There are two ways to forecast: based on statistics and based on opinion. What are the advantages of each? Eddie Davilla explains both options on managing inventory.
- Operations is about good planning. Operations is about being prepared to meet demand. To meet demand, you need to have the right people and the right materials. To meet demand, you need to have a good process, but to meet demand, you need to know demand. Most companies don't know how many customers they'll have tomorrow, much less next week, or next month. But in our modern world of electronic devices, social networking, and digital transactions, there's no shortage of data. Companies now have mountains of data to create sophisticated forecast based on what happened in the past. Quantitative forecasting relies on historical numeric data to predict what will happen in the future. And most people find comfort in this type of forecasting. It's unbiased. It's based on real data so we tend to trust it. But there's also something called qualitative forecasting. This relies on the opinions of people. This probably doesn't make you feel as comfortable, right? But, in the absence of historical data, it might be all we have. And believe it or not, under the right conditions, it can be extremely accurate, often much more reliable than quantitative forecasting. That seems odd because we've been trained to think that people are dumb and groups of people are supposed to be mindless mobs. That's not exactly accurate though. Most of you count on mobs to help guide your decisions everyday. Don't believe me? Well, tell me, how does Google generate their search results? They see what others with similar searches have clicked on. You aren't asking an expert to guide your internet experience. You're asking a group of people what they have clicked on. Their opinions and preferences are shaping your clicks and they seem to be doing a pretty good job. I mean, you and I keep going back to Google to search for almost everything. So, under which conditions can people make good decisions? There's a pretty good book called "The Wisdom of Crowds." It's written by James Surowiecki. In it, he tells all sorts of stories about how crowds make good decisions. One story is about a lost submarine. A naval officer puts together a team of people with all sorts of backgrounds: sub-salvage experts, naval officers, mathematicians. Each has some helpful knowledge. He individually presents these people with possible scenarios. He tells the people that whoever guesses the closest will win a prize. Each person, while isolated, talks about the likelihood of each scenario. The chief officer compiles each person's results and guess what? Two months later, the sub was found 220 yards from where the group had collectively guessed. So, what made that group so successful? Diversity of the group. All sorts of people with helpful bits of knowledge. Isolation. This empowered each person to think freely without fearing what others might think. And they also had the naval officer to sift through their responses. He was like Google. The point here is that companies have so many available tools, so many statistical formulas, lots of data, some based on fact, some based on opinion, and in the right hands, that data can guide us to helpful forecasts. And forecast don't just have to be for demand. We can try and forecast holding cost, ordering cost, the cost of materials and inventory, the speed of production. And as we try to make sound decisions about inventory, scheduling, waiting lines, or any other operations issue, we need to understand that accurate numbers help us formulate better answers to our problems. In business, almost everything that will happen in the future is unknown so having a firm foundation in data analytics and forecasting is vital.
- Understanding operations management
- Making key inventory decisions
- Balancing holding costs and ordering costs
- Choosing a production strategy and facility layout
- Managing waiting-line systems
- Defining quality and improving quality
- Managing business processes