Archives for category: Forecasting

In the book, How to Measure Anything, it is discussed how ranges are more preferable to a target number. I have had this discussion around annual planning time and I get blank stares. What I found was a better way, at least from my organization, to sell this concept and work through it with a team.

When working with a team around annual planning, we look at what the plans were for this year and what has worked and what hasn’t. We look at the gap from plan and last year and discuss what could have been done differently. This gives us some context on what should be worked on. The goal may or may not be known but either way we are focused on misses from last year.

The next step is looking at other opportunities. These are items that have not been covered or based on next year’s products or innovation items and processes that will be changed over the budget time horizon. These are captured with the current gap for further discussion.

Now that all the items are captured, a three dimensional matrix is developed. It is easy to hard, and simple to complex. There can be complex items that are easy because the organization is behind the initiative just as there are simple and hard items because though it is a simple item to fix, the organization isn’t aligned. The image of the process is below.

 

3 Dim Matrix

 

From this exercise the team now has a range of possibilities from Do it’s to moderate projects to major initiatives. This now becomes the range of values that the plan can encompass. What becomes the discussion now, is where in the range does the target land.

Does the target land within the range of values? If not, what needs to change on the financial side to fit it into the range? This can be a creative writing exercise more that hard science but at this point the dollars are a jumping off point anyway.

Does the target in the range but on the side of major initiatives? If so, how can the team makes some of those more doable? Is it by breaking them up into smaller projects? Is there a way to increase the impact of other projects? Is the target not realistic? That is a funny one I put in.

This has been a useful way to get leadership to understand and start thinking about ranges from the historical target focus. This may also giving the team the ability to treat the annual budget process closer to what it really is, an annual guess.

 

 

Advertisements

Nothing hurts plans and processes like change. Well, there may be another one. That is the illusion of knowing the future. Having a forecast is almost always helpful. It is when you believe that it is the future that it becomes an issue. One area that is very relevant is in sales and production organizations is mix and volume, especially when looking near the end of a product life cycle or limited time offers.

Mix can always mess up the works for a site or sales organization. When mix changes in manufacturing, it can mean that the products that were easy and high margin are now difficult and though the margin may seem higher, the conversion cost per unit or raw material costs go up as well. Can Continuous Improvement, CI, help? Maybe but it will also need engineering, CapEx, and other support to protect the facility from the next “great idea.”

Volume is another item that CI will need more support. When volume is greater than 120%, there are certainly CI projects that can help, especially related to Standard Work and basic troubleshooting skills that can ramp up new employees quicker than traditional OJT. What happens when you are 80% of plan? That is more difficult. The business needs to be flexible to the change in volume and CI projects should keep employees finishing projects and making improvements as long as possible before “right sizing” the workforce. This takes some discipline and rules for making the right choice.

Continuous Improvement can help mitigate impacts of volume and mix but the process needs a commitment to the process and help along the way when the market or the business changes. Changes come in various forms but volume and mix can make or break a site or process. CI will need all the organizations resources to move fast enough to mitigate changes and support the process when things look the worst.

One of the issues I frequently run into is when I discuss risk or future events or what impact a certain action will have, I get usually one of three answers. I am sure it will work, there is no way it’s going to make a difference or I don’t know. All of these options are weak compared to thinking probabilistically. Also, there are also other options not even discussed or considered. For the discussion here, I just want to focus on a future with only 2 mutually exclusive outcomes to make the process and concept easier. In future posts, I would like to explore this part of the process in more detail. I have put this into a pretty simple 3 step process.

Step 1 – Determine what you know about the subject and put a % towards what you think you know. You want to take some time with this to consider all the factors you know. If you don’t know anything or very little, use 50%. If you are very confident and have a lot of knowledge of the issue, may be you are close to 100% or 0% depending on what you think. This step should entail someone slowing down, stay away from the knee-jerk reaction and think about what you know, the circumstances of the issues at hand and the people at the table.

Step 2 – As information becomes available, modify your assumptions and potentially change the %. This means as time goes on, items related to the event or issues are changing, be aware of these changes and quantify them as best you can.

Step 3 – After the event occurs, review what you have learned. Would you have put the % initially knowing what you know now? This is not the statement; I would be 100% solid on this because I know who wins. It is more like what other information should I review before setting my starting point? What part of the process did I follow and what I didn’t.

I will be using the 2016 NBA Finals as an exercise using me as the protagonist. I will be the first one to tell you I don’t watch much NBA basketball but being from Cleveland and watching how they were beat by the Golden State Warriors in 2015, I wanted to see how well they would do and use my limited knowledge of the game to work this process out.

I want to make sure

Step 1 – Come up with an initial percentage. Let’s look at the Cav’s winning the NBA Finals. Some ground rules

If you know nothing about the subject, use 50%. This means that a flip of a coin is just as accurate as your guess.

Stay away from 0 and 100% unless you are absolutely certain of the outcome

For the Cavs vs. Golden State finals, before they started, I figured the Cavs had about a 35% chance of winning. Here was my thought process for this:

I worked from a 50%, flip of a coin, process because I recognized I don’t know much about basketball

From 50%, I knew the following items though the percentages are pretty arbitrary.

Golden state had home court advantage – about 10%

Golden State had the best record – about 10%

Golden State had just come back from a 3-1 deficit to win the west – about 5%. That gives the Cavs about 25% chance.

Cavs were rested and everyone was healthy – about 10%. That moves it back to 35% chance.

The Golden State Warriors win the first 2 games – That moved me to 10% from 35%. I just couldn’t see them winning it all though I believed they could win at least one game

Cavs win a game at home – big. Move up to 20%. This was an improvement over the 10% not so much that they won because I thought they could win one, it was more around how many points they beat them by.

