 |
Description  |
|
|
FIELD OF THE INVENTION
This invention relates to a method of monitoring the delivery status of
selected pieces in a batch delivery through a delivery service.
BACKGROUND OF THE INVENTION
In a direct delivery campaign, such as the mailing of a seasonal catalog to
customers, timely delivery of mail is important. Thus, there is a need to
monitor mail delivery to aid in planning such a campaign, and to identify
and correct problem areas in an ongoing campaign.
One known method of monitoring the delivery status of mail is used by the
postal service to monitor delivery between different post offices. This
method provides no information concerning delivery to the ultimate
destination.
Other known monitoring methods utilize return forms or postcards which are
included with a piece of mail. A respondent upon receiving the mail piece
fills in requested information on the form or postcard and mails it back
to the originator of the mailing. However, because of reliance on a return
mailing, such a monitoring method provides responses which are not
received quickly enough to take corrective action in the event that
problems exist. Also, the delivery status information must be manually
correlated once it is received. With a large mailing this can be
burdensome.
The present invention is intended to overcome these and other problems with
delivery monitoring methods.
SUMMARY OF THE INVENTION
It is an object of this invention to provide a method for monitoring the
delivery status of selected pieces which improves the reporting response
time so that problems can be identified quickly and accurately.
In accordance with this invention selected pieces directed to respondents
are included in a batch delivery which are forwarded by a delivery
service. Each piece is encoded with indicia uniquely identifying the
respondent therefor. After the pieces are forwarded to the respondents the
date of receipt is telephonically reported to a data processor by the
respondent. The date of receipt is recorded and a report is generated. The
indicia may comprise a seven digit number which uniquely identifies the
respondent by name and address, and a three digit number which uniquely
identifies the piece. The data processor replies to the respondent using
voice synthesizing means instructing the respondent to provide receipt
information. The receipt information may be a four digit number
representing the date on which the respondent receives the mailing piece,
and an additional digit representing the condition of the received piece.
The receipt information is recorded, for example, on a magnetic disk
memory to be later used for generating daily and final reports.
The invention may be used, for example, to monitor delivery status of
selected mailing pieces which are delivered by the postal service.
Another feature of this invention is that one of the seven respondent
identification digits is a check digit. The check digit is used to verify
that the reported seven digit number is correct.
A further feature of this invention is the use of a push button operated,
multiple frequency coded telephone by the respondent to report information
to the data processor.
Yet another feature is the generation of an advance notification postcard
which is mailed to the respondents in advance of a mailing to advise the
anticipated delivery of an encoded mailing piece.
Still another feature of this invention is the insertion of a list of
respondents onto a customer list for the batch mailing so that postal
discount levels and handling characteristics are not altered.
A further feature of this invention is the selection of respondents
according to a geographic distribution of the batch mailing.
Further features and advantages of this invention will be readily apparent
from the specification and from the drawings.
BRIEF DESCRIPTION OF THE DRAWINGS
FIG. 1 illustrates a typical mailing piece showing an encoded mailing label
affixed thereon;
FIG. 2 illustrates in greater detail the mailing label encoded with indicia
identifying the respondent and the mailing piece;
FIG. 3 illustrates an advance notification postcard;
FIG. 4 is a block diagram illustrating the invention;
FIGS. 5a and 5b are a flow diagram of the delivery monitoring method;
FIGS. 6a-6c are a flow diagram for the data processor; and
FIGS. 7a-7c are a flow diagram of the insertion method for including the
mailing pieces in a batch mailing.
DESCRIPTION OF THE INVENTION
Referring first to FIG. 1, a mailing piece 10 is illustrated with a mailing
label 12 affixed thereon. The mailing piece 10 is to be forwarded to a
respondent through the postal service in a batch mailing. Such a batch
mailing may be to send a seasonal catalog to customers. Included in the
batch mailing are additional catalogs to be mailed to respondents to
monitor date of receipt of the catalogs and condition of the catalogs when
received.
Alternatively, comparable delivery services can be monitored. The term
"mail" used herein includes other such forms of delivery.
The mailing label 12 is illustrated in greater detail in FIG. 2. The label
12 has indicia including the respondents' name, address and a multi-digit
number 14 which uniquely identifies the respondent and the mailing piece.
The delivery respondents are preselected and trained as to what action to
take when an encoded mailing piece 10 is received. The respondents are
selected so that their representation corresponds with the U.S. Postal
Service delivery statistics file. Thus, the database of respondents
represents all postal respondents, on a smaller scale.
When a respondent is recruited he or she is issued an instruction card (not
shown). The instruction card advises the respondent what steps to take
when an encoded mailing piece is received. Also described thereon are
procedures to follow in special situations, such as when a respondent will
be away from home.
