The first pull plan I ever ran was a mess. Twelve trades in a jobsite trailer, a roll of butcher paper taped to the wall, and a box of sticky notes that ran out halfway through. By the end we had a beautiful wall of yellow squares that told the story of how we'd top out the structure and get the building watertight. Two weeks later that wall was gone — someone had propped a window open and half the notes were on the floor. We'd built the plan and lost it in the same month.
That's the gap software fills in pull planning. Not the thinking — the trades still do the thinking — but the capturing, the connecting, and the keeping. Phase pull planning is where you get every trade in a room and work backward from a milestone to agree on how the work will actually flow. It's some of the most valuable coordination time you'll ever spend on a project. The trouble is that a whiteboard exercise is only worth as much as what survives the week after. This article is about how to run pull planning well and how the right tool keeps the plan alive from the trailer wall all the way down to the weekly work plan your foremen execute.
Pull planning, and why "backward" matters
Most schedules are built forward. You start at day one, string activities left to right, and hope the milestone lands where the contract says it should. Pull planning flips that. You start at the milestone — "dried in by June 12" — and ask the room a single relentless question: what has to be done immediately before that, and who does it? Then you back up one more step and ask again. You keep pulling activities off the finish line until you reach the work that can start today.
The reason this works isn't philosophical. When a trade foreman commits to a handoff date because he personally worked backward from the milestone and saw what depends on him, that commitment holds up far better than a date a scheduler assigned from an office. The people who will do the work define the sequence and the durations. Your job as the superintendent is to facilitate, not to dictate, and then to make sure the plan doesn't evaporate.
Do this at the transitions that actually matter: before a major phase kicks off (foundations, structure, envelope, rough-in, finishes), ahead of a hard milestone like substantial completion or a tenant turnover, and any time a big change — a design revision, a long-lead delivery slip, a phasing swap — knocks the project off its trajectory. You don't pull plan the whole job in one sitting. You pull plan the next chunk in enough detail to drive your six-week look-ahead.
Running the session so it's worth the room
Block two to four hours depending on how tangled the phase is. Rough-in for a hospital floor is not a two-hour conversation. Get the real people there — foremen and superintendents who can commit, not estimators standing in who'll have to "check with the field." A pull plan built by proxies is a wish list.
Here's a sequence that keeps a session moving instead of dissolving into an argument about who's slowing whom down:
- Frame the milestone (15 minutes). Say out loud exactly what "done" means and the date it's due. Make sure everyone agrees on the scope of the phase. Half of all pull-planning fights come from two trades quietly picturing different finish lines.
- Generate activities from the end (30–45 minutes). Each trade writes their own cards — one activity per card, a verb and a location, "hang and tape drywall, Level 3 north." Trades that don't write their own cards don't own their own dates.
- Sequence and expose the handoffs (30–45 minutes). Lay the cards out backward and connect them. The gold here is the handoffs — the moment work passes from one trade to the next. That's where jobs actually break.
- Assign durations and reality-check (20–30 minutes). Each trade owns its own numbers. Then add them up and see if the whole thing fits inside the milestone window. It usually doesn't the first time.
- Squeeze, don't wish (15 minutes). If it doesn't fit, decide together what changes — more crew, overtime, a resequence, splitting an area into two work fronts. Padding every duration by a day and calling it done is how you build a plan you can't hit.
A word on durations: work in whole crew-days, and keep a small buffer between trades where cleanup and inspection live. Frame-to-rough-in usually wants a day or two of slack so the electricians aren't tripping over the framers and so the inspector has something clean to look at. Buffers between trades aren't waste — they're what keeps one late trade from detonating the next four. But put the buffer at the handoff on purpose, visible to everyone, not hidden inside a trade's number where it quietly inflates the whole plan.
Sticky notes versus software — you want both
Don't let anyone sell you on either extreme. Physical sticky notes are unbeatable in the room: they're tactile, everybody can see the whole wall at once, and there's zero technology barrier for a 30-year plumber who doesn't want to touch a tablet. The energy of people standing up and moving cards is real, and it's hard to reproduce on a screen.
What sticky notes can't do is survive. They fall off walls, they don't distribute, and a photo of the wall is not a schedule — it's a picture of a schedule that nobody can update. The strongest approach for most teams is hybrid: plan on the wall with paper, then capture the result into a look-ahead scheduling tool that same afternoon while the logic is fresh in your head. That's where a location-based tool like LookAheadWall earns its keep — the handoffs and trade sequences you built on the wall become live activities and trade-flow connections you can actually maintain, not a photo buried in someone's phone.
