How to Build and Use a Tat Analysis Sheet
A Tat Analysis Sheet is a straightforward spreadsheet you use to track the time between when a request enters a process and when it exits. It sounds simple enough, but the real work is in how you define the start and stop points, which is where most people trip up. I built my first one around 2018 for a manufacturing line handling custom fabrication jobs. We had incoming orders bouncing between three departments — cutting, welding, finishing — and nobody actually knew where time was getting eaten. My manager wanted a single view of turnaround from order receipt to shipment. That's the Tat Analysis Sheet in its most basic form: timestamp in, timestamp out, gap calculated, bottleneck identified.
Tat Analysis Sheet structure
Set up five core columns. Column A is the Job ID or Request Reference. Column B is the intake timestamp — the exact moment the request is logged or received. Column C is the department or stage name, because a single job usually passes through multiple touchpoints. Column D is the completion timestamp for that specific stage. Column E is the elapsed time, which is just a simple subtraction formula between D and B, formatted as hours and minutes. The row-level layout matters. Each stage gets its own row for the same Job ID. So if a job goes through three departments, it appears three times, each with its own timestamps and calculated gap. This lets you see where a single order is sitting idle versus actively being worked on.
What actually goes into the cells
Don't guess at timestamps. Pull them from whatever system actually records them — an ERP, a ticketing tool, a shop floor log. If you're entering data manually, you're already introducing error. I've seen variance of up to forty minutes between what someone thought a job started and when it actually did, just because they were filling out a sheet at the end of the day from memory. For the elapsed time formula, use something like =TEXT(D2-B2,"[h]:mm") so negative values don't break your sheet. And make sure B and D are formatted as date-time cells, not text. I learned that the hard way when a sheet I inherited treated timestamps as strings and my subtraction formulas returned zeros across every row. Took me an hour to realize that before I rewrote the import logic.
Get the Full Details
Identifying the bottleneck
Once you have clean data, pivot or filter by the stage column and sort by elapsed time descending. The longest average duration per stage is your bottleneck. But here's the part beginners miss: the longest stage isn't always the constraint. Sometimes the constraint is the waiting time between stages, not the active processing time. I ran into this exact situation with a client last year. Their welding stage averaged 4.2 hours, which looked like the problem. But when I added a sixth column tracking queue time — the gap between one stage completing and the next stage starting — the real issue jumped out. Welding sat waiting an average of 6.8 hours before finishing even got to move the piece to the next station. The actual constraint was material handling between stations, not the welding itself. Fixing that reduced overall TAT by 38% without touching the welding process at all.
Common mistakes that ruin the sheet
One major error is mixing different time zones in the timestamps. If your intake team logs in EST and your warehouse team logs in PST, your elapsed times are wrong by three hours on every row crossing a regional boundary. Standardize on one time zone before you start collecting data. UTC if you're dealing with multiple locations, or pick the dominant one and convert everything else. Another mistake is including setup time inside the processing time without separating them. If a machine needs twenty minutes of prep before it runs a batch, and your sheet lumps that into the TAT for the batch, your numbers look worse than they actually are and you can't tell whether the delay is in the work or the preparation. Add a separate column for setup versus run time. It takes five extra minutes per row and it pays for itself immediately when you're trying to justify a new machine or staffing decision. Here's a less obvious one: don't include incomplete or still-in-process jobs in your average calculations unless you clearly mark them separately. A job that's been sitting at the finishing stage for eleven days skews your average and makes the process look broken when it's just an outlier. Filter those out or flag them, and calculate averages on completed jobs only. Then report the outlier count separately so management sees both the normal flow and the exceptions.
When a Tat Analysis Sheet won't help
If your process has fewer than ten distinct stages or your throughput is low enough that you can see every job move in real time, a full sheet is overkill. A whiteboard or a Kanban board does the same job faster. The Tat Analysis Sheet earns its keep when you have volume — fifty plus jobs per week across multiple stages — where human memory and casual observation can't track the patterns anymore. It also falls apart if your data sources aren't reliable. I worked with a logistics company where drivers were supposed to scan jobs in and out via a mobile app, but they'd scan batches of ten at the end of the shift instead of individually. The timestamps were close to right but systematically off, and the elapsed time column was basically decorative. No sheet in the world fixes bad input. You either fix the data collection process or you're just generating a fancy spreadsheet that tells you nothing useful.
Export and sharing
Once the data is collected, a few quick pivots turn the raw sheet into something actionable. Group by stage, average the elapsed time, and flag anything over a threshold you set. I usually recommend 1.5 times the running average as the trigger point — anything above that gets a separate row in a summary sheet with a notes column for investigation. Keep the raw data untouched in the main sheet. Clone it to a summary tab for reporting so you always have the original record. Version your exports with dates so you can compare month to month. Without that trail, you'll be looking at a single snapshot and wondering why numbers changed when you can't go back and check what actually happened. The template itself is free to build from scratch — no special software required. Excel or Google Sheets works fine. If you want a ready-made structure, search for "Tat Analysis Sheet template" on spreadsheet marketplaces, but don't buy anything expensive. The math is subtraction. The value is in how you set it up and what you do with the results.
