
[0126] (4) i=i+1, go to (1).
[0127] FIGS. 13A-13E show an example of how a mobile user might
interact with on embodiment of the present invention. A mobile user
drives by a fast food restaurant, such as Fast Food, Inc.
(fictional name), and decides to get a meal. Fast Food, Inc. is a
seller that maintains e-coupons through the conditional e-coupon
service provider. As shown in FIG. 13A, the mobile user can make a
request to the conditional e-coupon service provider to browse the
e-coupons available for Fast Food, Inc. by inputting the seller
name, as shown by reference character 1302. In this example, the
mobile user can make the request on a Personal Digital Assistant
(PDA) device 1301.
[0128] As disclosed above in one embodiment, the present invention
receives the request to browse e-coupons available for Fast Food,
Inc., determines a plurality of Fast Food, Inc. locations (or
stores) that are local to this mobile user and a plurality of
corresponding e-coupons from each Fast Food, Inc. location. As
shown in FIG. 13B, a mobile user receives the e-coupon availability
data of local Fast Food, Inc. and e-coupons available for each. The
first screen on the PDA could show the Fast Food, Inc. that is
physically closest to the mobile user. In this case, the PDA has
received a location at 110 Pine St., as shown by reference numeral
1303, and all the corresponding e-coupons available for that
particular Fast Food, Inc., as shown by reference numeral 1304.
[0129] FIG. 13C represents what the mobile user may see on a
separate page. The location of this Fast Food, Inc. at 2345 N.
Stein Rd. 1305 could be the second closest Fast Food, Inc. in
physical location to the mobile user. A listing, as shown by
reference numeral 1306, of all available e-coupons for that
particular Fast Food, Inc. location is also shown. Multiple
locations of various locations could be listed on one screen.
[0130] Should the mobile user choose a particular e-coupon, such as
the e-coupon for a Nuggets from the 110 Pine St. Fast Food, Inc.
location, the present invention would receive the request for the
e-coupon and then authorize whether the e-coupon should be
distributed as disclosed above. FIGS. 13D and 13E show two possible
screens that may be seen from the authorization step of the present
invention.
[0131] In the case that the e-coupon is distributed, FIG. 13D shows
receipt, as shown by reference numeral 1307, of the requested
e-coupon from the particular Fast Food, Inc. location. With the
e-coupon, the mobile user can redeem the e-coupon at the particular
location and receive the desired Nuggets at the e-coupon price or
discount.
[0132] In the case that the e-coupon is not distributed, FIG. 13E
shows a notification message, as shown by reference numeral 1310,
stating that the e-coupon was not available for that request.
Again, other options may be available, such as a prompt to request
the same e-coupon again, as shown by reference numeral 1311, and a
prompt to browse for other e-coupons, as shown by reference numeral
1312.
[0133] As shown in FIGS. 13D and 13E, some examples of further
options the system could provide are an option to receive driving
directions, reference numeral 1308, to the seller, request the same
e-coupon, reference numeral 1311, and an option to browse for other
e-coupons, reference numerals 1309 and 1312. In addition, the
system could automatically log a request for the same e-coupon
during a second processing cycle if the first processing cycle
ended without a distribution of the desired e-coupon. This
automatic feature of the present invention could be a field in a
mobile user or seller profile, such as field 622f or 633l.
[0134] While the invention has been described with respect to
specific examples including presently preferred modes of carrying
out the invention, those skilled in the art will appreciate that
there are numerous variations and permutations of the above
described systems and techniques that fall within the spirit and
scope of the invention as set forth in the appended claims.
* * * * *