Each respondent is assigned a seven digit number which uniquely identifies
the respondent and the respondent's address. The first two digits thereof
identify a group of respondents which the respondent is assigned to. A
particular mail order house may have its own list of respondents who are
only utilized for monitoring mailing of their own catalogs. Such a mail
order house will be assigned its own two digit group number. The next four
digits are randomly assigned to respondents in a group. The last of the
seven digits is a check digit which is used to verify the correctness of
the seven digit number when it is reported by a respondent. The check
digit is determined by multiplying each of the six digits by 2.sup.x where
x represents the position of the digit from right to left, starting with
the one position; summing the products; dividing the total by nine; and
subtracting the remainder from nine. For example, the respondent code
123456 has a check digit of 3 which is calculated as follows:
((1.times.2.sup.6)+(2.times.2.sup.5)+(3.times.2.sup.4)+(4.times.2.sup.3)
+(5.times.2.sup.2)+(6.times.2.sup.1))/9=26R6; 9-6=3
This formula is based on number theory and permits the check digit to be
used to minimize errors which could be caused by transposition of any of
the six digits.
A unique three digit number is randomly assigned to the particular catalog,
or other mailing piece, which is being monitored. The multi-digit number
14 comprises the seven digit and the three digit numbers.
It is desirable to advise each respondent in advance to expect a mailing
piece which is to be reported. Otherwise, the respondent may not notice
the multi-digit number 14 and fail to report receipt of the mail piece.
For this reason each respondent receives an advance notification postcard
18, see FIG. 3, prior to receiving the mailing piece notifying the
anticipated delivery thereof. The postcard 18 also advises the respondent
as to what action should be taken if the mailing piece is not received by
a given date.
With reference to FIG. 4, a data processing system 20 records and reports
information provided by respondents. Each respondent utilizes a telephone
22 to report receipt information. A telephone interface 24 is coupled to
telephone lines 26 which enable the respondents to communicate with the
telephone interface 24. The telephone interface 24 monitors incoming
telephone calls and includes a voice synthesizing circuit or converter 24a
for prompting the calling respondent to enter date of receipt information
into the data processor 20. If the respondent utilizes a push button
operated, multiple frequency coded telephone, the receipt information is
entered using the telephone's push buttons. The telephone interface 24
includes a circuit 24b for converting the multiple frequency codes into
ASCII coded characters required by the data processor 20. Alternatively, a
speaker independent voice recognition circuit or converter 24c can be
utilized so that the respondent can enter the receipt information by
reciting it into the telephone's mouth piece, or a pulse decoder 24d is
used if respondents report using rotary dial telephones.
The telephone interface is coupled to a first central processing unit (CPU)
28. Peripheral devices such as a CRT display 30 and a magnetic disk memory
32 are also coupled to the first CPU 28. The disk memory 32 stores the
receipt information reported by the respondents.
The first CPU 28 is a personal computer which can perform only a single
task at one time. An additional CPU is required in order to record receipt
information from respondents and print reports simultaneously. Therefore,
a pair of modems 34, 36 couple the first CPU 28 to a second CPU 38. The
second CPU 38, also a personal computer, is coupled to a CRT display 40, a
magnetic disk memory 42 and a printer 44. The modems 34, 36 transmit
receipt information from the first CPU to the second CPU. The printer 44
generates daily reports and batch reports. Alternatively, the user could
move magnetic disks between the two CPU's 28, 38, eliminating the modems
34, 36. If a multi-task capability CPU is used, the second CPU 38 is
unnecessary. The printer 44 is then coupled to the first CPU 28.
With reference to FIGS. 5a and 5b, a system flow chart illustrates in
detail the method of monitoring the delivery status of selected mailing
pieces from a batch mailing using the devices previously described with
reference to FIGS. 1-4.
At block 46 the respondent database is defined according to the postal
service delivery statistic file previously discussed. Each respondent is
assigned to a group at block 47; a 6-digit number, the group number and
respondent number is assigned at block 48. The check digit is calculated
and assigned to the respondent at block 49. The respondent database is
stored in a cross reference file at block 50.
