I have data for 4 years with variables year, month and sales volume only. I want to find the base volume and incremental volume. I want to use Marketing Mix Modeling. Can anybody tell how to define model equation with these variables and how can I use SPSS/R for the same ?

Is it possible to build model with only these variables?

Is there any other way or statistical method to find these volume ?

Please reply...

Thanks in advance....

  • $\begingroup$ Thanks for your reply...I came to know about this model from the this link en.wikipedia.org/wiki/… ... I think I can use this method to find the incremental and base volume....Please let me know If you need any other informations $\endgroup$ – Arushi Dec 26 '13 at 11:08

No, it is not possible to build such a model from the data you have (year, month, and sales volume). You did not list any information about the actual marketing mix (advertising, promotion calendars and levels, distribution, price). Plus, in most cases you will need finer time granularity (e.g. weeks), but that depends on the calendars.

You might want to start with Abraham and Lodish's work, published in Marketing Science in the 1980s and work forward using a Google Scholar search.

  • $\begingroup$ Thanks for answering...I can have month and week also in which promotions were there...Then can you please tell me how to proceed? $\endgroup$ – Arushi Dec 26 '13 at 11:04
  • $\begingroup$ I can't be specific about SPSS/R. Deseasonalize. Dummy out promo periods (plus post-promotion dip if present). For remaining periods, run a smoother (e.g. exponential smoothing) to get base. Consider whether you run this forward only, or forward and backward and then average the two. Forecast through the promotional periods. Reseasonalize. That's your base sales. Incremental = actual - base. $\endgroup$ – zbicyclist Dec 26 '13 at 14:49
  • $\begingroup$ Thank you so much. I like the approach...Can you explain this line "Forecast through the promotional periods "... I don't have much knowledge of statistics, that why asking to elaborate it more...Sorry for the inconvenience...Pls reply $\endgroup$ – Arushi Dec 26 '13 at 15:23
  • $\begingroup$ The exponential smoothing model will allow you to forecast. Because you will have eliminated seasonality and trend will be very slight during the short time periods, the forecast will essentially be flat. $\endgroup$ – zbicyclist Dec 26 '13 at 18:11
  • $\begingroup$ "Dummy out promo periods (plus post-promotion dip if present)" can you please explain this line..? $\endgroup$ – Arushi Mar 14 '14 at 16:16

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.