
I Compared WorkBuddy Against Manual Work on Concrete Overuse—Here's Where They Differ
At noon I was scrolling on my phone in the site office and saw a post saying that on a railway section in Hubei, using Glodon's system to reconcile planned concrete usage against actual consumption, the over-consumption rate dropped from 110% to 45% in half a year, creating over three million in savings. After reading it I put my phone face down on the table and looked back at the ledger for my own factory building project.
Honestly, over-consumption is something I know all too well. Planned usage is in the budget spreadsheet, actual poured volume is in the ready-mix plant's reconciliation statement, and in between there are the mixer truck drivers' signed receipts, the supervisor's pouring records, and change orders. These things are scattered across four or five places, and by the time month-end comes and they don't match, the concrete has long since set in the walls.
My project isn't big — thirty or forty truckloads a month — but it's messy enough. Before, I tried reconciling purely by hand, Excel open, checking the license plate number and time on each receipt truck by truck; my eyes glazed over and I'd still mix up rows. Later I ran WorkBuddy against the manual approach. It won't balance the books for me, but it can take over the file-flipping part, which saved me a lot of trouble. Below I'll explain how I set it up and where it differs from pure manual work.
First my environment: Win 11, WorkBuddy used for three months, desktop version installed — may not apply to everyone, especially if you're on a company intranet with encryption, in which case you may need IT to whitelist it.
The first difference is in the feeding stage. For manual reconciliation, I open the folder and click open each PDF and photo one by one. With WorkBuddy, I first create a project-specific folder, named something like Concrete-November, and throw in the reconciliation PDFs sent by the ready-mix plant, scans of the mixer truck receipts, and photos I took myself. Note: don't mix in files from other months — I stepped in this pit; last month I scanned in a few October slips and the summary came out with two extra truckloads. After dumping them in, open WorkBuddy, click File Organization in the left sidebar, and it scans through first; a file list appears in the interface with each file tagged with its recognized type, like reconciliation statement, receipt, image. Don't rush to have it work at this step — you need to eyeball the list yourself to see if any receipt got misidentified as a reconciliation statement. On my end about one in ten gets misidentified; just manually change the type.
The second difference is in how you give instructions. I took a detour here. At first, to save effort, I just told it "summarize the concrete usage in this folder," and the output was unusable because it didn't know which file to treat as authoritative. Later I learned my lesson and split the task up. First have it do only one thing: open Data Analysis, select those reconciliation statement files, and tell it "extract the date, license plate, and volume for each truckload of concrete, output as a table." The expected result is a table with one row per truck. At this point your key check is whether any volume has a misplaced decimal — I've seen 8.5 cubic meters recognized as 85, which you can spot at a glance.
The third difference is in the comparison afterward. Once both the receipt table and the reconciliation table are out, that's where the real time savings are. Still in Data Analysis, I tell it "using license plate and date as the key, compare the volume differences between the two tables and flag the inconsistent rows." At this step it gives you a half-finished product — it won't tell you who's right or wrong, just picks out the differing rows. The rest of the judgment is still on me; for example, two truckloads differed because the driver wrote the wrong truck number when signing, which a machine can't know.
My usage is actually simple: WorkBuddy turns scattered paper into comparable tables, and for the mismatches I still have to call the ready-mix plant myself. It can't replace my judgment, but it saves those hours of flipping files.
A few operational habits I'll mention in passing. First, after each job, don't let it auto-archive — I manually move the original files into a subfolder called Originals-Don't-Touch, so the next run won't double-count. Second, for the tables it produces, I habitually export an Excel copy with the date in the filename, like NovConcreteComparison-1115, because if you later modify the files, the results inside WorkBuddy may no longer match. Third, on permissions: I use it alone, no collaboration set up, but if several people in the crew look at it together, I'd suggest setting that project folder as read-only shared, with write permission for only one person — otherwise if several people dump files in at the same time, the scanned list gets messy.
Back to that news item at the top. That's a big section — with a system, dedicated staff, and processes; getting the over-consumption rate down to 45% is the whole apparatus working. A small project like mine doesn't have those conditions, but the books still have to be balanced. With my flow, monthly reconciliation went from an afternoon down to a bit over an hour, and the remaining time goes to checking the site, which beats staring at paper.
One honest word for first-time users: don't try to have it manage your entire cost ledger right away. Start by practicing on one month of concrete reconciliation — get the four steps of creating the folder, scanning the list, splitting the instructions, and comparing differences running smoothly, then apply it to other categories. Tools are dead; how you assign the work is what's alive.
Physix Frontier