Marginal Analysis of Project Schedules

May 11th 2018

A friend of mine who also does contract engineering was mentioning that a customer of his was getting rid of their engineer. Of course, I mentioned that if they needed more help to let me know. He said well they are just about out of money and have almost missed their market window, so they can not afford you. This is a common in problem I see with companies, not just startups, where they do not understand how to do some very simple and basic math which will help them stay on target and get products out the door.

Most companies’ setup Gannt charts and budgets for projects and then try their best to do Taylor’s method of project management. See https://en.wikipedia.org/wiki/Frederick_Winslow_Taylor

The problem is that these tools do not always help the guys doing the work do make/verse buy decisions. For example, I worked on a project in the past where they said “time and quality is everything, we do not care about the costs.” When it came time to order circuit boards the manager decided to save the company money and order the boards with several weeks lead time, when the project was already behind schedule and the boards were on the critical path. The manager was doing what he felt was best for the company, however this decision most likely costed the company more than ordering boards quicker turn around time. 

The key to making better decisions in projects is to understand some basic metrics for the project. We often hear that for projects you can pick two: Good, Fast or Cheap. Everyone wants all three but in reality, if you get two out of three you are doing better than most companies.. Even when people say they want good and fast, they can not help but include cheap, for example the manager above. Therefore, when projects start I like to ask a simple question. How much is it worth to change the schedule by 1 day? If we were 1 day late in the project what does it cost the company? If we are 1 day ahead what does it provide to the company? This marginal analysis of projects is very insightful and can make some decisions trivial.

One company I worked for was doing a cost reduction on their product. They were reducing costs by $5 through redesigning the board with new parts. They were estimating that by time the cost reduction was done they would be shipping 1 million units a year. Some quick math revealed that each day the cost reduction project was not in production would end up reducing the company’s profit by $13,699. So if you could spend $10k to reduce the schedule by 1 day it would be worth it. Armed with this simple metric the managers and engineer could make some  decisions very easy. For example, should I order boards with 2 week turn or 1 week for a $5k price difference? Well if you have nothing else to do and need those boards then ordering with 2 day turn for an extra $15k would be worth it to knock 3 days off the schedule.

That company knew their volume and cost saving, so the marginal cost was easy to calculate. This is not always the case. For example, if you are building a new product and not sure what the volumes will be it is harder to estimate the value per day. However, you can always estimate the cost per day for a delay. That is assume you have 5 engineers working on a project, and their full burdened cost to the company is $150k per year for each engineer (this includes office space costs, insurance, taxes, etc). Then the cost of the five engineers per day is $2,000. Note that it costs the company $2,000 per day to design and build the product, so the value to the company for the product has to be greater than $2,000, for company to be profitable!  I like to believe that the product value should be double the engineer costs, but this varies with companies. Either way if you can spend $2k to save a day it is worth it, because it not only saves a day of engineering but also means engineers can take that day and work on a new product, so your payoff is doubled.

This marginal analysis of the cost by day is very valuable, however keep in mind that the value per day might change over time for projects. Often at the end of a project there is a trade show or customer deliverable and thus the perceived cost per day goes up. Although this happens, it is a sucker’s game as the value for a day at the end of a project is actually the same value per day as at the start of a project. Both will save you the same amount of time, however it ‘feels’ different at the end of a project as so much has already been invested.

So the next time you are stressed over a schedule or deliverable, ask yourself what you could have done earlier in the project to prevent the problem today.   After consider the costs of further delays in the project you might find that adding experienced contract engineers to help is well worth the expense.