Published on 25-Mar-2011
"IBM SPSS Modeler helps improve the model and attain the balance between prediction and business profitability." - Julio Quiñonez, Director of Business Intelligence, Neck & Neck
Neck & Neck
BA - Business Analytics, BA - Business Intelligence, Smarter Planet
Smarter Solutions for Retail
In 2003 Neck & Neck, a Spain based children's clothing chain with 200 stores in nine countries that designs, produces and distributes clothing for children up to 14 years of age, began its customer loyalty club. The Neck & Neck Club offers a series of advantages like special rebates or personalized catalogs to its affiliates.
With a commercial base of 22,000 contacts on the international level, there was a clear need for knowing the customers in detail and approaching them in the most personalized way possible, while increasing the profitability of these actions. Neck wanted to reduce costs, make mailing campaigns profitable, and use existing data for recapturing customers who make purchases during one season but not during the following one
Neck & Neck identified the need to manage a propensity model that would indicate which customers should be sent catalogs, improve the segmentation models, and offer customer management decision making tools.
The campaign response rates increased by more than 25% and the average number of consumer sales grew by 15% in one year, even though fewer catalogs and promotions were mailed.
Neck & Neck improve customer intimacy and make better decisions throughout their organization. They attained an explosive growth and optimal localization of the target, tailoring the campaign to the needs of each customer. They realized cost savings from optimized campaigns.
To read a Spanish version of this case study, please click here.
In 2003 Neck & Neck, a children’s clothing chain that designs, produces and distributes clothing for children up to 14 years of age, began its customer loyalty club. With a commercial base of 22,000 contacts on the international level, there was a clear need for knowing the customers in detail and approaching them in the most personalized way possible, while increasing the profitability of these actions.
In the words of Julio Quiñones, Director of Neck & Neck Business Intelligence, “BI emerges as a result of the company’s reflection upon the importance of the customer loyalty club, achieving explosive growth and managing great amounts of information. Thanks to IBM SPSS Modeler from SPSS, an IBM Company, we can contact our customers through the means that they prefer and send them the ideal product, in accordance with the age or gender of the children, in addition to personalized and well-timed promotions”. The Neck & Neck Club offers a series of advantages to its affiliates: 5 euro checks for each 100 euros spent, personalized catalogs, and more. For purchases of greater than 500 euros, the checks are worth 10% of the total, and the purchaser may elect to take advantage of discounts before the official period.
Iradia Sampol, Account Manager at IBM SPSS, puts it this way, “as a retail company, it was clear that a comprehensive data mining process could offer high profitability in solving problems, such as in managing the sending of catalogs and identifying who the best customers are to achieve the greatest number of positive responses to well-timed promotions with the lowest cost to the company”. In general, Neck & Neck wanted to reduce costs, make mailing campaigns profitable, and use existing data for recapturing customers who make purchases during one season but not during the following one.
Internalization of the knowledge
“We had to integrate tools that gave us decision making support for internal model management, such as the problem of the loyalty club data management. Our database grew exponentially and our main problem was optimal management when sending out the catalogs”, affirms Julio Quiñones. Neck & Neck identified the need to manage a propensity model that would indicate which customers should be sent catalogs, improve the segmentation models, and offer customer management decision making tools. Different alternatives were evaluated, and although at first they began to work with consulting firms, “it is faster not to depend on third parties and internalize all of the processes”, added Julio. “What’s more, this allows us to have more autonomy and to save money, because the contracting consulting services are more costly. A further advantage is that by carrying out these processes in house we are training our own teams”. In 2006, they began to select technology providers with the approval of the management team.
Neck & Neck chose the IBM SPSS solutions for various reasons: “we found this to be a market reference, with easy access to service, including an office in Madrid, making it easy for us to learn, as well as having an accessible budget that allows us to grow by acquiring additional modules. The implementation was very simple and was carried out in less than six months after SPSS contacted us to evaluate our needs, until we carried out the testing and discovered of the functionality of the tool”. A single user solution was installed which could be expanded on a per-module basis with the option of contracting the server based version.
The next challenge was having available the knowledge for fully exploiting this tool. Neck & Neck did not hesitate to include intensive internal training programs with advanced modeling so that the knowledge would remain within the company. “Above all, we wanted to avoid committing the error of having interesting technology and not knowing how to use it. In the end we focused on having our employees being able to use it on their own. We don’t have much time and it is important to be able to export the models we have created to each country”.
Why IBM SPSS Modeler?
Modeler is the leading data mining tool from IBM SPSS. Ana María Molina, Marketing Manager of SPSS puts it this way, “today there are many companies that gather a great amount of data, but if they do not use these data appropriately, all the effort is wasted”.
