Value creating promotions you and your supplier


We know how it is. You're sitting and trying to figure our what tactics and discounts that optimise your promotion campaign. Or at least you want to get ballpark figures as you're about to enter the meeting with your supplier. Hopefully you can convince them that it's a good business case for them to spend a significant chunk of money on this particular campaign. 

Below, you see a diguised example: 

•7 campaigns for the same product
•The campaigns have different price discount and different tactics, with discount rangin from 20% to 40%, 3 for 2's and more

So, what campaign offer should I choose in order to maximize sales?


Actually in this case you might be lucky to find some pretty decent campaign to draw information from. A 40% discount seems to lift unit sales by a large factor. The tricky thing is that even large lifts might cause negative profit due to 1) lower margins caused by the discount 2) cannibalization and stockpiling of the products (the customer switches from another brand to this one, or stockpile so the won't buy next week when the regular price is back). 

So what to do?

What you need is an algorithmic approach, that can capture both issues outlined above. Moreover you need to take into account weather and seasonality factors. What is also needed when you forecast products that do not have much historic promotions, is the ability to use information about how similar products behave. This means using all the rich information you have in your  sales and product data. 

Once you have that in place you are ready to go and meet your supplier, creating a win-win as you can outline the promising outcomes of your optimised campaigns, strengthening your business case.