What is a job log in manufacturing quoting?

A running record, one row per job, of what was quoted, what a job actually cost and took to complete, and the outcome. It replaces quoting from memory with quoting from the shop's own recorded history.

Do I need software to keep a job log?

No. A basic spreadsheet or even a physical notebook works, as long as it's filled in consistently. The value comes from the habit of logging every job accurately, not from the sophistication of the tool.

How is this different from a CPQ or quoting software platform?

CPQ tools automate and speed up the quote-generation process itself, usually assuming a company with the budget and volume to justify the software. A job log doesn't generate anything automatically. It just gives the person quoting a real historical reference to work from, which is a different and much lower-cost problem to solve.

How many columns should a job log have?

Five or six is usually enough. That means part description, quantity, quoted price and lead time, actual cost and lead time, and a one-line outcome note. More columns tend to reduce how consistently it actually gets filled in.

Who should be responsible for keeping the job log updated?

One named person, usually whoever handles quoting day to day. Shared responsibility across a whole team tends to mean nobody consistently fills it in.

Does a job log help after an RFQ has already been won, not just at the quoting stage?

Yes. The actual-cost and outcome columns, filled in after a job closes, are what make the next similar quote more accurate. The log's value compounds over time as more real outcomes get recorded, not just at the moment of quoting.

What's the biggest mistake shops make when starting one?

Trying to log everything after the fact, in a batch, from memory. That reintroduces the same accuracy problem a job log exists to fix. Log at the time of quoting and at the time a job closes, not months later.

Job Costing Log Template for Machine Shops

Most job shops quote from memory. Someone who's been there a few years glances at a new RFQ, thinks back to something similar and estimates from a feel for what that kind of job usually runs. It works, until the person holding that memory is out sick the week a big RFQ lands or leaves the company or simply misremembers a detail that mattered. The fix isn't quoting software, a CPQ platform or a knowledge-management tool. It's a running log of every job quoted and what actually happened, one row at a time, kept in a spreadsheet or a notebook, that turns "I think we charged around this last time" into "here's exactly what we charged and here's what it actually cost us."

Why "Ask the Person Who Quoted It Last Time" Stops Working

The default institutional memory in a small shop lives in one or two people's heads. It works well right up until it doesn't. That person is out the week a similar RFQ arrives, the job in question was quoted two years and a hundred other jobs ago and the details have blurred or the person who knew the shop's real cost structure on that part family has since left. None of this is a failure of anyone's memory. It's a structural problem with keeping institutional knowledge in a place that can walk out the door or simply be unavailable on a Tuesday.

A shop that has never written this information down doesn't know it's missing anything, because the memory-based system works well enough, most of the time, for as long as the same one or two people stay in the room.

What the One-Page Job Log Actually Is

Not a CPQ tool. Not a knowledge-management platform. Not an ERP module. A single spreadsheet, a running quote history built one row per job, with a small number of columns that take under two minutes to fill in when a job is quoted and again when it closes.

Column	What goes in it
Part / job description	Enough detail to recognize it again: material, process, key dimension or feature
Quantity quoted	The volume the price was based on
Quoted price and lead time	Exactly what went out to the customer
Actual cost and actual lead time	What it really took, filled in after the job closes
Outcome	Won, lost or won-but-thin-margin, plus one line on why

That's the whole template. No forecasting fields, no automated cost rollups, nothing that requires training or buy-in from a software vendor. The value comes entirely from filling it in consistently, not from the sophistication of the tool holding it.

What Quoting Without It Actually Costs

A quote built from memory tends to drift in one of two directions over time and both are expensive. It drifts optimistic, because memory tends to round toward "that job went fine" and forget the overtime hours or the scrap rate that ate into the real margin. Or it drifts conservative, because after one bad surprise, the same person pads every future quote by feel rather than by data and starts losing winnable work to competitors quoting from something closer to their real numbers.

