Predictive Analytics Increase Sales, Reduce Campaign Costs

By: Sysware  05-Apr-2012
Keywords: Data Mining

This client needed to target their product to the ten thousand people most likely to buy.

Sysware worked together with Mapnet to identify those people. Mapnet geocoded the client's existing customer database and referenced the customer addresses to census meshblocks. Sysware then joined the NZ customer database file to the large database of information from the 2001 Census database. By running data mining procedures against this data, Sysware was able to identify the characteristics of current buyers of the product.

Sysware discovered that buyers had certain characteristics, including a high probability of being born overseas, a high income, internet access and tendency to have financial investments.

Sysware then identified the meshblocks containing residents with the highest concentration of these characteristics. The customer was given a list of targeted meshblocks, and a detailed report containing graphic representations of distributions and customer profiles. The list was sorted by propensity to purchase, the client used this to mail a high-quality promotional package to the top 10,000 addresses (excluding existing customers).

The results of the exercise speak for themselves.

  • Targeted address response rates were up 52% from the previous year.
  • Average customer spend in response to the addressed campaign increased by 284% on the previous year.

By utilising Sysware’s data mining expertise and Mapnet’s mapping capability, the client achieved a vastly improved response rate, and a huge return on their investment. Sysware was proud to be involved in such a successful collaborative exercise, which provides compelling evidence on the success of targeting on increasing sales and reducing costs.

Keywords: Data Mining

Other products and services from Sysware


Powering Up – Building a Data Warehouse

The project included providing OLAP data structures for end users to exploit, the development of web-based reporting for end users; and the establishment and optimisation of the reporting architecture. To be able to participate in this sharing of the electricity market, this power utility had to build an ultra-reliable data warehouse that collected real-time power pricing data 24 by 7.


Business Intelligence Solutions

We’ve worked with many leading organisations, helping them to increase their profits, optimise business and technical performance, analyse and report on their data, develop and implement powerful customer retention strategies and reduce their operating costs. By combining data mining with traditional audit techniques, you create effective and efficient tests for making cost recoveries, discovering revenue opportunities and identifying fraud.


Using Predictive Modelling to Reduce Churn

The marketing team decided to focus on high value customers who were most likely to leave and asked Sysware to help carry out a predictive modelling exercise to identify these customers. The data was fed into a SAS predictive modelling tool that uses neural networking techniques to build a model to predict churn behaviour and factors with little or no influence were removed.


Data Consolidation delivers value and efficiency

Sysware successfully tendered for the job and promptly delivered a cost-effective solution on time, to budget, meeting all expectations for cost, performance and quality. It is also a distributed application split into five regional instances, with a structurally identical database existing for each region. Using SAS Data Integration Studio, Sysware quickly delivered the project, enabling the required improvements, including.