Every scheduling tool on the market will tell you it "supports the Last Planner System." Most of them are lying, or at least stretching. What they usually mean is that you can type activities into a grid and mark them done. That is not Last Planner. That is a to-do list with a sales deck.
I've watched crews try to run pull planning and weekly work plans on tools that were never built for it — general PM platforms, a shared spreadsheet, a Gantt chart nobody trusts. It mostly works until the first constraint gets missed, and then everyone quietly goes back to the trailer whiteboard. If you're going to put real software in front of your foremen and subs, it has to earn its place. Here's what actually matters when you're judging whether a tool can carry the Last Planner workflow, from someone who has run the meetings.
The four things Last Planner actually needs the software to do
Strip away the feature lists and the Last Planner System is four connected moves: pull the sequence, look ahead and screen the work, make firm weekly commitments, then measure whether you kept them. Software that supports it has to hold all four together without you re-entering the same information four times. If a tool nails one and fumbles the rest, you've bought a very expensive silo.
A lookahead you can actually roll
The lookahead window — three, four, six weeks, whatever fits your job — is where make-ready happens. It's the room where you catch the missing submittal, the crane conflict, the pour that can't happen until the inspector signs off. The software's real job here isn't storing activities; it's making the weekly rollover cheap. When Monday comes and the window slides forward a week, you should not be rebuilding the board. New work should surface, completed work should drop off, and the constraints you flagged three weeks ago should still be attached to their activities.
Ask this in a demo: "Show me a lookahead in week one, then roll it to week two." If that takes more than a couple of clicks, your team will stop rolling it, and a lookahead you don't roll is just a stale printout.
Constraint tracking that names a person and a date
This is the part most tools get embarrassingly wrong. A constraint is not a checkbox — it's a promise someone owes the crew before work can start. RFI answer, material on site, prior trade complete, area released, permit in hand. The software has to let you attach a constraint to a specific activity, assign it to a specific person, give it a need-by date, and then nag when that date slips.
The test is whether the tool can answer, in five seconds, "what's going to stop us next week and who owns it?" If your constraint log lives in a separate spreadsheet from your schedule — which is how most jobs limp along — that answer takes an hour of cross-referencing, and by then the drywall crew is already standing around because the overhead MEP never got inspected. Good software keeps the constraint welded to the activity so the make-ready conversation writes itself.
Weekly commitments that only accept ready work
The whole discipline of Last Planner is that you don't commit to work that isn't ready. The last planners — your foremen — should only be putting an activity into this week's plan once its constraints are cleared. The software should make that hard to skip. When a foreman drags an activity into the weekly work plan, the tool should show whether its constraints are green. Committing constrained work should feel like an override, not the default path.
This is also where trade coordination lives or dies. Your weekly work plan is only as good as its handoffs. If the framers finish an area Thursday, the tool should make it obvious that the electricians are counting on Friday, and that a one-day slip cascades. Location-based visual plans — where you can see the actual areas and the sequence of trades moving through them — beat a flat activity list badly here, because that's how the work really moves: crews flowing through space, not rows on a chart. This is exactly the gap a purpose-built look-ahead board like LookAheadWall is trying to close, and it's a fair thing to hold any tool to.
PPC that computes itself and demands a reason
Percent Plan Complete is the heartbeat. At end of week you mark each commitment complete or not, and the software calculates PPC — commitments met divided by commitments made. That number is meaningless without the second half: when something didn't complete, the tool has to make you say why, and it has to make you pick from consistent variance categories (prerequisite work, materials, labor, design/RFI, weather, changed priorities, and so on).
The reason categories are the entire point. One week's PPC tells you nothing; twelve weeks of variance reasons tell you that 40% of your misses trace to late RFI answers, and now you have an argument to take to the design team with data behind it. If the software captures the completion but lets people skip the reason, you've automated the least useful part and thrown away the analytics that actually change behavior.
The field is the real customer
Here is the hard truth about construction software: it lives or dies in the field, not in the office. A tool that's beautiful on a project manager's laptop and miserable on a foreman's phone in the sun with gloves on will not get used, and Last Planner without field participation is just the office talking to itself.