A logged job removes the guessing on both ends. The actual cost column is either close to the quote or it isn't and if it isn't, the outcome line says why, a material price that moved, a tolerance that took longer to hold than expected, a customer change mid-run. That reason becomes available to whoever quotes the next similar job, whether or not the person who ran this one is still around to be asked.

How the Log Actually Speeds Up Quoting

The first several months of logging feel like pure overhead, filling in a row for no immediate benefit. The payoff shows up once the log has enough entries to search. A new RFQ arrives for a part that resembles something quoted eight months ago. Instead of re-deriving a price from raw material cost and a guess at cycle time, the person quoting pulls up the closest match, sees what it actually cost to run and adjusts from a real number instead of starting from zero. That's the difference between an estimate and a lookup and it's the single biggest lever the log has on response speed, faster than any software feature, because it's working from the shop's own real history instead of a generic cost model.

The Second Way to Use the Same Log

Looking up one similar job at quote time is not the only use for the log and over time it is not even the more valuable one. Once the log has a year or more of rows, scan it as a whole every quarter instead of one row at a time. Sort by part family or by won versus lost and look for a pattern a single lookup would never surface: quotes on one material running thin for three quarters straight, every job in one part family losing on lead time or a shop quietly pricing itself out of a category it used to win. A single row tells you about one job. A quarterly scan tells you whether the whole shop's pricing has drifted, in which direction and on what. It costs nothing beyond the time to look, since it is the same spreadsheet read a different way, not a new tool or a new column.

The only upkeep the log itself needs is staying searchable. After a couple of years of rows, archive anything old enough to be irrelevant into a second tab or filter by part family before searching, so a quick lookup at quote time stays quick instead of turning into its own kind of clutter.

What It Preserves When Someone Leaves

The quieter benefit shows up on the day someone who held a lot of that pricing knowledge retires, takes another job or is simply out for an extended stretch. A shop with two years of logged jobs hands the next person quoting a real reference instead of a blank page and a phone number to call the person who left. This is the part that never shows up as a cost on any single quote and shows up as a very real cost the one time it's missing.

The Ready-Made Version of This Log

Building the columns above from scratch takes a few minutes, but a ready-made copy removes even that friction. The free download (Augmino-One-Page-Job-Log-Template.xlsx) is the same six columns in this guide, already set up in a spreadsheet, with a cost variance column that calculates itself from the quoted price and actual cost so a quote that ran hot or thin is visible at a glance. It also runs the quarterly scan described above automatically: a small summary at the top of the same sheet keeps a live count of jobs logged, win rate, average cost variance and how many jobs ran more than 15% over or under quote, updating itself as rows get filled in instead of waiting for someone to read the whole log. Nothing beyond that. It is still just a spreadsheet.

Common Mistakes When Starting a Job Log

Too many columns. A log with fifteen fields nobody has time to fill in accurately becomes a log nobody fills in at all. Five or six columns, kept simple, actually gets used.
Filling it in after the fact, in a batch, from memory. The log's entire value comes from accuracy. A row filled in three months later from a vague recollection just reintroduces the same memory problem the log was built to fix.
No single owner. If filling in the log is "everyone's job," it's nobody's job. One person, usually whoever does the quoting, owns keeping it current.
Starting it only for big jobs. Small, repeat-order jobs are exactly the ones worth having real data on, since they're the ones most likely to get quoted again from memory.
Never actually searching it before quoting something new. A log that exists but never gets consulted before a new quote goes out is providing none of its real value. The habit of checking it has to be built alongside the habit of filling it in.

The Takeaway

A shop that quotes from memory is quoting from whatever one or two people happen to remember on a given day, which works until it doesn't. A one-page job log, kept consistently, replaces that memory with a real, searchable record of what jobs actually cost and how they actually went. It costs nothing to start, takes minutes per job to maintain and pays off the moment a new RFQ resembles an old one or the moment the person who used to hold all that knowledge in their head is no longer the only place it lives.

The One-Page Job Log: How Small Shops Quote Faster Accurately