At block 51 the originator of the batch mailing, e.g. the catalog printer
and mailer, selects the mail pieces to be monitored. A group, or groups,
of respondents to receive mailing pieces to be monitored as part of the
batch mailing by the data processor is selected at block 52. The groups
are selected using the 2 digit group code previously discussed. The cross
reference file block 50 stores the names, addresses and identification
numbers for the respondents, as previously discussed. The data processor
at block 53 selects available respondents from these groups and prepares
advance notification cards 18 which are mailed to respondents at block 54
notifying each respondent to anticipate an encoded mailing piece 10. If a
respondent will be away from home for any extended period of time, the
respondent can be "dialed-out" of the system by reporting in advance to
the data processor 20, their seven digit respondent code, a mail piece
code of 999, the month and day they will return, and a condition code of
1. Thereafter, those respondents will not be selected at block 53 until
the day on which they have indicated as their return date. Subsequent to
preparing notification postcards at the block 53, which are mailed at the
block 54, control advances to a block 56. The three digit code
representing the mailing piece is generated at block 56. Block 60 encodes
the mailing labels 12 with the ten digit number 14 uniquely identifying
the respondent and the mailing piece 10.
A method for inserting the monitoring mail pieces, into the batch mailing
is discussed in greater detail below.
The batch including the encoded mail pieces 10 is mailed to the selected
delivery respondents at block 62. The mail pieces 10 are received by
respondents at block 64. The respondent places a telephone call to the
data processor 20 at block 66. The data processor 20 waits at block 68 for
a call to be received. When the call is answered, of the data processor
20, by the first CPU 28 through the voice synthesizer 24a prompts the
respondent to enter the encoded multi-digit number 14 and receipt
information at block 70. This information is entered at block 72 by
touching the push buttons of the phone, and the call is terminated at
block 74.
The details of how the information is recorded and verified by the data
processor 20 is discussed in greater detail below.
After the call is terminated, decision block 76 determines whether or not
it is the end of the day. If it is not the end of the day, the system
returns to block 68. Otherwise, at the end of the day a daily report is
printed at block 78. Decision block 80 determines if the batch mailing is
complete. The batch is complete a predetermined time after mailing or when
reports are received from all respondents. If the batch is not complete,
the system returns to waiting block 68. If the batch is complete, a final
batch report is printed at block 82, the batch is statistically analyzed
at block 83, and the monitoring of the batch ends at block 84. The
analysis of the batch at block 83 is discussed in greater detail below.
Referring now to FIGS. 6a-6c, a flow diagram of the operation of the data
processor 20 is illustrated. The data processor 20 initiates a program for
the monitoring method at blocks 90-97. Particularly, the data processor 20
is initialized in a conventional manner at a block 91. At respective
blocks 92 and 93, the data processing system 20 gets the day and time and
the session number. This information may be entered by an operator using a
conventional keyboard, or may be automatically entered if the system is
configured to do so. At a block 94 the session number, date and time are
printed either on the CRT 30 or the printer 44, see FIG. 4. At a block 95
the transaction log files for the system are open, and file headers are
written at a block 96. Subsequently, at a block 97, phrase modules and an
interface are initialized to enable communications between the first CPU
28 and the telephone interface 24 for communicating over the telephone
lines 26. Block 98 causes the telephone interface 24 to initialize the
telephone channels to accept calls from respondents.
Block 100 continually checks each telephone channel for an incoming call. A
decision block 102 determines whether or not a call has been received. If
no call is received, the system returns to block 100. If a call is
detected, block 104 issues on off-hook command to answer the call. A
greeting such as "hello" is generated by the voice synthesizer 24a at
block 106. A try-counter is set to one at block 107. The try-counter
provides an indication of the number of times the respondent has entered
receipt information during a call. The use of the try-counter is discussed
in greater detail below.
The respondent is prompted at block 108 to enter a fifteen digit number.
The fifteen digit number comprises the ten digit number 14 and an
additional five digits. Four of the latter five digits represent the month
and day when the mailing piece was received. For example, for mailing
pieces received on April 9th the four digit number is 0409. The last digit
is a condition digit representing the condition of the received mailing
piece. The condition digit is 1=good, 2=fair, 3=bad, and 4=no receipt.
Decision block 110 determines whether or not fifteen digits have been
entered within a predetermined time period. If the fifteen digits are not
entered, a "try again" message is given at block 111 and the time period
is reset, as by returning to the block 108 to prompt the respondents to
enter the fifteen digit number. If the fifteen digits are entered within
the set time, a record is created in memory and the digits are edited at
block 112.
A decision block 114 verifies that the fifteen digit number is entered
correctly. Particularly, the check digit previously discussed is used to
verify the seven digit respondent number; the mailing piece code is
verified to be valid; the month and day are verified to exist, i.e. 1349
is an invalid code as there is no 49th day in a 13th month; and the
condition digit is verified to be 1-4. If the number is verified, decision
block 115 determines if the respondent has entered a mail piece code of
999, indicating that the respondent is to be dialed out. If so, the
respondent is removed from the list of available respondents at block 116
until the date on which the respondent indicates as a return date.
