Ogden Division — On-Time Diagnostic: Root Causes

Job Scheduling & Milestone Info, Task Pay Info, Tasks General, Unexpected Orders, and Job & Task & Delay Photos — Ogden Division, joined job-by-job

Methodology

"Target End Date" comes from TrackVia's Job Scheduling & Milestone Info report. "On time" means Major Tasks Completion Date fell on or before that target — any overage, even one day, is scored "Late."
Vinyl-primary houses excluded. 12 jobs were removed (a logged Vinyl task with no stucco/masonry base coat) — all 12 were in Railrunner (DAI Construction). Houses using vinyl only as a trim/accent alongside real stucco or masonry work are kept in.
The construction sequence used below (per Ronnie): Scaffold set → Paper/Wire & Housewrap → Weather Barrier Inspection → Siding & Brown Coat → (Rock/Brick, if the house has it) → Paint → Soffit & Fascia → Stucco Color → Scaffold removed. Gutters, tarps, VPO work, warranty/return visits, and other follow-up tasks sit outside this core sequence and are excluded from the stage-gap analysis in Section 3 so they don't distort it.

1. How close are we, really?

Sustained volume (10+ scored jobs/month) started 2026-01. Everything below uses 2026-01 through the most recent complete data (262 jobs).
50% 26-0132%26-0251%26-0352%26-0439%26-0553%26-0653%26-0736%
125
On time (48%)
137
Late (52%)
10.5
Avg days late (late jobs)
7
Median days late

How late, when late?

Days over targetJobs% of late jobs
1-2 days1712.4%
3-5 days3324.1%
6-10 days3827.7%
11-20 days3223.4%
21+ days1712.4%

2. Do documented delays explain this?

TrackVia's Job & Task & Delay Photos report lets crews log a Delay, Material Delay, or Scaff Delay against a job with a date range. It only has live/recent data (no full history), so this comparison is limited to the 159 jobs (61% of the sustained set) that finished after logging began (Apr 07, 2026).
5%
Late jobs with a logged delay
20%
On-time jobs with a logged delay
No — barely any of it is documented. Only 5% of late jobs have a formal delay record at all — and, if anything, on-time jobs log delays more often (20%). That's not because delays don't happen on late jobs — it's more likely that logging a delay tends to get it resourced and resolved before it costs the schedule, while the delays that actually cause lateness are happening off the books. The delay-tracking feature isn't capturing the real cause of lateness right now.

3. Where the time actually goes: the build sequence, stage by stage

Using the sequence above, each job gets a timeline of when it hit each stage (the last task in that stage to complete). "Idle time" is the sum of the gaps between consecutive stages — days where the job didn't move forward. 30% of raw task-duration records come back as exactly 0 hours (crews logging start = finish), so this stage-gap approach is used instead of task duration — it only needs completion timestamps, which are logged more reliably.
26 days
Avg total idle time — Late jobs
18 days
Avg total idle time — On-time jobs
Late jobs sit idle about 8 more days across the build than on-time jobs — this is still the clearest, most consistent signal in the data.

Where the gaps are biggest, stage to stage

Stage handoffAvg gap — LateAvg gap — On timeDifference
Rock/Brick → Soffit & Fascia15.0d (n=5)-0.0d (n=9)+15.0d
Siding & Brown Coat → Rock/Brick7.5d (n=82)3.1d (n=64)+4.3d
Siding & Brown Coat → Paint3.8d (n=48)0.9d (n=28)+2.9d
Housewrap/Paper-Wire → Siding & Brown Coat14.0d (n=136)11.4d (n=124)+2.6d
Siding & Brown Coat → Soffit & Fascia1.4d (n=6)-0.7d (n=30)+2.1d
Stucco Color → Scaffold Removed4.1d (n=99)3.0d (n=73)+1.0d
Paint → Soffit & Fascia2.8d (n=122)2.1d (n=80)+0.7d
Scaffold Set → Housewrap/Paper-Wire4.4d (n=112)4.0d (n=79)+0.4d
Soffit & Fascia → Stucco Color-1.7d (n=99)-0.2d (n=69)-1.5d
Rock/Brick → Paint-5.0d (n=77)-1.1d (n=55)-3.9d
The two handoffs worth acting on first — biggest gap and enough jobs behind them to trust the number: Housewrap/Paper-Wire → Siding & Brown Coat (the longest wait in the whole sequence — roughly 14 days on late jobs vs. 11 on time) and Siding & Brown Coat → Rock/Brick (about 7.5 days on late jobs vs. 3 on time — more than double). Both have 60+ jobs behind them on the late side alone. A few stages also show up completing out of the expected order (negative numbers) — worth a second look at whether the process is running the way it's meant to, or whether crews are circling back for touch-ups.

4. Which POs go with late jobs

Comparing the share of jobs that placed at least one unexpected/will-call order of each type. Only types with a meaningful sample shown below.
6.0
Avg # unexpected POs — Late
4.6
Avg # unexpected POs — On time
$1,549
Avg unexpected PO $ — Late
$987
Avg unexpected PO $ — On time
PO type% of late jobs with one% of on-time jobs with oneGap
Paint Order23%6%+17pp
Soffit Order31%18%+13pp
Initial Order30%20%+10pp
Color Order28%25%+4pp
Consumables Order88%84%+4pp
Gutter Order7%6%+0pp
Vinyl Order7%7%+0pp
Paint and Soffit will-calls stand out the most — Siding and Rock orders, despite being the biggest task categories overall, are actually not disproportionately tied to lateness. I checked whether these orders get placed during the stage-gap windows above (i.e. crews stalled waiting on the reorder) and the timing mostly didn't line up — most Paint will-calls land either well before or after the gap, not inside it. So this reads as "late jobs run into more material problems overall," not "this specific PO caused this specific wait" — still useful, just not a one-to-one story.

5. Other factors checked

FactorLate jobsOn-time jobsVerdict
Started later than originally planned40%39%Not a driver — same rate either way
Jobs with a task at 2x+ typical duration83%82%Inconclusive — duration data too noisy
Bottom line: idle time between construction stages — especially the wait for the base coat/siding crew after housewrap, and for the rock/brick crew after that — is the clearest driver. Late jobs also carry more material problems (especially paint and soffit reorders), but that's a correlation, not a clean cause-and-effect on timing. Documented delays barely touch the late population, meaning most of this is happening without anyone flagging it in the system. Starting late doesn't predict finishing late.