Welcome back to Beyond Deadlines newsletter, a free perk for people looking to improve in Planning and Scheduling. Each week, we provide tactics, prompts, jobs and food for thought. We want you to succeed today, tomorrow and throughout the rest of your career.

Most schedulers read total float one activity at a time.

You click an activity, see 47 days, and move on. Fine, it's not critical.

That's the wrong unit of analysis. Float is a distribution, not a value. Sort the whole schedule by total float and the bands tell you where your logic is broken.

Here's the triage table. Working days, and these are diagnostics, not laws.

≤ 0 days. Critical, or sitting behind a constraint. Trace and explain every single occurrence. No exceptions.

1 to 10 days. Near-critical. This is your watch list, and it's where next month's critical path is already hiding.

10 to 30 days. Comfortable parallel work. Leave it alone.

30 to 60 days. Decoupled parallel work. Usually fine.

60+ days. Either genuinely parallel scope or missing logic. You have to look.

90+ days. Almost certainly a wiring gap. First question: does this activity have a real successor?

That last band is where the money is. An activity with 90 days of float is telling you one of two things. Either it's truly independent of everything else on the job, which is rare on a real project. Or nothing downstream depends on it, because you never tied it to anything.

Large float across many activities usually means missing logic, not spare time.

Once you've got your bands, don't start fixing. Sort the offenders by root cause first, because each cause has a different fix.

Procurement buffers. Deliberately off-critical. Leave them. Procurement float is earned float, and tightening deliveries just to make the float numbers look busy is how you end up storing switchgear you can't install.

Genuinely flexible scope. Leave it.

Missing crew-continuity ties. Fix with logic. Tie unit n to unit n+1 so the model stops assuming one crew can be everywhere at once.

Under-tied utility and commissioning interfaces. Fix. These are the ones that bite at the end.

Mis-modeled roll-up tasks. Rework them.

Scope waiting on external input. Flag it. Don't guess.

Six causes, six different responses. A blanket "clean up the high-float activities" pass treats all of them the same and breaks the three you should have left alone.

Never fix float values with constraints.

A constraint that pulls a 90-day-float activity back into range hasn't fixed anything. It's hidden the missing successor and added an artificial pin your next negative-float investigation will have to strip out.

Fix float with defensible real-world ties. Or accept it as true and say why.

Float isn't slack you get to spend. It's the model reporting on itself.

Read the report.

Wispr Flow turns your voice into clean, professional text you can send the moment you stop talking. Not rough transcription you have to clean up. Actual polished text — ready for email, Slack, or any app.

Speak the way you think. Go on tangents. Change your mind mid-sentence. Flow strips the filler, fixes the grammar, and gives you text that reads like you spent five minutes writing it.

89% of messages sent with zero edits. Millions of professionals use Flow daily, including teams at OpenAI, Vercel, and Clay. Works on Mac, Windows, and iPhone.

Start flowing free

Claude CoWork

Most people still use AI like a faster search box.

Ask a question. Get an answer. Paste it somewhere else.

That's assistance.

Anthropic just released Cowork, an experimental product that pulls Claude out of the chat window and puts it to work across your files and your connected apps.

Multi step tasks. Actual work. Not just answers.

That's delegation.

And the difference matters more than it sounds.

"How should I analyze these reports?" is a question.

"Analyze these reports, compare them, flag the major issues, and build me a summary" is an assignment.

One is a tool.

The other is closer to a teammate.

Why this should get your attention

Construction is almost perfectly built for this.

We generate mountains of information every single week. Schedule updates. Daily reports. RFIs. OAC minutes. Procurement logs. Change orders. Cx documentation. Progress photos. Twelve spreadsheets nobody owns.

The problem was never a lack of data.

The problem is connecting it fast enough that the decision still matters.

So picture handing over a folder. This week's update, the OAC minutes, the procurement log, and the field progress reports.

Then giving it the assignment:

Review these files. Flag anything threatening the next six weeks. Compare those risks against the critical and near critical paths. Build me the issues list for tomorrow's planning meeting.

That's not a smarter search box.

That's a first pass analyst who never gets tired of opening files.

The scheduler isn't going away

I deployed GPT-4 across a portfolio of 4,998 projects and a $6 billion budget.

Here's what that taught me: AI is not replacing the scheduler. It's making the scheduler faster, sharper, and more valuable.

But it absolutely raises the expectations of the role.

Right now too many planners spend eight hours gathering information and two hours thinking about it.

Flip the equation.

Let the machine do the gathering, the sorting, the comparing, and the first pass.

You spend your hours challenging assumptions, walking the site, testing scenarios, and helping leadership choose.

Show me a tactic and I'll replace it with a process. Show me a process and I'll replace it with a system. Show me a system and I'll replace it with automation.

That ladder just got a lot shorter.

The advantage won't be knowing how to use AI.

Everyone is going to know how to use AI.

The advantage will be knowing what to hand off, and what still requires construction judgment.

Software can be learned in weeks. Judgment is built over years.

The best planner of the future may not be the one who can personally process the most information.

It'll be the one who builds the best system for turning information into decisions.

So here's my question for you:

What's the first task on your desk you'd hand off completely, if you trusted the output?.

Director, Planning and Scheduling

Area Scheduling Lead

Integrated Planning Site Lead

^{We have no connection to these jobs or companies. Our goal is simply to help you land the job of your dreams.}

This week’s episode we dive into building portfolios. Watch or Listen on Spotify, Apple Podcasts or Youtube.

Thank you for reading.

Check out Micah, Greg and Beyond Deadlines on LinkedIn.

Sign up for our LinkedIn newsletter.

Subscribe to our YouTube Channel.

Listen to the podcast on Spotify or Apple Podcasts. Visit our website.

See you next week,