2026-10-14 · 5 min read

How to Build a 3-Week Look-Ahead from a P6 Schedule

The master programme tells you where the project needs to be. The look-ahead tells the site team what they are doing on Monday morning. If those two documents drift apart, the programme becomes a wall chart and the look-ahead becomes a wish list.

In this article I will walk through the way I build a 3-week look-ahead from a Primavera P6 schedule. It is not complicated, but there are a few decisions that make the difference between a look-ahead people use and one that gets printed, pinned up and ignored.

What a 3-week look-ahead is for

A look-ahead is a short-window extract of the programme, broken down far enough that foremen and supervisors can plan labour, materials, access and permits against it. Three weeks is a common window because it is long enough to order materials and book plant, and short enough that the plan is still realistic.

It has three jobs:

1. Translate programme activities into work that a crew can start and finish.
2. Expose constraints early: missing information, materials, access, permits, preceding trades.
3. Feed progress back into the master schedule so the two stay aligned.

If your look-ahead is not doing the third job, it is a separate plan, and you now have two programmes to manage.

Step 1: Update the P6 schedule first

Never build a look-ahead from a stale programme. Before you extract anything, run your normal progress update: actual starts and finishes, remaining durations, percent complete, and then schedule (F9) with the data date set to the start of the update period.

Check that nothing obviously broken comes out of the reschedule. Activities with a start date before the data date and no actual start are a red flag. So are activities that suddenly jump weeks to the right. Fix those before you go further, or your look-ahead will carry the same errors.

Step 2: Filter the window

In P6, create a filter for the look-ahead window. A typical setup is:

• Activities where the Start is on or before the data date plus 21 days, and
• the Finish is on or after the data date, and
• the activity is not completed.

This catches activities that are already in progress, activities due to start in the window, and long-running activities that span it. Save the filter so you can reuse it every week. I usually also save a layout with the columns I need for export: Activity ID, Activity Name, Original Duration, Remaining Duration, Start, Finish, Total Float, and any activity codes for area, trade or subcontractor.

Step 3: Group by the way the site works

The programme is often grouped by WBS, which usually follows the contract or the design packages. Site teams think in areas and trades. Before you export, regroup the layout by area (or level, or zone) and then by trade or responsible party.

This is where activity codes earn their keep. If the schedule has consistent codes for area and contractor, the look-ahead almost builds itself. If it does not, adding them is one of the best investments you can make in a schedule, because it pays back every week.

Step 4: Export to Excel

Use the P6 export to Excel (File, Export, Spreadsheet) or copy straight from the layout. Either works. What matters is that you export the same columns every week so your template formulas keep working.

In the template I keep two tabs:

• Programme extract: the raw export, pasted in every week without editing.
• Look-ahead: the working sheet that references the extract and adds the site-level detail.

Keeping the raw extract untouched means you can always trace a line on the look-ahead back to a programme activity. That traceability is what keeps the two plans aligned.

Step 5: Break activities down where needed

Some programme activities are already small enough for a crew to plan against. Others, such as "Level 2 containment and cabling, Zone B, 25 days", are not. On the look-ahead, break those into the tasks the supervisor will actually allocate: by run, by room, by grid line or by system.

Two rules I stick to:

• Every look-ahead task must roll up to one programme activity. Keep the Activity ID in a column against each sub-task.
• Do not add new scope on the look-ahead. If something is missing from the programme, it goes back into P6, not just onto the spreadsheet.

Step 6: Add the constraint columns

This is the part that makes a look-ahead useful. For each task, add simple columns:

• Information (drawings, RFIs, approvals)
• Materials (on site, ordered, not ordered)
• Labour (crew assigned or not)
• Plant and access (scaffold, MEWPs, cranes, permits)
• Preceding work (handed over or not)

Mark each one as clear or not clear. A task is only "ready" when everything is clear. In the weekly meeting, the conversation then becomes about clearing constraints on next week's work instead of explaining why last week's work did not happen.

Step 7: Show it visually

A simple bar chart in Excel, with one column per working day and conditional formatting to fill the cells between start and finish, is enough. You do not need a Gantt chart tool for this. Colour the bars by status: ready, constrained, in progress, complete.

Keep the layout readable when printed on A3. A lot of look-aheads are still read on paper or on a tablet at the workface.

Step 8: Run the weekly loop

The look-ahead is a weekly cycle, not a one-off document:

1. Update P6 with actual progress.
2. Re-extract and refresh the template.
3. Review with supervisors and subcontractors: what is ready, what is constrained, who is clearing what.
4. Record what was planned versus what was done.
5. Feed changes back into the programme where logic or durations are wrong.

Step 4 is optional but worth doing. Tracking the percentage of planned tasks that were actually completed each week gives you an early, honest signal of whether the plan is reliable, long before slippage shows up on the critical path.

Common mistakes

Building the look-ahead from scratch every week. It takes too long and breaks traceability. Use a template that refreshes from the extract.

Letting the look-ahead and programme diverge. If the site is working to a different sequence, the programme needs to change, or the programme is fiction.

Too much detail. If a supervisor needs to scroll through 400 lines to find their crew's work, they will stop looking. Filter by area or trade and give each person their slice.

No constraint tracking. A look-ahead without constraints is just a shorter Gantt chart.

The template

I use a basic Excel layout on large technical builds: a raw extract tab, a working look-ahead tab with sub-task rows, constraint columns, a daily bar chart driven by conditional formatting, and a simple planned-versus-complete tracker. It works with an export from P6 or from MS Project, as long as you keep the column order the same.

That template is coming soon to newsletter subscribers. Join below and you will get it when it is ready, along with the weekly newsletter. If the layout does not suit your project, adapt it. The method matters more than the spreadsheet.