Integrated Project Delivery sounds great in the contract room. Owner, architect, and builder share one risk pool, sign one agreement, and everybody's incentives finally point the same direction. Then you get to the field and discover that a shared contract doesn't automatically produce a shared plan. The owner is still late on finish selections, the design team is still cranking out RFIs, and your drywall sub is still standing around because the ceiling grid ahead of him got hung wrong. IPD gives you the legal framework for collaboration. It does not give you the operating system. That's what the Last Planner System is for, and it's where scheduling software actually earns its keep on these jobs.
I've run both traditional design-bid-build and a couple of genuine IPD jobs, and the difference in how planning has to work is bigger than most people expect going in. Below is what actually changes on the ground, and how to run Last Planner so IPD delivers what it promises instead of just costing more in meetings.
Why IPD and Last Planner Fit Together
The Last Planner System was built on a simple idea: the people who do the work should be the ones who plan it, and a schedule is only worth anything if the promises in it are reliable. That maps almost exactly onto what IPD is trying to do. Both assume the master schedule is a target, not a script. Both assume the real coordination happens week to week among the people holding the tools. And both live or die on trust.
The practical link is the commitment. In a hard-bid job, a sub who misses his date argues about it, files for delay, and protects his margin. In IPD, that same sub is sharing the risk pool, so a broken promise costs him directly. Last Planner is the machinery that turns that shared incentive into a weekly cadence of reliable promises. IPD gives people a reason to make good commitments; Last Planner gives them the structure to actually make and track them.
Pull Planning With Everyone in the Room
Pull planning is where IPD changes the most. On a normal job, the superintendent and the trades build the sequence backward from a milestone and the owner and designer find out later. On IPD, the owner and the design team are in the pull-planning session, and that changes the sticky notes on the wall.
Run it the same way you always would: pick a milestone, work backward, one activity per card, and let each trade declare what they need from the trade before them. The addition in IPD is that design deliverables and owner decisions become cards on the wall too. When the fire-alarm sub says "I can't rough-in until I have the approved device locations," that approval date is now a card owned by the design team, sitting right there in the sequence with a name and a date on it. That's the whole point — you're making the information handoffs as visible as the physical ones.
A few things I've learned running these:
- Cap the session. Three to four hours max per phase pull. People fade, and a tired room makes soft commitments.
- Make the owner's decisions concrete. "Owner selects flooring" is useless. "Owner approves Level 3 flooring spec by March 12" is a commitment you can hold.
- Photograph the wall and get it into your look-ahead tool the same day. A pull plan that lives only on butcher paper is dead the moment the room clears out. The value comes from pulling those handoffs into a rolling look-ahead schedule the whole team refreshes weekly.
The Look-Ahead: Where Design Constraints Live
On a hard-bid job your six-week look-ahead is mostly physical: what needs to be built, by whom, in what order. On IPD you're carrying two kinds of constraints in the same view, and the second kind is what usually kills you.
The physical constraints are familiar — predecessor work complete, materials on site, area released, inspection passed. The information constraints are the ones IPD forces you to schedule: an approved submittal, a resolved RFI, a coordinated model, an owner decision. Treat them identically. If the RFI answer that unblocks the plumbing under-slab is due in two weeks, it goes in the look-ahead as a constraint with an owner and a need-by date, exactly like a concrete pour would.
The discipline that pays off is screening every activity for constraints before it enters the weekly work plan. Nothing goes into next week's committed work until every constraint on it is clear. In IPD that means an activity can be blocked by a designer or the owner, not just a sub — and because everyone's in the same accountability structure, you can actually say so in the room without starting a war. A shared look-ahead tool matters here mostly because it keeps one version of the truth. When the design team, the owner's rep, and six trades are all looking at the same constraint log, the "I didn't know you were waiting on me" excuse disappears.
Weekly Work Plans and Reliable Promises
The weekly work plan is the commitment layer. This is the meeting where each foreman looks at the coming week and says, out loud, "Yes, I can do that," or "No, I can't, because X." The rule that makes it work: only make ready, constraint-free work goes on the plan. If you let people commit to work that still has open constraints, your promises become fiction and your whole system loses its credibility inside a month.