What you're looking for in a tool is honest support for the way pull planning works: activity cards you can create and drag fast, organization by trade or by area (swim lanes), visible dependency links so a handoff isn't just an arrow you drew and forgot, and dates that recalculate when you move something instead of forcing you to re-do the arithmetic by hand. If capturing the plan is slower than writing it, people won't do it — and then you're back to a wall of notes on the floor.
From pull plan to look-ahead to the weekly work plan
This is the part that separates a pull plan that changes a job from one that decorates a trailer. The pull plan is your phase roadmap. It is not your execution plan. The bridge between them is the look-ahead and the make-ready process.
As the pull-planned activities roll into your six-week window, you screen each one for constraints — the seven usual suspects: information (approved shop drawings, RFIs answered), materials (on site or confirmed en route), manpower, equipment, prerequisite work complete, space (a work front that's actually available), and conditions/approvals (permits, inspections signed off). An activity isn't ready to hand a crew until every one of those is clear. That screening is the whole point of the look-ahead: you're finding the problems six weeks out, when there's still time to fix them, instead of the morning the crew shows up.
Only activities that survive make-ready — constraints removed, genuinely ready to build — earn a spot in the weekly work plan. That's the commitment your foremen make in short-interval scheduling: not "we hope to," but "we will, because nothing's in the way." When that discipline holds, your Percent Plan Complete climbs, and the reasons your commitments miss start pointing you at the real, repeatable problems instead of the weather. A tool that carries the same activities from pull plan to look-ahead to weekly plan without re-typing them is what keeps that thread from breaking between the trailer wall and Monday morning.
Where pull planning goes wrong
The failure modes are predictable, which means you can head most of them off:
- The key trade is a no-show. If the electrician or the mechanical foreman isn't there, you're guessing on the trade that usually drives the sequence. Reschedule rather than plan around a hole.
- One loud voice runs the room. Pull planning dies when the biggest personality dictates everyone else's durations. Facilitate deliberately — go trade by trade and make the quiet ones commit out loud.
- Activities too vague to build. "MEP rough-in" is not an activity; it's four trades and three weeks. If a card can't be handed to one crew for one clear chunk of work, break it down.
- Optimistic durations nobody believes. The number a foreman gives with the whole room watching is usually more honest than the one on a bid sheet. Push gently when a duration looks like a wish.
- The output gets lost. The one I opened with. Capture it the same day, or the plan quietly becomes a rumor.
It complements CPM — it doesn't replace it
Pull planning and critical path scheduling aren't rivals. Your CPM schedule holds the contractual dates, the overall structure, and the critical path the owner cares about. Pull planning fills in the trade-level logic and week-to-week coordination that a CPM schedule is too coarse to capture. The two have to agree at the milestones — if your pull plan says the structure tops out two weeks after the CPM milestone, you don't have a scheduling tool problem, you have a project problem, and pull planning just surfaced it early. That early warning is a feature, not a failure.
Remote and distributed teams
Not every team can get in one trailer anymore. Distributed pull planning works when everyone shares one live planning view on screen, video keeps it face-to-face enough to read the room, and the session gets captured so the foreman who couldn't make it can still see the logic and his commitments. Cloud-based look-ahead scheduling makes that practical — the same plan is visible to the field superintendent, the PM in the office, and the sub across town, and the mobile companion app puts each crew leader's committed work in their pocket when they walk the floor. It's not quite the energy of the wall, but it beats losing your best coordination time because two key trades are on different jobs that week.
Is it actually working?
Track a few honest measures instead of assuming. Did the phase finish the way you pull-planned it, or did it drift? Did handoff disputes and trade stacking go down? Did the pull plan make your look-ahead more accurate — fewer surprise constraints, more activities ready when their week came up? And ask the foremen plainly whether the sessions were worth their afternoon. If they roll their eyes, you're running the session wrong, not proving that pull planning doesn't work.
Done right, pull planning turns a wall of sticky notes into a commitment the trades made to each other and can be held to. The software's job is humble but essential: capture what the room decided, connect it to the look-ahead and the weekly work plan, and keep it current as the job moves. Get the people and the backward logic right, capture it before it hits the floor, and you'll spend a lot less time firefighting handoffs you could have seen coming six weeks out.