Are your sales forecasts causing an "earthquake" in your manufacturing operations?

Add bookmark
Seth Marrs
Seth Marrs
05/27/2014

The boundaries between tectonic plates are the most volatile places on earth. The friction between them has taken millions of lives and wreaked incalculable levels of damage. Yet, ironically, without this friction the earth wouldn't have evolved into the habitable planet it is today.

Sales and operations function at the business equivalent of the boundary between two tectonic plates. The friction between these two teams can create problems, but it also enables the business to evolve.

See if you can relate to either of these scenarios:

Scenario #1: Your sales team is confident in the strength of your market and is chasing a number of leads that they believe will turn into business. On the basis of those prospects, they predict that they will need X amount of units.

Your manufacturing team takes those predictions and uses it to determine how much product needs to be created. But suddenly, a lot of those strong leads don’t turn into business and you’re left carrying a lot of unsold stock.

Scenario #2: Your sales team underestimates the level of demand for your product. Suddenly, customer orders are flying every which way and manufacturing has no way of keeping up. Customers are unhappy as their orders go out late, everybody is stressed and the finger pointing starts about who is to blame.

In an ideal world, neither of these scenarios would exist. Sales would be able to exactly predict their future sales and match them to the manufacturing product build cycle. Manufacturing would, in turn, be able to deliver product in a steady flow that directly matches market demand.

But, in reality, many supply chains ignore the sales forecast when doing material planning. They've been burnt by awful sales predictions and then blamed for not having enough stock to meet demand. So they keep asking sales for numbers but only use them to supplement build planning.

So where does it go wrong?

Sales is good at understanding products, the value proposition, developing customer relationships and closing deals. But they are not fortune tellers.

In most cases visibility into the sales cycle is too short and also requires historical trending analysis to get to as accurate an overall number as possible.

And trending is not a core competency for sales. Data analysts are best positioned to accurately handle trending requirements.

To make matters worse, sales cares about revenue whereas supply chain needs units by product. So what was already a complex data task has now become even more challenging. This problem repeats itself in businesses all over the world.

In the upcoming Manufacturing Process Excellence conference this June in Chicago I’ll be discussing this friction in more detail along with potential solutions to help improve these dynamics.

Here are three key areas of focus:

#1: Don't let your sales staff do the trending:

You will likely have a gap between sales cycle and manufacturing build time which will require you to use trending. Put this task in the hands of a data analyst. While you’re at it, move manufacturing trending to this person as well. He or she can then work on sustainable models to better predict the future and use sales and operations as an input rather than a source.

#2: Stretch the sales cycle while shrinking build time:

Maximum visibility into your sales cycle will increase your accuracy as sales are experts in their deals so are much more accurate at forecasting than trending.

#3: Iterate the model:

Once you have a solid model in place keep comparing it to the adjustments you're being asked to make. At a certain point the model will evolve into a more accurate predictor than a manually adjusted forecast. This will give sales more time to sell and manufacturing better visibility to what they need to build.

But what do you think? Have you encountered similar problems in your company? How have you helped manage the friction between sales and manufacturing?

[eventpdf]


RECOMMENDED