System and method for selecting optional inserts with optimal value in an inserting machine5367450Abstract A method of making a selection of optional enclosures for a mailpiece including required enclosures. The method comprises the steps of assigning to each of the optional enclosures a weight, a cost and a benefit, creating a list of the potential compositions of the mailpiece, the list including a total of 2.sup.n combinations of the optional enclosures, where n equals the number of optional enclosures, computing a total weight of the mailpiece for each of the potential compositions, computing a value of the mailpiece for each of the potential compositions, the value being any computable function of the benefit and the cost of each optional enclosure included in each of the potential compositions, and selecting from one of the potential compositions based on the one having the optimal value. Claims What is claimed is: Description FIELD OF THE INVENTION
TABLE I
______________________________________
List of Possible Enclosure Combinations
______________________________________
0. A
1. A + E1
2. A + E2
3. A + E3
4. A + E4
5. A + E1 + E2
6. A + E1 + E3
7. A + E1 + E4
8. A + E2 + E3
9. A + E2 + E4
10. A + E3 + E4
11. A + E1 + E2 + E3
12. A + E1 + E2 + E4
13. A + E1 + E3 + E4
14. A + E2 + E3 + E4
15. A + E1 + E2 + E3 + E4
______________________________________
The total number of combinations is 16. In the general case of "n" optional enclosures the total number of combinations is equal to: ##EQU1## where [] denotes the number of ways to choose a k-element subset from an n-element set. In the case of n=5, the number of combinations is 32, and in the case of n=10, the number of combinations is 1024. The n=10 case covers a vast majority of the practically encountered situations. However, even for a larger n the number of combinations to be analyzed is quite manageable even for a modest modern microprocessor. Which one of the combination alternatives is the "best" one? The answer to this question depends on a criterion for the best. If an objective numerically valued function "VALUE" can be defined on the set of all possible combinations then the maximal value of this function can define the best combination. In a particular embodiment the objective function "VALUE" can be defined as the difference (or the ratio or another function) between the total benefit and the total cost corresponding to a particular combination. The choice of a particular "VALUE" function depends on the considerations employed by the owner of the inserting system. For example, it may depend on the accounting system used by the owner. For the purpose of the present invention, this choice is irrelevant since the scheme described below will work equally as well with any "VALUE" function. (It is assumed, of course, that the computational effort required for calculation of the "VALUE" function depends little on the actual nature of this function which is certainly true for all practical applications.) The total cost, which is the cost of postage and the cost of producing enclosures, can be determined based on the total weight of the mailpiece, worksharing level and the sum of enclosure costs. In the example use herein, it is assumed, for the sake of simplicity, that the mailpiece to be assembled is not prebarcoded and not a member of presort group (i.e. it is not subject to a discount and will be paid at the full postage rate) and that the weights in ounces of assembly A and enclosures E1, E2, E3, E4 are as follows: Weight (A)=1.65 Weight (E1)=0.2 Weight (E2)=0.15 Weight (E3)=0.1 Weight (E4)=0.05 For the sake of simplicity, Weight (A) includes the weight of the mailing envelope. It is also assumed that the cost of producing (or price paid for) each of enclosures E1, E2, E3, E4 are 2.cent., 2.cent., 3.cent., and 3.cent.respectively. Then all the possible compositions of the mailpiece will have the incremental costs (i.e. the costs excluding the cost of making assembly A which is set to 0) in 1992 U.S. postal rates listed in Table II.
