Dec 1, 2016
My forecasting formula keeps getting it wrong
We often meet inventory planners who have tried other inventory technologies and forecast engines. Despite the bold claims these other applications tout, many planners have found them to be unhelpful and even downright incorrect.
Given all the tools available today, what’s the disconnect here? Why do these inventory forecasting tools continue to come up with sub-optimal order recommendations? Let’s dive in and see what the problems are.
Is there a perfect forecasting engine?
First, let’s dispel the myth that there is one forecast engine that will deliver perfect forecasts for all of your products all of the time. That, unfortunately, is a unicorn, and you’re not likely to find one anytime soon.
But before you give up on the concept of a forecasting tool, remember that going back to dumping data from your ERP or Quickbooks into a spreadsheet isn’t giving a forecast, either. You certainly can’t rely on checking every single record and calculation one by one, day after day. You still need a way to properly plan, an inventory forecast that will guide you to have the right amount of stock at the right locations exactly when you need it.
You’ll want to start with a reliable forecast engine that will create “reasonable” forecasts for you. That’s where a lot of tools stop, but there are more issues you may want to address. Many inventory-based businesses need a way to identify and prioritize the biggest problems and most rapid changes that the engine can’t cope with. Let’s go over a few key points.
Start with a top-down approach. Check the overall sales using the previous units sold at today’s unit cost, versus the future forecast also at today’s cost for all the items together. Does everything look good?
Drill down. Take a look at key product lines. Do you see anything abnormal?
Look for exceptions. Find the items where the upcoming forecast looks quite a bit different than the sales history. Use the value of the variance as a sort order – after all, if you buy too much of an item where the forecast was too high and it was cheap, does it really matter?
This kind of exception analysis and item testing is critical to really understanding what’s going on with sales and marketing, not to mention the importance of such activities to your business’s future in the market.
Incorporate the exceptions
Next, you have to be able to override the forecast when you’ve identified exceptions. You’ll also want to build forecasts for new items that don’t have any usable data to work off of.
You simply cannot manage your inventory properly and create optimal purchase orders without digging into these exceptions, which routinely arise. You’ll probably also want to do a basic exception analysis on items by comparing your sales for the month so far to the amount that had been forecasted to this point in the month. That might sound a bit simplistic, but if you try it, you’re likely to find a number of items that needed some extra scrutiny.
Be patient and reap the rewards
This is a process that does take time and focus, which is why it’s important to remember how relevant it is to your company’s bottom line. It is unlikely that there’s an another area in your business where new tools and technology can improve your business as much as re-assessing your inventory management process.
With the right tool and a little bit of attention, you will be well on your way to reducing your investment in stock while also improving customer fill rate. That might sound like a unicorn compared to your current process, but it is a real option that is available for the wide variety of SMBs that manage inventory.