So the mobile experience can't be a read-only afterthought. Your foremen need to update commitments, close out constraints, and mark completions from where the work is. That means big touch targets, fast entry, and a layout that survives a cracked screen and bright daylight. It also means the tool has to tolerate a bad connection — the far corner of a below-grade parking structure has no signal, and if the app throws away a foreman's updates because the bars dropped, he's done with it. A companion crew-leader app that shows people exactly what they're on the hook for this week, without making them log into the full platform, is often what finally gets subs to engage.
When you evaluate the field side, put the actual phone in the actual foreman's hand during the demo. Watch his thumbs. If he's pinching and zooming and squinting, no amount of desktop polish will save you.
Subs have to be able to walk in the door
Last Planner is a team sport, and the team includes people who don't work for you. If your subcontractors can't easily see and update the plan, you're back to running the whole thing yourself and calling it collaboration. Watch for two failure modes here. First, tools that charge per seat so aggressively that nobody wants to license their subs — then the subs never touch it and the data rots. Second, tools with onboarding so heavy that a two-man specialty sub gives up before their first weekly meeting.
What you want is dead-simple, role-appropriate access: a sub can see their commitments and the constraints they own, update their own work, and not accidentally blow up someone else's schedule. Role-based permissions and a low-friction invite matter far more here than any advanced feature.
Where integration helps and where it's a distraction
Vendors love to sell integration. Some of it earns its keep; a lot of it is checkbox theater. The connection worth caring about is to your master schedule — the CPM plan that sets the milestones the lookahead has to feed. Being able to tie weekly commitments back to the master activities keeps the short-interval plan honest and shows you early when the field is drifting off the contract dates.
Linking to drawings and documents is genuinely useful too, because half of constraint resolution is "which detail governs this?" and having the sheet one tap from the activity saves a trip to the trailer. BIM and cost integrations are nice-to-haves; don't let a slick 3D demo distract you from whether the daily workflow is any good. A tool that integrates with everything but is painful to run a Tuesday planning session on has optimized for the sales cycle, not the job.
The boring requirements that quietly sink projects
A few things never make the highlight reel but will wreck you if they're weak. Speed under load is the big one — a planning meeting means fifteen people hitting the same schedule at once, and a tool that crawls when the whole team is on it will lose the room in the first session. Test it with a real project's worth of activities, not the tidy little demo file.
Then the basics: reliable uptime, real data backup, sensible access control so the right people can edit the right things, and an audit trail so you can see who changed a commitment and when. Cloud-hosted is the norm now and usually the right call — anywhere access, automatic updates, no server to babysit — but confirm the vendor's backup story rather than assuming it. And look at the company behind the tool. Scheduling data is the memory of your job; you do not want to be migrating it off a platform whose vendor folded mid-project.
How to actually run the evaluation
Don't buy off a feature matrix. Run a real week of your real job through a trial and watch what happens:
- Build a genuine lookahead from a live project, constraints and all, then roll it forward a week. Time it.
- Have a foreman — not you — enter and update commitments from a phone on site.
- Invite one sub and see how long it takes them to do something useful without a training call.
- Deliberately fail to complete a couple of commitments, then see how cleanly the tool captures the variance reasons and rolls them into a PPC trend.
- Load it up and have several people hit it at once, the way a Tuesday planning meeting will.
The common pitfalls are predictable. Feature-bloated platforms nobody fully learns. Desktop-first tools that treat the field as second class. Generic project management suites bolted with a "lean" label that don't respect how make-ready and PPC actually work. And integration gaps that leave your constraint log stranded in yet another spreadsheet. If a tool survives a real trial week without triggering any of those, you've found something worth rolling out.
The bottom line
Software doesn't run Last Planner — your foremen and your discipline do. What good software does is remove the friction that makes people quit the method: the painful rollover, the disconnected constraint log, the PPC math nobody wants to do by hand, the field app that's a chore. Judge any tool by whether it makes the honest weekly rhythm easier to keep, not by how many logos are on its integration page.
Get the fundamentals right — rollable lookaheads, constraints tied to owners and dates, ready-work commitments, self-computing PPC with real variance reasons, and a field experience your crews will actually tolerate — and the technology fades into the background where it belongs. That's the goal. When the tool disappears and all you're left with is a crew that reliably does what it said it would this week, the software did its job.