What IPD adds is that the designer and owner make promises in this meeting too. The architect commits that the revised details will be issued Thursday. The owner's rep commits that the change directive will be signed by Wednesday. Those are tracked exactly like a trade commitment, and they show up in the score.
Measuring PPC When Everyone Is Accountable
Percent Plan Complete is the number that keeps the whole thing honest. You count the commitments made last week, count how many were completed exactly as promised, and divide. No partial credit — a task that's 90 percent done was not completed. A healthy team lands in the 80 to 90 percent range; if you're consistently above 95, your promises are probably too soft, and if you're stuck below 70, you've got a make-ready problem upstream.
Here's the part that unsettles people the first time on an IPD job: PPC applies to the owner and designer too. When the design team promises drawings Thursday and delivers Monday, that's a missed commitment and it goes in the number. This is uncomfortable, and it's supposed to be. The value is in the variance analysis, not the blame. Every missed commitment gets a reason code — waiting on prerequisite work, materials late, design incomplete, crew short, owner decision pending — and you chart the reasons, not the names. When you sort a month of misses by reason and the biggest bar is "design incomplete," you've just found where the project is actually bleeding, and because everybody's in the same risk pool, you can fix it instead of papering over it.
Software helps here in a genuinely boring, genuinely essential way: it does the counting and the reason-code rollups automatically. Tracking PPC and variance by hand across a dozen parties in a spreadsheet is how the metric quietly dies by week five. In LookAheadWall the commitments and their completion status come straight off the weekly work plan you already built, so the score is a byproduct of running the process rather than a separate chore nobody wants to own.
Design and Construction on the Same Timeline
The biggest waste I see on IPD jobs isn't in the field — it's in the seam between design and construction. Builder input during design is the whole selling point of IPD, and Last Planner is how you make it operational instead of aspirational. Pull the design deliverables into the same backward-passed sequence as the construction work, so the design team can see that the mechanical room layout has to be locked six weeks before the equipment ships, which has to happen eight weeks before you set it. Design gets a real deadline tied to a real physical consequence, not a soft "as soon as you can."
A rule of thumb worth stealing: give yourself a buffer between an information handoff and the field work that depends on it. A day or two isn't enough. When a submittal approval feeds a fabrication lead time, you want that approval landing well ahead of the "must-have" date, because it never comes back clean the first time. Build the float in on purpose or the field will absorb it as delay.
Big Room, Target Value Design, and the Culture Problem
IPD teams love the Big Room — everyone co-located, planning together, screens on the wall. It's genuinely powerful, and it's also where Last Planner sessions belong: the pull plans, the weekly work plan meetings, the PPC review all happen there with the look-ahead up on the screen so nobody's working from a stale printout. If your team is distributed, the substitute is a single shared plan everyone can see and edit in real time; the co-location matters less than the shared truth.
Target Value Design fits the same way. When the trades are in the room early, the foreman who's installed a thousand of something can tell you the sequence that saves three days and the detail that saves ten grand, before the design is frozen. Last Planner gives that input a place to land — it becomes a sequencing decision in the pull plan rather than a good idea nobody captured.
Now the honest part. The hardest thing about running Last Planner on IPD isn't the method, it's the culture collision. You've got an owner who's used to being handed a schedule, a design firm that's never had its dates scored, and trade partners who've spent careers protecting themselves from exactly this kind of transparency. Three things get you through it:
- Start at kickoff, not mid-project. Retrofitting Last Planner onto a job that's already defensive is twice the work.
- Score the process, not the people, out loud and consistently. The first time a missed owner commitment gets treated as a fact to learn from instead of a fight, the room relaxes.
- Put everyone on one tool. When each company brings its own system, the plan fragments and the shared visibility that IPD depends on evaporates. Pick one look-ahead platform, give every party a role-appropriate view, and make it the single source of truth.
The Bottom Line
IPD writes collaboration into the contract. It does not make collaboration happen — that's a daily practice, and Last Planner is the practice. Run real pull planning with the owner and designer at the wall, carry information constraints in your look-ahead alongside the physical ones, make reliable weekly promises across every party, and score PPC honestly for everyone including the folks who aren't holding tools. Do that from day one, keep it all on one shared plan the whole team can see, and IPD stops being an expensive contract structure and starts being a genuinely faster, calmer way to build.