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

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

categories revealed that the str<strong>on</strong>gest (average) effects are: extra promoti<strong>on</strong>al volume (ranging<br />

between categories from 16% up to 96 %), followed by brand switching (ranging between<br />

categories from 7% up to 47%), and purchasing so<strong>on</strong>er (ranging between categories from 0%<br />

up to 21%).<br />

The promoti<strong>on</strong>al increase in unit sales was decomposed in two different ways. The<br />

first decompositi<strong>on</strong> is based <strong>on</strong> distinguishing between purchases that would have taken place<br />

anyway at this moment, or purchases that are changed in time. The current additi<strong>on</strong>al<br />

expenditures are <str<strong>on</strong>g>of</str<strong>on</strong>g> interest, accounting for about 2/3 <str<strong>on</strong>g>of</str<strong>on</strong>g> the extra promoti<strong>on</strong>al unit sales. The<br />

sec<strong>on</strong>d decompositi<strong>on</strong> distinguishes between purchases coming from regular versus n<strong>on</strong>regular<br />

c<strong>on</strong>sumers. This different point <str<strong>on</strong>g>of</str<strong>on</strong>g> view showed that about 2/3 <str<strong>on</strong>g>of</str<strong>on</strong>g> the unit sales<br />

increase is due to brand switching, coming from n<strong>on</strong>-regular c<strong>on</strong>sumers. <str<strong>on</strong>g>Promoti<strong>on</strong></str<strong>on</strong>g>s do turn<br />

out to be quite effective in drawing c<strong>on</strong>sumers from competitive brands.<br />

What is the c<strong>on</strong>tributi<strong>on</strong> <str<strong>on</strong>g>of</str<strong>on</strong>g> decomposing the promoti<strong>on</strong> unit sales effects using<br />

household level data instead <str<strong>on</strong>g>of</str<strong>on</strong>g> store level data? The different findings dealt with in this<br />

secti<strong>on</strong>, based <strong>on</strong> the two different ways <str<strong>on</strong>g>of</str<strong>on</strong>g> decomposing the promoti<strong>on</strong>al effects, can <strong>on</strong>ly be<br />

made explicit using household level data. These micro level data <str<strong>on</strong>g>of</str<strong>on</strong>g>fer a higher level <str<strong>on</strong>g>of</str<strong>on</strong>g> detail,<br />

leading to a direct understanding <str<strong>on</strong>g>of</str<strong>on</strong>g> for example choice behavior, and the relati<strong>on</strong>ship between<br />

purchase behavior and c<strong>on</strong>sumer socioec<strong>on</strong>omic characteristics (Russel and Kamakura 1994,<br />

Bucklin and Gupta 1992). Bucklin and Gupta (1999) stated that research is needed to<br />

develop simple, robust models that will provide better estimates <str<strong>on</strong>g>of</str<strong>on</strong>g> promoti<strong>on</strong>al sales that are<br />

truly incremental for the manufacturer. Van Heerde et al. (2002) decomposed the sales<br />

increase due to sales promoti<strong>on</strong>s using a unit sales approach, based <strong>on</strong> store level data,<br />

although they state that household data provide the best opportunity for decompositi<strong>on</strong>. But,<br />

store level data are far more likely to be used by managers and they are also more<br />

representative. A disadvantage <str<strong>on</strong>g>of</str<strong>on</strong>g> using store level data instead <str<strong>on</strong>g>of</str<strong>on</strong>g> household level data is that<br />

it is impossible to recover the individual purchase behaviors that underlie models <str<strong>on</strong>g>of</str<strong>on</strong>g><br />

household behavior (brand choice, purchase timing, and purchase quantity decisi<strong>on</strong>s) (Van<br />

Heerde et al. 2002). Their unit sales decompositi<strong>on</strong> distinguishes three sources: cross-brand<br />

effects (the units that other brands lose at the time <str<strong>on</strong>g>of</str<strong>on</strong>g> the promoti<strong>on</strong>), stockpiling effects (sales<br />

that are shifted from other weeks to the current week), and category expansi<strong>on</strong> effects (the<br />

remainder <str<strong>on</strong>g>of</str<strong>on</strong>g> the unit sales increase). Unit sales that come from households that accelerated<br />

190

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

Saved successfully!

Ooh no, something went wrong!