I am a marketing intelligence student, going an internship in B2B company. Over the years the company has saved a lot of customer and sales data. One of my assignments is to make projections of their future sales. For each customer: when + what + how much will they probably order?
They want to turn these projection into (customer and item specific) graphs of future orders, and use these as a base for purchasing and production: “When do we need what orders to be ready?” and as a tool for sales reps: “Which customers should have ordered by now?”
When I’m finished making the projections, I should tell IT to change the CRM system in such a way that it automatically, at the end of the year, creates the projections for the following year.
They asked me to make the projections based on their order history (date + item(s) + quantity). Other explanatory variables that I could possibly include are:
- Business type (is the customer a distributor or end-user)
- Application category (how the customer uses the productL 7 categories)
- Location (zone: country and city of the customer)
- Account manager (our employee responsible for the account)
- Paying info (pre-paid, post-paid, etc.)
- (I could also combine spreadsheets to add their sample history: when did they receive which product sample(s))
I really have no clue where to start with this part of my assignment. I tried chopping it up in smaller pieces and just start with yearly projections: “how much do we expect customer X to order or product Y in 2017?” And then turn that into monthly projections later, but even with this yearly projection I don’t know where to start.
Therefore, my question is: how to do this? Where should I start? Are there specific modelling techniques that I should look into? And maybe: is it realistic that I, as a marketing intelligence intern with an average brain, work out this module within 3 months?
Looking very much forward to hearing from you.