What Process Mining Actually Is
The discipline of learning how work really flows, from the trail the work leaves behind.
Process mining is a mature discipline in other industries and nearly unknown in restoration. It reconstructs how work actually moves through a business from the digital footprints it leaves, and shows where it waits, loops, and diverges from how everyone assumes it goes. Restoration is finally in a position to use it.
There is a gap in almost every business between how work is supposed to flow and how it actually flows. The process on the whiteboard is clean: one step leads tidily to the next. The process in reality has detours, waits, and rework that nobody designed and that few people can see, because no single person ever watches a job travel the entire way through. Process mining is the discipline of closing that gap, by reconstructing the real process from the evidence the work leaves behind rather than from anyone's memory of how it goes.
It is well established in manufacturing, logistics, and finance, where it has reshaped how those industries find and remove waste. In restoration it is nearly unknown, and that is worth pausing on, because it is not unknown for any good reason. It is not that the discipline does not apply to restoration work. It is that the data needed to do it was never assembled. That is the part that has recently changed.
Work Leaves A Trail
The foundation of process mining is a simple observation: every step in a business leaves a digital trace, and almost every trace carries a timestamp. A job is created, and that moment is recorded. An estimate is written, scheduled, started, invoiced, paid, and each of those events is logged somewhere with a time attached. Individually these timestamps are just administrative exhaust, the ordinary residue of running the business. Strung together in order, they become something far more valuable: a precise record of the actual path a job took, from beginning to end, with the time it spent at each stage.
Process mining reads that trail across thousands of jobs and reconstructs the real flow. Not the idealized version anyone would describe if you asked them, but the one that genuinely happened. The whiteboard says estimate, then schedule, then start. The data might say estimate, then a three-day wait, then schedule, then a callback that sent the job backward, then start. The wait and the callback were always part of the process. They were simply never visible, because no one was ever in a position to see the whole sequence laid out at once.
Patterns, Not Anecdotes
The power of doing this at scale is that individual stories give way to patterns, and patterns are something you can actually act on. One job that stalled is an anecdote, easy to explain away as a bad week or a difficult customer. A thousand jobs that consistently stall at the same handoff is a finding, and a finding points somewhere specific. Process mining turns the vague question "why did this job take so long" into the answerable one "where does our work reliably lose time," which is a question that has a real address and a real fix.
In practice it tends to surface three things, and each one is a place where time and margin quietly leak. It shows where work waits, the gaps between steps where a job sits doing nothing while the clock runs. It shows where work loops back on itself through rework, the callbacks and redos that the clean version of the process never mentions. And it shows where the actual path diverges from the intended one, the cases where the real sequence simply is not the sequence anyone believes is happening. None of these are visible from inside a single job. All of them become obvious once the full trail across many jobs is in view.
Why It Works From Records, Not Opinions
What makes process mining distinctive, and worth the effort, is that it does not depend on anyone's recollection of how the process goes. It reads what the records show. This matters not because people are unreliable in some embarrassing sense, but because of an ordinary limit on everyone: no one can hold thousands of jobs in their head at once, and the day-to-day view from inside the work naturally focuses on the job right in front of you. Each person sees their own stretch of the road clearly and the rest of it dimly, and the parts that are dim are exactly the parts where problems hide.
The timestamp trail does not have that limit. It keeps the whole history, every job, every step, including the wait that slipped everyone's mind and the rework that never quite came up at the meeting. So the proposition is not to second-guess what experienced operators know. It is to complement that hard-won knowledge with a clear, complete read of how the work actually ran, drawn from evidence the business already produced as a byproduct of doing the work. Experience tells you what usually happens. The trail tells you what did happen, every time, and the two together are stronger than either alone.
Key Finding
Process mining reconstructs how work really flows from the timestamps it already leaves behind. The gap between that and the version on the whiteboard is where a great deal of the time, and the money, quietly goes.
A Way To Start Seeing Your Own Flow
You do not need software to begin thinking like this, and starting to notice is most of the value. Pick a single recent job and walk it backward in your mind, not as it should have gone, but as it actually went. When was it created? How long did it sit before someone picked it up? Where did it wait for an answer, an authorization, a signature, a part? Did it ever go backward, repeating a step that should have happened only once?
Almost every operator, doing this honestly for the first time on a real job, finds at least one wait or one loop they would not have predicted, a place where days went somewhere and no one quite noticed. That single observation is process mining in miniature, done by hand on one job. The discipline simply does it across thousands of jobs at once, so the one-off observation becomes a reliable pattern you can trust and act on. The question worth carrying is this: if you could see the true path of every job your business ran last year laid side by side, where do you suspect the waiting actually happens? You likely have an instinct. The whole point of reading the trail is to find out whether your instinct is right, and to learn the parts you could not have guessed.