All public courses using SAS are held at SAS, Marlow:

and cost for two-day courses 1000 GBP (plus VAT) per person which includes lunch and refreshments. Reduction to 900 GBP (plus VAT) per person for multiple bookings from the same organisation.

To book a place on a course either click here, or phone +44 (0)118 934 4265, or print and complete an application form (pdf format 115 Kb) .

Course dates for 2008/2009 :

• Propensity Scoring Models: 13th-14th Nov 2008, 12th-13th Feb 2009, 11th-12th June 2009, 5th-6th Nov 2009
  
• Segmentation of a Customer Database: 3rd-4th Nov 2008, 26th-27th March 2009, 24th-25th Sept 2009
  
• Planning & Interpreting of Marketing Campaigns: 18th-19th Sept 2008, 4th-5th June 2009
  

Planning and Interpreting Marketing Campaigns

Using SAS/STAT^® to analyse responses to complex marketing campaigns.

The Course

Your marketers want to test...

• several forms of an offer,
• several versions of the creative,
• different types of discount

...all in the same campaign.

Without careful planning two things typically happen:

• the number of different tests becomes very large with sample sizes smaller than intended,
• a small number of the many combinations of offer, creative and discount are tested and when you come to try and analyse the results it is hard to draw useful conclusions about the benefits of the different components of the campaign.

How can the campaign be designed to try and address all the questions simultaneously and efficiently?

Classical experimental design techniques can be used to plan efficient marketing campaigns where several features are tested together.

This course explores the potential of designed experiments and the extent to which good design procedures, like factorial experiments and case-control studies, can support database marketing trials.

It begins by discussing the efficient analysis of simple tests which a straightforward comparison between several values of a single customer attribute - a situation which often arises where customer segments are simply labelled as different, without the structure in the differences being taken into account. We will consider aspects like

• sample size (including the size of control groups),
• potential adjustments for unmatched test groups,
• the evaluation of differences in response rates as well as in the value of the responses.

We then introduce the concept of experiments with several factors and the opportunity to measure the level of interaction between different campaign factors. We will show how careful exploitation of the factorial structure can create a "win-win" situation where you can increase the information extracted from the test without a proportionate increase in sample numbers.

When there are many factors to be investigated in a campaign, it will not always be possible to manage the complete set of combinations. It is often possible to choose a special subset of the combinations which still allow the main issues to be addressed. The benefits and dangers of these fractional factorials are discussed.

We bring out the advantages of introducing as much balance as possible into the test while showing that it is perfectly possible, with the appropriate type of analysis, to make sense of tests where the tests groups are of very different sizes or inherently unbalanced in key characteristics.

The analysis of marketing campaigns and the assessment of their benefits can often be sharpened up by taking into account customer attributes not explicitly used in the design. For example, the revenue obtained from each customer in the three months prior to the trial can be used to "control expectations" about the likely responsiveness of each customer, and we will address this type of Covariance Analysis.

The course ends by looking at "meta analysis" of historical campaigns and tests simple methods that can be used to investigate comparisons which were not embedded in any single test.

Prior Expertise:

This course does not assume prior statistical expertise. It is suitable for an analyst with knowledge of the context and some SAS skills (ability to write a simple data step and to use simple base procedures).

The SAS^®; Tools used in the course are: PROC GLM and PROC LOGISTIC.