Cavs lose at home – Less than 5%. At this point, the Cavs were down 3-1, the team had to win in Golden State twice and there had only been 2 other teams in history to get to a game seven and no one had ever won the Finals when they were down by 3-1.

Cavs win in Golden State – Move to 25%. This 20% move was due to the point spread in the win in Golden State, and the motivation and momentum for the Cavs. Also, the Cavs were coming home and I thought they could win game six as well.

Cavs win big in Cleveland Game 6 – Move to 35%. I only moved 10% because the Cavs still had to go back to Golden State and the starters played the whole game 6. The Cavs won big in this game but I figured that they didn’t leave anything in the tank for game 7.

Cavs at the start of the second half.  35% they were down by 8 at this point and though they looked pretty good in the second quarter, they were still losing.

Cavs end of 3rd quarter – 40% – The Cavs played great during the third quarter but they were still down by one. I could not help thinking that they had given everything they could and were still coming up short.

7 minutes left – 50%. It was a very tough quarter. Each team would have a lead and then they would be even up. This happened several times in the fourth quarter. Now someone may say that this means I didn’t know anything like I stated at the beginning of the discussion. This one is a little different because of the way I arrived at the number. I was now adding and subtracting percentages based on what I was sing. The fact that it came up to 50% was a coincidence.

5 minutes left – 50%. This was the hardest call. Both teams looked exhausted. Each team would drag down the court, wait until there was less than 10 seconds on the clock and then try to get the ball in the basket. There were errors and fouls on both sides and I didn’t make any changes. At this point, I thought either team could win.

50 seconds left – 70%. This jumped 20 points because of the Kyrie Irving 3 pointer with 50.6 s left. Not only were the Cavs up by 3, Kyrie still looked fresh where everyone else on the court looked punch drunk.

10.6 seconds left – 75% – This is when LeBron got fouled going to the basket. I only jumped it by 5% because LeBron was down on the court grabbing his wrist / elbow and was not getting up. Also, he may be unable to make the two free throws needed to put the game away. He missed the first one and then made the second. 4 point game but that is just a 3 and a foul away.

4 second left – 80%. I know, I know, there wasn’t much time left at this point but Curry had the ball and Kevin Love was guarding him. Kevin had done a great job earlier in this game as well as the regular season and some of the other play-off series but in this series, other than game 7, he has had limited impact. I was concerned he was going to foul Curry as he went up for a 3 pointer.

Buzzer – 90%. This is a joke, sort of. I still couldn’t believe that the Cavs had won. This is my point about personal bias. My bias based on all the history that I had been through with the teams of Cleveland; I was still struggling to take it all in.

After review, would I have done something different than the 35%? I saw at least a couple of predictions that stated the Cavs had a 25% chance of winning when I thought 35%. I don’t think I would have changed my percentage but I would do more independent research as well as reflect and discount my biases a little more but I think I would have posted a 35%.

Now think about this process and how you could use it as well as think about those around you. When you come up with an idea for something that is novel or unique and have to get people behind it, whether they know it or not, they will have to go through this process. It is an epistemological question. What does the individual need to convince him that this is the way to go? Sometimes they may not be able to tell you. Remember they are coming from a vested interest, naming themselves and it is hard to move away from you to improve your opinions, options and questions.

3 rules:

Review what you know and pick a percentage

As information becomes available modify your percentage

Once the event passes, review what you did and how you thought. What would you do differently and what did you leave out?

I have worked with a company related to TPM and engineering activities. While I was doing this, unknown to me, they were also moving from a localized forecast and scheduling model to a centralized model. I started asking questions when site after site I went to had inventory in the production areas, outside warehouses, and Lunch rooms. Well, the last one’s a stretch but anyway I started asking about the issue. I figured the move to central forecasting was obvious since everyone was filling up but I was surprised about the central scheduling part. Corporate governance was driving a schedule attainment metric (moving from a day to hours) may make sense in some situations but it was creating havoc in the warehouse and operation side.

I get trying to save money on SKU level forecasts out 12 to 14 weeks. Using the resources that are doing that should be used to find better signals and help the plant react or produce at the rate of consumption , not predict the future. But at least give the local sales and marketing group as well as scheduling the ability to change the schedule appropriate for what they are seeing. Taleb in Anti-Fragile, discusses this phenomenon in depth and what I take from it is two-fold. Some items are not scalable. Going from sales reps who are talking directly with the direct customers, they come up with fancy algorithyms. My first question is how much better is your system compared to naïve or other simple models run on a smart spreadsheet?

The other issue brought up by Taleb is when one makes a system bigger, the bigger the errors. This may seem simple but it actually gets worse mathematically as connections cross disparate fields. This makes risk management not just difficult but bordering on the impossible.

What is the knee jerk reaction to this inventory growth? I have seen the following:

Sales gets pushed to make the numbers either now or in future periods / quarters.

Operations get shuddered at some point due to low market demand to plan. Read Reality did not line up with the guesses.

CapEx gets squeezed, either from the infrastructure or productivity budget. The emergency CapEx balloons as the creative writing begins.

Discount the stuff that the customer didn’t want anyway.

Tweak the math in the black box so that it “better” predicts the future.

I have not seen a company dismantle the system they are currently using for forecasting for a simpler model. I have had sales / sites / and mid-level leadership to go do what they want instead of worrying about what is coming out of the black box.

I am a grown-up and know that companies, even small ones, have to spend some portion of their time forecasting demand. I would use the simplest model that got 60-70% of the demand and then use tribal knowledge or local expertise to tweak as see fit and let operations work with scheduling to make the most of their operations.