This webpage contains the instances used in the paper

Talla Nobibon, F., Leus, R. and Spieksma, F.C.R. (2011).  Optimization models for targeted offers in direct marketing: exact and heuristic algorithms.  European Journal of Operational Research, 210, 670-683. (pdf) (DOI)

The instances are saved as .txt files. Each instance is identified by:

• its group; S1 for 100 clients, S2 for 200 clients, S3 for 300 clients, M1 for 1000 clients, M2 for 2000 clients and L for 10000 clients.

• its hurdle rate R, which is either 5 for 5%, 10 for 10% or 15 for 15%.

• its number of products n, which is either 5, 10 or 15.

• an integer indicating whether we have a tight budget (1), an average budget (2) or a large budget (3).

• either the letter s indicating that each client can receive only few offers or the letter l if each client can receive a higher number of offers.

Therefore, an instance named S1-5-10-2-l refers to an instance with 100 clients, a hurdle rate of 5%, with 10 products, an average budget and where each client can receive a higher number of offers. We next present the instances per subgroup.

• S1, n = 5  (this subgroup contains 18 instances)

• S1, n = 10  (18 instances)

• S1, n = 15  (18 instances)

• S2, n = 5  (18 instances)

• S2, n = 10  (18 instances)

• S2, n = 15  (18 instances)

• S3, n = 5  (18 instances)

• S3, n = 10  (18 instances)

• S3, n = 15  (18 instances)

• M1, n = 5  (18 instances)

• M1, n = 10  (18 instances)

• M1, n = 15  (18 instances)

• M2, n = 5  (18 instances)

• M2, n = 10  (18 instances)

• M2, n = 15  (18 instances)

• L, n = 5  (18 instances)

• L, n = 10  (18 instances)

• L, n = 15  (18 instances)

Each instance has m+4 lines, where m is the number of clients.  The first line with three columns contains successively the number of clients, the number of products and the hurdle rate. The next m lines each have 2n+1 columns. The ith line contains the data corresponding to the  ith client. At that line, the first n columns represent the cost of offering each product to that client while the next n columns report expected revenue per product if the considered product is accepted by that client. Finally, the last column contains an integer representing the maximum number of offers that can be made to that client.

The last three lines contain n columns each. The last line reports the fixed cost for each product. The last but one line reports for each product the budget available while line m+2 gives for each product the minimum number of clients who must receive an offer of that product is used during the campaign.