Julio Quiñones adds, “when you start doing model analyses you don’t have much time. We were aware that we were a smaller business that just incorporated all of the capabilities of a consulting company and internalizing them while keeping the company from increasing its number of employees. In order to achieve this we needed simple software that was adapted to ourselves which, above all, would help us with data mining. What for? For example, to better choose the customers to whom we should send our catalogs”.
Another of Neck & Neck’s problems was to find the ideal variables for establishing the model as to who these preferred customers were: “you can lose a lot of time searching for ideal variables (profitability, usefulness of purchase, purchasing capacity, and age). SPSS is of great help because it shows us whether or not there is improvement in the model of our selected variables. When we started making models with the consulting company, we knew that two variables were established that helped them to select these preferred customers: segmentation by family type (number or age of children) and customer activity (dormant, active, and inactive). With the IBM SPSS solutions we learned to find these ideal variables ourselves, such as customer profitability, and Modeler helps us establish their level of importance within the model. All in all, Modeler helps improve the model and attain the balance between prediction and business profitability”.
The other Neck & Neck business problem that Modeler has helped solve is the existence of a customer profile that one season buys but does not come back later. This is to say, it is a problem of reactivating customers. In order to solve this problem we must define which customers to contact for reactivating their purchasing behavior. Neck & Neck had previously carried out testing with promotional activity and had built up a data history.
Once these customers were identified, we offered them a special benefit during a certain period in order to attract those who had not bought anything for three months. This worked, but it was necessary to better identity which customers were more likely to respond favorably.
“SPSS helped us to define customer variables (what makes it more probable for a customer who has not bought anything in three months to come back?) In order to do this we analyzed the age of the children (the lower it was, the more likely there would be a purchase), the path the customer has taken with us (accumulated purchases or profitability) with different variables, etc. IBM SPSS Modeler provides the possibility of seeing how the model would improve its predictions and adjusts it according to new elements that are added. We have been including another series of variables that we have been adjusting more (number and age of children, how long ago the latest purchase was made, amount, profitability, etc.). SPSS allows us to understand whether this model with certain variables is better than the previous one”.
Since it is scalable, Modeler further allows a limitless number of data, cases and variables to be used. The performance of the tool is directly proportional to the power of the computer.
“After six months of implementation I met with the management team to show them the campaigns throughout the season and compare them with the ones from a year ago. We found that, with a lower number of catalogs and promotions sent, the campaign brought us a greater response rate along with an increase in the average purchase”, stated Julio Quiñones.
The difference between the two campaigns is considerable: before the installation of Modeler at Neck & Neck, and with the help of the consulting company, a campaign response rate between 15 and 20 percent was achieved. Currently, with a significantly lower number of mailings, this rate has risen to between 20 and 25 percent. This is an enormous increase in profitability when taking into consideration that the average customer purchase is between 100 and 120 euros. If 90,000 impacted customers have a 20 to 25 percent response rate, we can see why just one campaign was able to pay for the investment and training costs.
“It is clear that our growth is linked to SPSS and we will be expanding models and the license in order to achieve a server based system”, said Julio. Modeler has a series of options with different segmentation choices in order to continually improve management and business model.
“We still have to develop the sequential cross sale analysis in order to analyze and model customer purchases at different events; as they come two or three times per season, they create a purchase chain. The idea is to create this based on customer type and present them with an offer after having analyzed why one product leads to the acquisition of the next”. In order to achieve this they have not hesitated in acquiring additional modules.
For Julio, “in making a decision that requires this kind of investment it is essential to return results, acquire technology according to company growth, and, most importantly, keep in mind the market standard. It should be an easy to learn tool and information should be offered such that the greatest possible number of employees can benefit from it”.
Products and services used
IBM products and services that were used in this case study.
Footnotes and legal information
SPSS Inc., an IBM Company Headquarters,
233 S. Wacker Drive, 11th floor
Chicago, Illinois 60606
SPSS is a registered trademark and the other SPSS products named are trademarks
of SPSS Inc., an IBM Company. © 2010 SPSS Inc., an IBM Company.
All Rights Reserved.
IBM and the IBM logo are trademarks of International Business Machines
Corporation in the United States, other countries or both. For a complete list of IBM
trademarks, see www.ibm.com/legal/copytrade.shtml.
Other company, product and service names may be trademarks or service marks of
References in this publication to IBM products or services do not imply that IBM
intends to make them available in all countries in which IBM operates.
Any reference in this information to non-IBM Web sites are provided for convenience
only and do not in any manner serve as an endorsement of those Web sites. The
materials at those Web sites are not part of the materials for this IBM product and use
of those Web sites is at your own risk.
© Copyright IBM Corporation 2010