27.10.2012 Views

Analysis of Sales Promotion Effects on Household Purchase Behavior

Analysis of Sales Promotion Effects on Household Purchase Behavior

Analysis of Sales Promotion Effects on Household Purchase Behavior

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

6 DATA DESCRIPTION<br />

6.1 Introducti<strong>on</strong><br />

To study the effects <str<strong>on</strong>g>of</str<strong>on</strong>g> sales promoti<strong>on</strong>s <strong>on</strong> household purchase behavior, c<strong>on</strong>tinuous<br />

household-level scanner data <strong>on</strong> product items is combined with retail outlet data regarding<br />

their promoti<strong>on</strong>al activity (the so-called causal data). The household data comes from GfK<br />

C<strong>on</strong>sumerScan and the retail data comes from IRI/GfK InfoScan. Both data sets cover the<br />

period from the last quarter <str<strong>on</strong>g>of</str<strong>on</strong>g> 1995 (IV 1995) until the last quarter <str<strong>on</strong>g>of</str<strong>on</strong>g> 1997 (IV 1997), 9<br />

quarters in total. GfK C<strong>on</strong>sumerScan is the market leader with respect to household panel data<br />

in The Netherlands and is part <str<strong>on</strong>g>of</str<strong>on</strong>g> the pan-European market research agency GfK. For this<br />

study, data <str<strong>on</strong>g>of</str<strong>on</strong>g> various frequently-purchased packaged c<strong>on</strong>sumer goods is used, covering a<br />

variety <str<strong>on</strong>g>of</str<strong>on</strong>g> food/beverage products. We use <strong>on</strong>ly a subset <str<strong>on</strong>g>of</str<strong>on</strong>g> the available data (c<str<strong>on</strong>g>of</str<strong>on</strong>g>fee, s<str<strong>on</strong>g>of</str<strong>on</strong>g>t<br />

drinks, fruit juice, potato chips, candy bars, and pasta). These categories are chosen because<br />

<str<strong>on</strong>g>of</str<strong>on</strong>g> their frequent purchase and promoti<strong>on</strong> activity character. Of the 2.25 years covered, we use<br />

the first 13 weeks (the last quarter <str<strong>on</strong>g>of</str<strong>on</strong>g> 1995) to initialize some model variables. We use the<br />

remaining 104 weeks <str<strong>on</strong>g>of</str<strong>on</strong>g> the period (1996 and 1997) for model calibrati<strong>on</strong>. The two different<br />

levels <str<strong>on</strong>g>of</str<strong>on</strong>g> data used (household panel and retail panel) are elaborated up<strong>on</strong> in the subsequent<br />

secti<strong>on</strong>s (Secti<strong>on</strong> 6.2 and Secti<strong>on</strong> 6.3). Secti<strong>on</strong> 6.4 describes linking these two different data<br />

sets in more detail. Secti<strong>on</strong> 6.5 deals with the representativeness <str<strong>on</strong>g>of</str<strong>on</strong>g> the resulting household<br />

analysis data set used in this research to answer the research questi<strong>on</strong>s. We c<strong>on</strong>clude this<br />

chapter by providing some general descriptive statistics <str<strong>on</strong>g>of</str<strong>on</strong>g> the analysis data set.<br />

6.2 <strong>Household</strong> Level Scanner Data<br />

<strong>Household</strong> scanner panel data <str<strong>on</strong>g>of</str<strong>on</strong>g>fers certain unique opportunities for understanding<br />

c<strong>on</strong>sumer behavior and deriving implicati<strong>on</strong>s for marketing acti<strong>on</strong>s (Gupta et al. 1996). To<br />

decompose purchase behavior into the choice (brand, store), purchase timing, and purchase<br />

quantity effects, analysis must be c<strong>on</strong>ducted at this disaggregate (household) level (Bucklin<br />

et al. 1998). The GfK household scanner panel comprises 4060 households. This household<br />

panel represents the Dutch populati<strong>on</strong> <str<strong>on</strong>g>of</str<strong>on</strong>g> households. A household is defined as <strong>on</strong>e or<br />

105

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!