Thereafter, the system returns to block 100 to check for subsequent
incoming telephone calls. If the respondent has not dialed out, block 117
writes the receipt information to the transaction record which was created
at block 112. The voice synthesizer 24a issues a confirmation message at
block 118, repeats each recorded digit at block 120 and prompts the
respondent at block 122 to confirm the fifteen digits which were recorded.
Decision block 124 determines whether or not a confirmation signal is
received. If a confirmation signal is received the voice synthesizer 24a
generates a "thank you" message at block 126 and an on-hook command is
given at block 128 terminating the call. The system returns to block 100
to check for subsequent incoming telephone calls.
If the digits entered are not verified at decision block 114 or the
recorded digits are not confirmed at decision block 124 the system moves
to block 130 which determines whether the try-counter is greater than
three. If the try-counter is less than three block 132 increments the
try-counter by one, and the voice synthesizer 24a prompts the respondent
to "try again" at block 134, the system returns to block 108 and the
respondent is again prompted to enter the fifteen digits. If the
try-counter is greater than three at block 130, indicating three instances
of an incorrect number being reported or failure to confirm, the voice
synthesizer 24a generates a "call back" message to the respondent at block
136. The system issues an on-hook command at block 138 terminating the
call and an error statistic record is written at block 140. The system
returns to block 100 to check for subsequent incoming calls.
It is desirable that the batch mailing be presorted so as to take maximum
advantage of discounts offered by the postal service as well as to insure
desired handling characteristics. To pre-sort the batch, customer address
information is stored in order of customer zip code on a magnetic storage
tape. One such method of pre-sorting is described in U.S. Pat. Application
No. 668,989, filed Nov. 7, 1984, now U.S. Pat. No. 4,674,052, issued Jun.
16, 1987, assigned to the assignee of the present invention, the
spe.cification of which is hereby incorporated by reference. As described
therein, the postal service provides discounts according to the quantity
of pre-sorted pieces being mailed to the same carrier route, 5 digit zip
code post office, or a region or city having the same first three digits
in their zip codes. When mail pieces which are to be monitored are added
to a batch mailing, their insertion into the batch should not affect the
discount levels and handling characteristics which would otherwise be
available. In fact, the batch should be pre-sorted in the same quantity of
postal packages and sacks as there would be if no such monitored mailing
pieces were added. This allows the results of the monitoring to accurately
reflect the delivery pattern of the batch itself.
A method of inserting the monitoring mail pieces into the batch mailing is
illustrated in the flow diagram of FIGS. 7a-7c. This method is implemented
using a programmed data processor (not shown) similar to that in FIG. 4. A
list of customers to receive mail pieces in the batch mailing is provided
on a medium, such as a magnetic disk memory for access by the data
processor. A list of delivery respondents is provided on a second such
medium. The data processor, according to this invention, inserts the
respondent names onto the list of batch customers so that the discount
level is unchanged.
At block 150 the data processor retrieves a record for the first
respondent. A record for the first customer is retrieved at block 152.
Decision block 154 determines if the retrieved respondent and customer
have the same 5 digit zip code. If the zip codes are the same, block 156
determines the number of customers having that zip code. The number of
customers for each zip code is stored in a summary record on the storage
medium. Decision block 158 determines whether the number of customers
having the zip code is low. If the number of customers is not low,
indicating that the addition of this respondent will alter the discount
level and handling characteristics otherwise provided by the postal
service, the respondent is skipped at block 160. Otherwise, the respondent
is added to the batch output list at block 162. The output list is the
list of all recipients of mail pieces in the batch mailing, including
customers and respondents. After the respondent has been skipped at the
block 160, or the respondent has been added to the output list at block
162, the customer is added to the output list at block 163.
Decision block 164 determines if there are any more customers. If not,
insertion is complete. If more customer records are available, the next
customer is retrieved at block 165 and decision block 166 determines
whether or not there are more respondents. If additional respondents are
available the data processor retrieves the next respondent at block 168.
Thereafter, the insertion cycle continues at decision block 154, discussed
above.
If decision block 154 determines that a customer and a respondent do not
have the same 5 digit zip code, decision block 170 determines whether or
not the first three digits of the respondent's and customer's zip codes
are identical. If decision block 170 determines that the first three
digits of the zip codes are identical, block 172 determines the number of
customers having the same first three digits. The determination at block
172 is made in the same way as the determination made at block 156,
discussed above. Decision block 174 determines whether the number of
customers having the same three digits is low. If the number of customers
is not low, indicating that the addition of this respondent will alter the
discount level and handling characteristics otherwise provided by the
postal service, the respondent is skipped at block 160. Otherwise, the
respondent is added to the output list at block 162, and the insertion
method continues thereafter.