TABLE II
______________________________________
Incremental Costs
______________________________________
0. Cost (A) = Postage (1.65 oz) = 52.cent.
1. Cost (A + E1) = 2.cent. + Postage (1.85 oz) = 54.cent.
2. Cost (A + E2) = 2.cent. + Postage (1.8 oz) = 54.cent.
3. Cost (A + E3) = 3.cent. + Postage (1.75 oz) = 55.cent.
4. Cost (A + E4) = 3.cent. + Postage (1.7 oz) = 55.cent.
5. Cost (A + E1 + E2) = 4.cent. + Postage (2.0 oz) = 79.cent.
6. Cost (A + E1 + E3) = 5.cent. + Postage (1.95 oz) = 57.cent.
7. Cost (A + E1 + E4) = 5.cent. + Postage (1.9 oz) = 57.cent.
8. Cost (A + E2 + E3) = 5.cent. + Postage (1.9 oz) = 57.cent.
9. Cost (A + E2 + E4) = 5.cent. + Postage (1.85 oz) = 57.cent.
10. Cost (A + E3 + E4) = 6.cent. + Postage (1.8 oz) = 58.cent.
11. Cost (A + E1 + E2 + E3) = 7.cent. + Postage (2.1 oz) = 82.cent.
12. Cost (A + E1 + E2 + E4) = 7.cent. + Postage (2.05 oz) = 82.cent.
13. Cost (A + E1 + E3 + E4) = 8.cent. + Postage (2.0 oz) = 83.cent.
14. Cost (A + E2 + E3 + E4) = 8.cent. + Postage (1.95 oz) = 60.cent.
15. Cost (A + E1 + E2 + E3 + E4) = 10.cent. + Postage
(2.15 oz) = 85.cent.
______________________________________
[Table II is based on the current rates in the USA. It can be easily changed to any other rate structure around the world. Also the worksharing option can be included without major complications.] It follows from the table that the minimal cost combination is the original assembly A without any optional enclosures. This is trivial and the least interesting case. A "benefit function" is defined as follows. For purposes of describing the present invention, it will be assumed that each optional enclosure has a numerically valued benefit associated with it. For example, it can be an expected value of incremental business which the mail sender anticipates to generate as a result of inclusion of given inserts into an envelope. It can be determined, for instance, as the total incremental dollars generated as a result of the aggregate mailing divided by the number enclosures of a given type inserted in the mailing. For example, if as a result of sending 1,000 enclosures advertising sale of a piece of furniture the furniture store usually sells only one such a piece for $500, then the expected benefit of one advertising enclosure is $500/1,000 =50.cent.. Of course, the furniture store would not pay 50.cent. per piece but would have to consider some acceptable margin for profit, for instance 80%. In this case the furniture store would be willing to pay 10.cent. per advertising enclosure and realize gross profit margin of $400. The scheme may be as complex as desired. For example, the results of advertising can be measure for two different mailings, one in 1,000 pieces and another in 10,000 pieces and the difference can be evaluated. The results can be measured and normalized or known statistics can be used. The demographic information can be easily taken into account in arriving at estimated benefits for optional enclosures. Moreover, if the demographic information is available to the control computer during the mail assembly process, e.g., via a control document, it can be used to modify benefits "on the fly". Generally, there are well known methods for measuring effectiveness of direct mail advertising which include well defined and understood procedures. See, for example, The Dartnell DIRECT MAIL AND MAIL ORDER HANDBOOK, by R. S. Hodgson, Third Edition-1980, Appendix O. It is assumed that the benefit of each of the enclosures E1, E2, E3, E4 are 60.cent., 12.cent., 10.cent. and 25.cent. respectively. Then all the possible combinations will have the benefits listed in Table III. (The benefit of assembly A is set to 0 to simplify the description herein and the benefit is by definition an additive function to each of the combinations.)
TABLE III
______________________________________
Benefit of the Enclosures
______________________________________
0. Benefit (A) = 0
1. Benefit (A + E1) = 60.cent.
2. Benefit (A + E2) = 12.cent.
3. Benefit (A + E3) = 10.cent.
4. Benefit (A + E4) = 25.cent.
5. Benefit (A + E1 + E2) = 72.cent.
6. Benefit (A + E1 + E3) = 70.cent.
7. Benefit (A + E1 + E4) = 85.cent.
8. Benefit (A + E2 + E3) = 22.cent.
9. Benefit (A + E2 + E4) = 37.cent.
10. Benefit (A + E3 + E4) = 35.cent.
11. Benefit (A + E1 + E2 + E3) = 82.cent.
12. Benefit (A + E1 + E2 + E4) = 97.cent.
13. Benefit (A + E1 + E3 + E4) = 95.cent.
14. Benefit (A + E2 + E3 + E4) = 47.cent.
15. Benefit (A + E1 + E2 + E3 + E4) = 107.cent.
______________________________________
If the function VALUE defined as the difference between the benefit and the cost, then this results in Table IV.
