Survey Findings Sorted and Counted for Pricing

Build Concierge reads the report, matches every finding against your own categories, counts how many fall into each, and returns the totals.

Your categories, your matching rules, and your decision about which category wins when a finding could belong to two.

Because reports are written by people, and people describe the same defect five different ways.

An inspection, a condition survey or a compliance report comes back with dozens of findings, each written in a surveyor's or an engineer's own words. To price the remedial works you need to know how many of each type there are. So somebody reads the report line by line with a tally sheet, deciding which category each finding belongs in.

That takes hours.

It's also inconsistent, because the next report might be read by someone who draws the lines differently.

And it's error-prone in a specific way: a finding described in unfamiliar wording gets missed, not through carelessness but because it didn't look like the thing they were counting.

The obvious cost is time, and delay. A report that takes a day to tally is a day before the client gets a price.

The expensive cost is the miss. A remedial that was in the report but never got counted isn't in the price, and it gets found on site by an engineer who then can't do it. That becomes a variation, a return visit, or absorbed work. All three cost more than the tally would have.

The quiet cost is inconsistency. If two people count the same report differently, you can't compare one job to another, and your own pricing benchmarks are built on sand.

1. Reads the report, picking up the findings and ignoring the parts that aren't findings.
2. Matches each finding against your categories, using the phrase rules you've defined rather than a loose description of what each category means.
3. Counts each distinct finding separately, including where several appear together.
4. Applies your precedence rules, so a finding that could fall into two categories is counted once, in the one you've said takes priority.
5. Returns a count per category, ready for the next step.

It matches on the words that are in the report. It doesn't form a view about the condition of anything.

Worth deciding early, because this is what turns counts into a priced schedule.

• Emailed to whoever prices the work
• Recorded as a note on the job
• Used as quantities against priced items, building the schedule directly
• Used to raise a job per category of remedial

Most people start with the first and move to the third once they trust the counts.

• Your categories. This is the entire build, so it's the entire conversation.
• The matching rules for each one: which words or phrases put a finding into it.
• Your precedence rules for findings that could fall into two categories.
• A decision on where the counts need to end up.
• Reports arriving somewhere they can be read.

We run the category workshop.

Expect this to take real time, and expect it to surface categories your team has been applying without ever writing down.

We define the matching by the words themselves, not by what a category means.

A category called "damage" produces inconsistent counts, because two people reading the same report will draw the line in different places. A category defined by the phrases that actually indicate it returns the same answer every time.

We test it against reports you've already tallied.

Twenty historic reports, counted by hand, compared against the automated counts. Every disagreement is either a rule to fix or a category nobody had defined.