Production tracking: what it is and how to set it up in a week

Production tracking is the habit of recording, as it happens, which items have been made, by whom and when. That is all. It is not a planning system, a costing model or a dashboard. It is a trustworthy answer to one daily question: what is finished, and what is not?

Without it, the answer lives in people’s heads. The supervisor knows, the owner asks the supervisor, and the sales team asks the owner. Every question costs someone a walk to the floor, and every answer is slightly out of date by the time it is given.

With it, anyone can look at a list and see that 42 of today’s 60 items are done, that three orders are part made and that one order has been waiting since Monday. You find problems while there is still time to fix them.

The biggest mistake in a first attempt is trying to record everything. Operators then spend their day typing and the data is thrown away. Start with the smallest set of events that answers real questions.

Notice that “made” and “loaded” sit at the heart of it. If you only ever capture those two events, you already know what is finished, what is waiting and what has gone. Extra stages such as cutting, framing and upholstery can be added when the basics are solid.

Track pieces, not just orders. A single order often contains several items, and several parts per item. An order is only ready when every part is ready.

A scan point is a fixed place where a worker scans a label to say “this has happened”. Good scan points are few, obvious and unavoidable.

Rules for picking them

• Put the scan where the item naturally pauses, such as the end of a bench or the door of a bay.
• One scan point means one meaning. Do not make people choose from a menu at a busy bench.
• Make the scan the last step of the job, so it is only ever done once the work really is finished.
• Keep it within arm’s reach of the worker, with a scanner on a stand or a short cable.
• Never rely on memory. If the scan can be skipped, it will be skipped on the busiest day.

A sensible first set

1. Made: at the end of final assembly or finishing.
2. Put away: where the item goes into a rack or bay, if you use them.
3. Loading: at the van or the dock door.

You can capture each event with a handheld scanner, a USB scanner that types like a keyboard, or the camera on a phone. Pick the cheapest option that survives your environment. A dusty upholstery shop and a clean assembly room need different hardware.

You do not need a long project. A small factory can have a working system in five working days if it keeps the scope narrow. Here is a plan you can adapt.

Day 1: Map one order from start to finish

Take a real order and follow it through the building with a pen. Write down every place it stops, who touches it and what changes. Circle the two or three moments that matter most. Those are your scan points. Agree the list of products and how many pieces each one ships as.

Day 2: Decide your labels and codes

Decide what is on a label. At a minimum: a unique barcode, the product name, the size and colour, the order reference and who it is for. Labels should be one per piece, so an item that ships in three parts gets three labels. Order a roll of labels and confirm that your printer works.

Day 3: Choose and test hardware

Pick scanners for each scan point. Test them on real labels, in the real light, with real gloves and dust. Check that a scan registers on the first try. If a scanner needs two goes, people will stop using it.

Day 4: Load orders and print labels

Get tomorrow’s orders into whatever you are using, whether that is a sheet or a system. Print the labels and attach them to the first batch. Put the production list on the wall or on a tablet so the team can see what today’s target is.

Day 5: Run a live pilot with one team

Choose one product line or one shift. Brief the team for five minutes: why you are doing it, what to scan and what to do if the scanner fails. Be on the floor the whole day. Write down every problem. By the afternoon you will know what to change.

Do not switch the whole factory on at once. One line working well is worth more than five lines working badly. Add the next line the following week.

In week two, widen the pilot, add the loading scan and start reading the numbers. By week four most teams stop thinking of it as a project and treat it as how the factory works.

Once scanning is running you will have more data than you can use. Resist the urge to build twenty charts. Three measures give you most of the value.

Made per day

The number of pieces scanned as made each day. Plot it for a few weeks and you will see your real capacity, including the effect of holidays, absences and problem products. It is the most honest number in the factory because nobody can argue with a scan.

Part made orders

The number of orders where some pieces are done and others are not. Every one of these is a delivery waiting on a missing piece. The list should be short and shrink each day. If it grows, a part of the process is leaking, often a shared component or one slow workstation.

Age in stock

How long finished pieces have been sitting before they are loaded. In a make to order business this should be days. A piece that has waited two weeks is telling you something: the customer is not ready, the order is incomplete or it was forgotten.

Review these in a ten minute meeting at the start of each day. Agree one action per problem and name an owner. Dashboards do not improve production, conversations about them do.

There is no shame in starting simple. The right tool is the one your team will really use every day.

Paper tickets and a whiteboard are a good way to prove the process before you invest. A spreadsheet is fine while one person owns it. Both stop working when several people need the same answer at the same time, or when a single order breaks into many pieces.

At that point, scanning software removes the double entry. Skuflo builds a production list from your orders, prints one label per piece and records a piece as made the moment it is scanned, with optional photos and worker names. You can try the whole flow on a 14 day free trial before spending anything on process change.

• Tracking too much on day one. Start with made and loaded.
• Asking operators to type. Scanning is quick, typing is not.
• Putting the scan somewhere inconvenient. Walking costs time, and people will skip it.
• Not telling the team why. People will scan reliably if they understand it saves them from being chased.
• Using the data to blame. If a scan leads to a telling off, people will find ways to avoid scanning.
• Never closing the loop. If nobody looks at the numbers, the team notices and stops caring.

The surest sign that tracking is working is boring: fewer interruptions. Sales stop phoning the floor, the supervisor stops being a human database, and the first thing you do each morning is read a list, not hunt for one.