TABLE IV
______________________________________
Value of the Enclosures
______________________________________
0. VALUE (A) = -52.cent.
1. VALUE (A + E1) = 6.cent.
2. VALUE (A + E2) = -42.cent.
3. VALUE (A + E3) = -45.cent.
4. VALUE (A + E4) = -30.cent.
5. VALUE (A + E1 + E2) = -7.cent.
6. VALUE (A + E1 + E3) = 13.cent.
7. VALUE (A + E1 + E4) = 28.cent.
8. VALUE (A + E2 + E3) = -35.cent.
9. VALUE (A + E2 + E4) = -20.cent.
10. VALUE (A + E3 + E4) = -23.cent.
11. VALUE (A + E1 + E2 + E3) = 0.cent.
12. VALUE (A + E1 + E2 + E4) = 15.cent.
13. VALUE (A + E1 + E3 + E4) = 12.cent.
14. VALUE (A + E2 + E3 + E4) = -13.cent.
15. VALUE (A + E1 + E2 + E3 + E4) = 22.cent.
______________________________________
It is clear from Table IV that the optimal value of 28.cent. is found in line 7 which corresponds to the selection of enclosures E1 and E4. Thus, in accordance with the present invention optional enclosures E1 and E4 will be selected for the mailpiece. This is different from the selection of optional inserts pursuant to a "topping off" method. According to a "topping off" method, for example as in U.S. Pat. No. 4,639,873, inserts E2, E3 and E4 would be selected because the three optional inserts can be added to the mailpiece without exceeding the one ounce postage category of the mailpiece determined from the weight of the required enclosures W(A). In general, the "topping off" method will produce the same results as the present invention only when the cost of material (Table II) and the benefit (Table III) is set to zero for all the optional enclosures. In such a case, the inserting machine cannot be operated to maximize the potential value of the mailpieces being assembled. Once cost and benefit are assigned to the optional enclosures an optimal value can be determined for each mailpiece, thus providing means for making an optimal selection of the optional inserts for each mailpiece. Referring now to FIG. 2, in accordance with the present invention there is provided a flowchart of an algorithm for computing optimal composition of a mailpiece based on optimal value of optional inserts. The overall process of optimization proceeds as follows with n being the total number of optional enclosures. In the example described herein, n=4 is the number of optional enclosures denoted by E1, E2, E3 and E4. At step 100, three numerical attributes, namely weight (w1, w2, . . . , wn), cost (c1, c2, . . . , cn) and expected benefit (b1, b2, . . . , bn) for all optional enclosures are entered into an inserting machine control computer. The weight (wa) of the non-optional enclosures is also entered. R(W) is the rate function (or rate table) which defines the postage to be paid for the mailpiece with the weight W. If a change in the rate function is necessary, it is also entered. (It is noted that the expected benefit can be modified "on the fly" as described in U.S. Pat. No. 4,817,042.) After the expected benefits for all optional enclosures are entered, at step 102 the control processor (36 in FIG. 1) determines all the potential compositions of the mailpiece to be assembled. At step 104, the control processor computes the total weight (W1, W2, . . . , W2.sup.n) for all the potential compositions of the mailpiece. It is noted that W1 denotes an empty set, i.e., a mailpiece without any optional enclosures. At step 106, the control processor computes a postage rate for each of the potential compositions. At step 108, the control processor computes the Value for each of the potential compositions. At step 110, the processor selects the maximal value from the list of values computed for potential compositions. Finally, at step 112, the processor sends control signals to the appropriate optional enclosure feeders to realize the maximal value for the mailpiece. The foregoing steps are repeated for every mailpiece. As previously described, the total number of potential compositions is 2.sup.n where "n" is the number of optional enclosures. Typically, the number of such enclosures is between 2 and 8and therefore the total number of possible combinations is between 4 and 256. For each of the possible combinations, the function VALUE is computed and the maximal value (which always exists) is selected. Then the combination corresponding to this maximal value is selected and the control system of the inserting machine executes the actual assembly process. As a practical matter, the benefit attributes of all optional enclosures are not always known and sometimes cannot be estimated. In this case, the unknown benefits