Predicting Lateness

People don't know that they're going to deliver late. I've worked with multi-billion dollar companies and individuals, and often when someone says "I can deliver that on-time," they genuinely believe it. As project managers, it's important for us to interrogate that assertion and break down the path to completion with enough detail to see when tasks start to slip.

Realizing that you're not going to finish that project when you promised. Edgar Degas

I've written around this issue before, but haven't specifically addressed it yet:

• So close: in my "Ideal Jobs and Superpowers" post, I wrote, "We could know with precision how on-time or late we were multiple years out, and pinpoint it to specific part numbers," but not why that was unusual.
• Self-deception in scheduling (more on this below) in "Musings on Procrastination"
• Applying the tools in "Integrated Master Schedule": "When the supplier reported to me...I would adjust the start and finish dates for the current task based on what they told me for the in-work task... This allowed me to use the latest information from the supplier, but not require me to trust their final delivery dates: they are often much less reliable than the dates of the in-process work.

Too Big to Think About

Landing gear has extremely long lead times: a "hot" production line can still has a lead time of two years from order to first delivery, and even longer for a new product where design and qualification are required. A large structural part might have a lead time of 6-12 months (not including raw material). For these "long-lead" parts, at the time of order placement the supplier does not know if they will be late. Although these are not tiny companies, they are still light on office staff and project managers. I've seen these schedules laid out in Excel (with varying formula quality), but they rarely account for every operation and almost never include competing needs for those resources (which means more waiting time). When looking at a year-long schedule, it's common (and understandable) for a person to say to themselves, "Of course we can deliver! What moron couldn't deliver a big piece of steel in 12 months?"

But history has shown that intelligent and dedicated people can take far longer than 12 months to deliver a big piece of steel. Parts queue up in front of critical machines, the operations themselves take time, quality issues crop up and need to be resolved, outside processors (heat treatment, paint, anti-wear coatings, etc.) get full and can't meet turn-around times... the schedule is complex and full of opportunities to lose time, and it's impossible for anyone to keep all of the needed information in their heads. People also can be overly-optimistic about the ability to make up time (more on this below): "We'll just add a third shift," or "We'll pay to expedite this through the processor," might save the schedule, but often just prevents it from slipping further.

"Too big to think about" calls for coaching the experts into developing a detailed-enough schedule. If they can lay everything out from contract award through delivery and put durations for each task, a project manager can track status for each. As I discussed in "Integrated Master Schedule", that initial lay-out plus the expected finish date for the current process step yields more accurate answers than relying on someone blithely saying "we'll make it up later with overtime."

Over time, this method gets more accurate: for the second, third, or fourth production lots, the project manager can refer to the actual process steps and durations from the preceding production lots and use those to build the schedule. Do this in collaboration with the shop managers or supplier contacts: when they say "It'll take me X weeks," reply, "It took you 1.5X weeks the first time. I want to believe you, so show me on the previous-lot actuals where we're going to save time." This engages the detail-level thinking that hopefully does reduce the schedule duration, but at the least it forces the supplier to confront their own assumptions and enables them to think in chunks that aren't incomprehensibly big.

Self-Image and Self-Deception

In the October 19, 2025 episode of the Writing Excuses podcast, the hosts discuss how important it is to tell people that you're going to be late: others have held time in their schedules for you, and can shuffle that time around if given enough warning but not at the last minute. Writer Erin Roberts had a great way of describing why people don't know they're going to be late until they are:

What happens is you get into a cycle of, like, optimism and shame... You wake up optimistic that today you will suddenly write 10,000 million words. Like, cause you’re like if I just get in the zone, if I just do everything perfectly, it’s going to be fine. It’s going to be fine, fine, fine. And then as the day goes on, you’re like, oh, my God, it was not fine. Things happened. I needed to eat lunch at one point, why did I do it? And then you think, oh, my gosh, I’m so ashamed, I don’t want to say I’m failing. Maybe tomorrow I’ll fix it, and I’ll be the person...I’ll be the best million-percent version of myself. And I think you can get into that cycle until the point that you actually hit the deadline. At which point, then you’re sending like really sad emails, being like I don’t know, I thought I was going to do it. [Emphasis added]

"Optimism and shame" hit me as so familiar. We like to think of ourselves as capable people who can make our commitments happen, and overestimate our ability to make up for lost time (optimism) and then don't want to share when we're still struggling (shame). Project managers need to show empathy to others (and themselves) to reduce that shame, and gently challenge the optimism. This is where risk buffers built into the schedule can be emotionally helpful: if you can tell a person "look, I'm not slipping the whole schedule, I'm just showing our risk buffer going down and we can add the time back if you finish quickly," it lowers the burden of communicating and might get better assessments.

This dovetails with "Musings on Procrastination" because the procrastination is driven by both optimism and shame. As I quoted Peter Fenzel in that post,

It requires the self-forgiveness to recognize that it's not going to get done immediately: that you're going to only to sit down and do part of it [shame], you're not going to be able to sit down and do the whole thing and don't get so "up" about what you think of what you're capable of that you're just going to be able to bang it out [optimism].

Lateness can feed back on itself into procrastination and yield even more lateness. This is atypical in manufacturing because the process is more known and structured, but I imagine is common in creative and thought-intensive projects.