How accurate does the forecast need to be?

addtoany linkedin

In getting ready for a trip I went into the drug store to buy travel sized toothpaste and contact lens solution. Looking at the packaging, I started to wonder how accurate a forecast needs to be. (You know you’re consumed with everything supply chain when that’s what you think about while shopping!)

I’m sure no one was predicting the need for these products in 100ml sizes a couple of years ago. And what if the airlines lifted the size requirement on liquids or reduced it to 50ml? What chaos would that cause the demand planners of the world? Walking to the front of the store I noticed some Olympic wear.

As you know, Vancouver just finished hosting very successful Olympic and Paralympic games. I could only imagine the heroics and horrors that were experienced to make these games the success they were. Everything from scheduling materials for the new venues to the clothing, flags, food and everything else required for the games. Will the promotions to sell off Olympic paraphernalia make up for the excess inventories now on the shelves and in the warehouses?

In a discussion thread on the supply chain expert community, Joshua Gao asked what your “Vision of the Supply Chain” is? Well, if we look to the past many things are different from our grandparents' supply chain. Two of the biggest stand out. First, customers are more demanding. I mean that in a positive sense, in that customers can quickly research products, understand trends in technologies and purchase what they want with a few clicks of a mouse. The second is that supply has become more fragile. Outsourcing, margin pressures and even catastrophic events can cause supply challenges. So this gets us back to the vision of the future and the question, how accurate does the forecast need to be.

In the past, good enough may have worked because there were fewer demand and supply pressures. But today and in the future, is it better to have an accurate forecast or should the focus be on handling the deviation? If the focus is to manage the deviation and leverage your supply chain as a competitive advantage, then how much effort should go into developing the forecast if you know it is going to be wrong anyway?

This is where it would be helpful to get your feedback since the answer may vary based on industry etc. Does the forecast need to be more accurate given the supply chain challenges of today or do you just need some number to start with since you will have to handle change regardless what the forecast states? How close does the forecast need to be, 40%, 60%, 80%?

Just one final request for feedback: if you were involved in any Olympic related supply chain stories, it would be great to hear them. Maybe your story will make the podium and win gold, silver or bronze!


Brad Blonkvist
- March 29, 2010 at 2:17pm
I think that is a great topic. Every supply chain oriented professional probably thinks about these issues at one time or another if not daily.

My limited experience with forecasting is that it needs to be repititious and provide some level of color commentary to the receiver. Forecasts that fluctuate wildly without explanation are quickly dismissed. In fact, proactively telling your suppliers we are doing x planning in 2 weeks and will probably change the forecast based on it is helpful information.

Precision is probably not in the cards for many forecasts. But sharing the thought process and leading indicators that indicate a need to change the forecast is critical.

The other observation is that the forecast needs to be what the market is indicating not your budget/sales target. This is obvious but often times not adhered to in a consumer products setting as it often requires a significant set of sign-offs to change the forecast. Sign-offs such as that can hide the true market demand.

Good luck
Puneesh Lamba
- April 01, 2010 at 4:13am
My two cents:
1. Forecasting for long term (more than a year) needs to be seen as something that provides patterns & general expected behaviour to us. This acts a guideline to supply chain profesisonals who then take informed decisions by doing mroe scietific research and data gathering. So the accuracy above 70% should be able to help
2. Forecasting for near term needs to be seen in combination of forecast+lead time impact. It means even if we are not able to forecast correctly, focus on reduced lead times will be able to nullify the forecast inaccuracy as your supplier as well as your inhouse manufacturing will be able to deliver products on much shorter notice and you will be able to handle demand volatility more efficiently (and with lower losts)
- April 03, 2010 at 6:11am
In my opinion any forecast is as acurate as closer you are at the customer and s better the "bonus-games" of the the sales heros and product- or MKT-managers are known :-)
To be serious: In my former company I was the Director Operations and responsible for manufacturing, engineering, purchasing, planning ( and some other stuff).
We, as the manufcaturing site controlled as well the european central warehouse stock on an daily base. We gathered all informations from the internation PMs, looked at the history of their clients, talked frequently to R&D for new and enhancement projects to the actual portfolio, had frequent meetings with the purchasing managers about leadtimes and balanced all the informations against the stock and the production time. With approx 10.000 enditems and 3000 RMs we were able to hit a delivery goal of more than 90% and, as well very important could balance our production capacity based on the actual needs.
Such achievements work only if there is a dedicated Forecast-analyst who is linked to production and not to the product managers or MKT-departements
Christopher Lim
- April 06, 2010 at 2:04am
Forecasting should be minimized as invariably, they may turn out to be inaccurate. I know there are scientific methods that can minimize the variability in forecasting but there is still that uncertain factor especially when rules or policies are changed. I still believe in a responsive supply chain. Suppliers producing to actual demand as fed thru. Flexible yet high manufacturing throughput as well as transit time velocity is key to giving key competitive advantage. Suppliers should produce to demand in as low an MOQ as possible. A tall order but this practice has been done many times in the automotive industry and I believe the knowledge and know-how can be customized to fit all other industry as well. In fact, if you recall Dell has done this very well in years past. Some of that knowledge and practices could still be used to improve the supply chain velocity, flexibility and responsiveness.

Leave a Reply