The encyclopedia · Marketing & Brand · Marketing decision · 2001–2003
Rhenania rose from #5 to #2 in German mail order by optimizing catalog mailings
Dynamic multilevel modeling set how often catalogs should go out; Rhenania overtook rivals within a year and bought two competitors.
Rhenania
the move
Rhenania, a German direct mail-order company, was spending heavily on catalogs without knowing whether each mailing paid for itself. Its market position sat at number 5, and the reflex response to weak sales, mailing more, was burning money.
Researchers built a dynamic multilevel modeling (DMLM) approach that uses response elasticities to compute the optimal frequency and targeting of catalog mailings. Customer segmentation governed which customers received which catalog, and RFM analysis combined with a CHAID algorithm flagged customers who should receive a reactivation package instead.
Within one year, Rhenania moved from number 5 to number 2 in the market. The model was so effective that the company acquired two competitors, one a subdivision of Springer Verlag. The follow-up Marketing Science paper, which won the 2003 ISMS Practice Prize, reports the same rise and acquisitions.
why it works
- Optimizing mailing frequency stopped paying for mailings that did not pay back.
- Segmentation aimed each catalog at the customers most likely to respond.
- Reactivation packages converted dormant customers that another catalog would have missed.
what transfers
Frequency is a decision, not a default: measure the response elasticity of every mailing wave and let the model set the cadence, as Rhenania went from number 5 to number 2 within a year.
what came after
Rhenania's climb made it strong enough to absorb two rivals. The DMLM work won the 2003 ISMS Practice Prize and became a reference case for treating direct-marketing frequency as a mathematical optimization.
references
- Optimizing Rhenania's Mail-Order Business Through Dynamic Multilevel Modeling — Interfaces (via RePEc)
- Optimizing Rhenania's Direct Marketing Business Through Dynamic Multilevel Modeling (DMLM) — Marketing Science abstract
spotted an error? The archive wants to know.