Slow down, build faster

Question: Let's say you're building 501 homes. There are 100 tasks per home. Each task takes 1 day, and you have 1 dedicated team for each task. On Day 100, the first home is complete, and you have 99 other homes following on in production. On every day after Day 100, another home is completed. How long until all 501 homes are complete?

Answer: 100 (for 1st home) + 500 (1 home per day x 500) = 600 days.

Question: You realize that quality is suffering because 1 day is too rushed for some of the tasks. For 50 of the 100 tasks, you double the duration to 2 days. You have 1 dedicated team for each 1-day task, and 2 dedicated teams for each 2-day task. Your first home now takes 150 days to complete. How long until all 501 homes are complete?

Answer: 150 (for 1st home) + 500 (1 home per day x 500) = 650 days

Amazing, right? Intuitively, you would think that adding 50 days to the production of 1 home would add a LOT more than 50 days to the production of 500 homes. How is this even possible?

The secret is that for every 2-day-duration activity, you always have 2 of these activities taking place at the same time. That means that every 2 days you complete 2 of these tasks, which is the same speed as completing 1 per day. So, yes, your 1st home will take 50 more days to complete, but every day after that you are still completing 1 more home.

Wow!

But it gets even better.

Let's assume that when you were enforcing the 1-day duration on all 100 tasks, 1 out of 5 tasks were done poorly and required 1 day of rework to be made good. On this project there are 50,100 tasks. This means 10,020 of them need rework. Your team averages 83.5 tasks per day so the rework will take 120 days.

Result? Your actual duration under the 1st scenario is 600 + 120 = 720 days.

If there is zero rework under the 2nd scenario, the longer build time per house gives a faster project build overall.

But not only that, it's better to have realistic planned durations and resource up accordingly. Otherwise, you're only fooling yourself. If you budgeted 2 days for a task, but find that it is regularly taking 3 days to complete, just accept the 3-day duration and ensure you always have 3 dedicated teams and 3 of these activities in progress at all times. Then you'll produce 3 completed tasks every 3 days which is exactly what you need to maintain your 1-home-per-day production rate. This is obviously better than having 2 teams trying to do the task in 2 days but failing half the time.

Yes, there is a tradeoff for allowing a longer duration per home: you will have more homes in production at a time. In our example we had 150 in production in the 2nd scenario, versus only 100 in the 1st. Or did we? Remember that in the 1st scenario, we had over 10,000 tasks to rework, so while we planned to have only 100 homes in production, we'll probably have closer to 200, if we count the ones that are “complete” but have a long rework list.

Takeaway: insist on 100% completion with zero defects for all tasks, and work backwards from this requirement to set planned durations. There is simply no more powerful way to hold a project timeline than by eliminating rework. Having set your durations, allocate more teams to longer-duration activities until you achieve the same daily production rate for every activity in the build. You will increase predictability, profitability and quality, while massively reducing stress. What's not to like?