can be set to zero or to reasonably small values and the process of the present invention will select the minimal postage assembly. If the inclusion of optional enclosures was paid for by a third party then the benefit for these enclosures can be set based on the amount paid per item and a known weight distribution of the intended mail run. In this case, the third party is assured that all the enclosures will be sent while the mailing party, i.e., the party which is providing the insertion and mailing service, will be able to minimize the total postage paid for the mailing. The value based algorithm in accordance with the present invention can be modified to accommodate sliding postal rates or any other arbitrarily complex postal rates as long as the postal rates are algorithmically computable based on the weight, worksharing or other desired attributes. The benefit value of the optional enclosure can be also set or modified by the control document or a control file for computerized data base driven inserting machines. A very similar approach can be applied for optimization of the entire mail run. For example, if the entire mail run consists of 10,000 pieces and weight distribution of mandatory enclosures for all the mail pieces in the run is known before the process of actual mail assembly, i.e., insertion, sealing, postaging etc., takes place, one can define the value function for the entire run. This function would take into account the difference (or ratio or any arbitrary computable function) between the benefit and the cost for the entire mail run. Thus, the determination of whether to include or not a given enclosure is based not on the total value of the given mailpiece but on the values for all mailpieces. This, of course, requires a prior knowledge of the weight distribution of mandatory enclosures for all the mailpieces in the mail run. Referring now to FIG. 3, in accordance with the present invention there is provided a flowchart of an algorithm for determining the optimal composition of a mail run. The process of optimization of the mail run proceeds as follows with n being the total number of optional enclosures and m being the total number of mailpieces in the mail run. At step 200, three numerical attributes, namely weight (w1, w2, . . . , wn), cost (c1, c2, . . . , cn) and expected benefit (b1, b2, . . . , bn) for all optional enclosures are entered into an inserting machine control computer. The weight (wa) of the non-optional enclosures is also entered. R(W) is the rate function (or rate table) which defines the postage to be paid for the mailpiece with the weight W. If a change in the rate function is necessary, it is also entered. After the expected benefits for all optional enclosures are entered, at step 202 the control processor (36 in FIG. 1) determines all the potential compositions of a mailpiece to be assembled. At step 204, the control processor computes the total weight (W1, W2, . . . , W2.sup.n) for all the potential compositions of the mailpiece. It is noted that W1 denotes an empty set, i.e., a mailpiece without any optional enclosures. At step 206, the control processor computes an array of postage rates (R(wam+W2.sup.n)) for each of the potential compositions for each of the mailpieces in a mail run. At step 208, the control processor computes an array of all possible compositions of the mail run. Then, at step 210, the control processor computes a list of potential values of the mail run by computing a value for each of the potential combinations of mailpieces in the mail run by selecting one element from each row in the array of possible compositions to compute a list of potential values of the mail run. The value function is shown as a general function of benefits, costs an postage rates F[b1, b2, . . . , bn, c1, c2, . . . , cn, R(wai+Wji)], such as the difference between benefit and cost as shown in FIG. 2. It will be appreciated that other functions, such as a ratio of benefit to cost may also be used. Finally, at step 112, the processor selects an optimal value from the list of values computed for potential compositions of the mail run. While the present invention has been disclosed and described with reference to a single embodiment thereof, it will be apparent, as noted above that variations and modifications may be made therein. It is, thus, intended in the following claims to cover each variation and modification that falls within the true spirit and scope of the present invention.
|
Same subclass Same class Consider this |
||||||||||