If decision block 170 determines that the respondent and customer do not
have the same first three digits of their respective zip codes, the
respondent is placed in a hold table at block 176. Decision block 178
determines if the hold table is full. If the hold table is full, an error
message is written at block 180 and the program is aborted at block 182.
The hold table will normally only be full when there has been a large
number of respondents having an incorrect first three digits of their zip
code according to the geographical selective distribution of the batch
customers. If decision block 178 determines that the hold table is not
full, the program checks the customer file 3 digit count at block 184 to
determine the number of customers remaining having the same first three
zip code digits. Decision block 186 determines if there are any such
remaining customers, and if so, returns to block 168 for the insertion
method to continue. If decision block 186 determines if there are no
remaining customers having the same first 3 zip code digits, an error
message is written at block 188 and the program is aborted at block 190.
Because the respondents were initially selected according to geographical
distribution of the customers, there should never be a situation where
there are more respondents than customers. Therefore, if no customers are
remaining, a problem has occurred, such as customer records being removed
or truncated from the file. If the program is aborted at either block 182
or block 190, the data processor must be operated to restart the system,
as necessary or desired, as by returning to the system start preceding
block 150, see FIG. 7a.
If decision block 166, previously discussed, determines that there are no
more respondents, while more customers remain in the customer file, the
hold table is checked at block 192. Decision block 194 determines if there
are any respondents in the hold table with the same first 3 zip code
digits as the currently retrieved customer. If no such respondents are in
the hold table, the program advances to block 163 and the customer is
added to the output list. If decision block 194 determines that there are
matching respondents, block 196 determines the number of customers having
the same first 3 digits. The determination at block 196 is made in the
same way as the determination made at block 156, discussed above. Decision
block 198 determines whether the number of customers having the same 3
digits is low. If the number of customers is not low, indicating that the
addition of this respondent will alter the discount level and handling
characteristics otherwise provided by the postal service, all respondents
in the hold table having the same first 3 digits are cleared at block 200,
and the customer is added to the output list at block 163. Otherwise, the
respondent is retrieved from the hold table at block 202, and the
respondents entry in the hold table is cleared at block 204. Thereafter,
respondent is added to the output list at block 162, and the insertion
method continues thereafter.
Thus, respondents are only added to the customer list if their addition
does not alter the postal pre-sort discounts available for the batch or
alter the handling characteristics provided thereby. Similarly, if no
customer in the batch mailing has the same 3 digit or 5 digit zip code as
the respondent the respondent will not be added to the customer list. This
results in respondents having a geographically selective distribution
corresponding to the batch mailing distribution.
By utilizing a data processor for receiving telephonic communications from
respondents this invention permits problems in a mailing of a batch to be
identified quickly and accurately with minimal need for personnel. If
there is a problem with a particular batch mailing, the postal service can
be notified immediately to correct the problem. Also, daily and final
batch reports are obtained automatically without need for manual data
entry. Furthermore, the use of the unique ten digit code minimizes errors.
The daily report generated by the data processing system can be formatted
to indicate date of delivery according to geographical locations of the
respondents. Also, the reports can be formatted to indicate the date of
receipt according to a pre-sort level for the mailing, receipt or
condition according to the pre-sort level. Information such as pre-sort
level of a given mailing can readily be input into the system at the
initiation of a batch mailing. Such reporting options allow the monitoring
method of this invention to readily provide more detailed reports then
were available with prior known monitoring methods.
The postal service has a standard for mail delivery time based on the
distance which the piece is mailed. Samples of the reported information
are analyzed according to their distances and whether they were pre-sorted
according to carrier route, 5-digit zip code or 3 digits of the zip code.
These samples are compared to the postal service standard to insure that
the observed delivery times agree with the expected delivery times. Each
sample is non-parametric, i.e., no assumptions are made regarding the
distribution of the respondents. To insure accuracy, multiple samples of
delivery performance are analyzed using the "bootstrap" method of
statistical analysis. The "bootstrap" method is described in an article
entitled "ComputerIntensive Methods in Statistics", by Persi Diaconis and
Bradley Efron, Scientific American, May 1983, Vol. 248, No. 5, which is
hereby incorporated by reference. According to the bootstrap method, a
large number, such as 1,000, of random samples of no more than 5% of the
respondents each are analyzed to determine a bootstrap measure of
accuracy. If a statistically reliable difference exists between actual and
expected delivery time the postal service can be notified.
* * * * *
|
|
 |
|
 |
Description |
|
