Why does my bolt count on paper never match what is on the shelf at inventory time?

Your paper record says 6.5 yards. The yardstick says 5.75. Nothing was stolen. The gap is the accumulated residue of a dozen ordinary counter habits, each one too small to notice on the day it happened: the cut that ran generous, the sample pulled for the display wall, the return that came back to the shelf but never back to the record, the eighteen inches sold as a remnant from a bolt nobody logged against.

Fabric drift is different from hard goods shrinkage because the unit is continuous. A rotary cutter either is or is not on the shelf. A bolt is a number that can be wrong by a quarter yard in either direction forever, and errors do not cancel out, because almost every one of them runs in the same direction: less fabric than the record says.

Below are the specific mechanisms, roughly in order of how much yardage they usually account for, and the counter routine that closes each one. The dollar figures are worked from stated assumptions so you can swap in your own.

The generous cut and what a quarter inch per sale costs in a year

Nobody wants to send a customer home short. So the cutter squares up, adds a little for the straightening the customer will do at home, and cuts a hair past the line. Every time.

Work the arithmetic. Assume 60 cuts a day, six days a week, fifty weeks: 18,000 cuts a year. Assume a quarter inch of overage on each one, which is 0.00694 yards. That is 125 yards a year given away.

At an assumed landed cost of $6.75 per yard, that is $844 of fabric. At an assumed shelf price of $16 it is $2,000 of retail value that never rang up. And critically, it is 125 yards that your record still thinks you have.

The fix is a policy, not a scolding

Decide what the overage is and make it deliberate. Many shops cut a consistent half inch over on every cut and simply record the sale as the customer's yardage. That is a choice, and it is fine, as long as the record accounts for it. The alternative is to record the actual cut. Pick one, write it on a card at the cutting station, and make sure every part-timer cuts the same way.

Keep reading: Should I run beginner classes myself or pay a teacher a cut of every seat sold?

Preshrink, straightening the cut edge and mill yardage tolerance

Yardage loss starts before the bolt reaches the shelf. Three sources:

• Mill tolerance. A bolt marked 15 yards is not guaranteed to be exactly 15. Short bolts happen and long bolts happen, but you only find out by measuring.
• The straightening cut. Trimming the leading edge square costs an inch to several inches on a fabric that came off the roll badly.
• Flaws. Mill flaws marked with a flag mean you cut around them and lose whatever sits between.

The habit that fixes this costs about ninety seconds per bolt: measure the bolt when you unpack it, before it goes on the shelf, and record the measured yardage, not the yardage printed on the end. If you find a consistent short measure from one distributor, that is a credit request, not a rounding error.

Sample yardage and shop models pulled without a record

Every shop model, class sample, block of the month display and trunk show piece started as bolts on your own shelf. That fabric is a legitimate marketing expense and often the best money you spend. It is also the single biggest source of unrecorded yardage in most shops, because it is pulled by the owner, at night, in a hurry, when the register is closed.

A sample quilt might consume eight to twelve yards including backing. Six samples a year at ten yards each is 60 yards, roughly $405 at the assumed $6.75 landed cost. If none of it is recorded, your on hand figures are 60 yards optimistic and you will reorder late on every one of those bolts.

Treat sample pulls as a transaction type, not an exception. Ring them through as an internal sale at zero price against a "shop sample" customer or account. The yardage leaves the bolt record. The cost lands in a marketing bucket where you can actually see what your display wall costs you per year. Both facts are useful.

Keep reading: How do I build a class kit that actually turns a profit instead of eating my scraps?

Returns, exchanges and voided sales that never reverse the yardage

Cut fabric is usually final sale, so returns are rare, which is exactly why nobody has a routine for them. The problems are quieter than a return:

• Voided sales. The transaction is voided at the register, so the money is right, but the fabric was already cut and it goes back in a bin or onto the shelf as a remnant. The bolt record was decremented and never restored.
• Exchanges. A customer swaps two yards of one print for two yards of another. Two bolt records need to move, and usually only one does.
• Held cuts. Fabric cut for a phone order that is never collected. It sits behind the counter for weeks, then goes back to the shelf unlogged.

Rule: fabric physically going back onto a bolt or into remnants must generate a yardage entry, always, even when no money moves. Attach it to the void or the exchange so it is one action, not two things a busy person has to remember.

Ends of bolt sold as remnants outside the bolt record

The last two yards get pulled off the cardboard, folded, priced by hand and dropped in the remnant basket. From that moment the bolt record is either sitting at two yards it does not have, or it was zeroed and two yards of real inventory now exist nowhere in your system.

Both versions cause trouble. The first makes you think a print is still available when it is not, and you will promise it to a customer finishing a border. The second means you cannot answer the question every quilt shop gets asked: do you have any more of this?

Handle the transfer explicitly. When a bolt drops below your remnant threshold, say one yard, close the bolt record and open the remaining piece as a remnant with its own yardage, its own price and a link back to the original SKU. It takes fifteen seconds and it keeps the answer to "any more of this" honest.

See how QuiltCounter handles this for quilt shops and fabric retail

A cycle count rotation that does not close the shop

A full physical count of a wall of bolts is a two day job you can only do once a year, which means eleven months of every year you are working from numbers of unknown quality. Cycle counting fixes that by counting a slice at a time, continuously.

A workable rotation for a shop with roughly 1,200 bolts:

1. Divide the floor into zones of about 50 bolts each, following how the shelves are actually arranged: by color wall, by manufacturer, by batiks, by wide backs, by flannel.
2. Count one zone per open day. Fifty bolts measured with a yardstick is about 25 to 35 minutes for one person, done in the first slow hour after opening.
3. Count high value and fast moving zones more often. Batiks, wide backings and current collections twice a cycle; long tail basics once.
4. Record the variance, not just the new number. The variance is the data. A bolt that reads a half yard short every single count is telling you something about a habit, not about that bolt.
5. Review variances monthly and look for the pattern: one zone, one shift, one product type.

At 50 bolts a day across six days, a 1,200 bolt shop completes a full pass every four weeks without ever closing.

Approach	Time cost per year	Days per year with fresh numbers
Annual full count	Two people, two days, shop closed	Roughly 30
Daily cycle count, 50 bolts	Roughly 30 min per open day	Every day

The cycle count costs more total hours. It buys you accurate numbers every day of the year instead of one week in January, and it never costs you a day of trading.

Setting a shelf minimum that triggers reorder before the bolt is dead

Drift hurts most at the reorder point, because a record that overstates yardage delays the reorder until the bolt is already unsellable. A customer needs three yards for a backing and you have 1.8.

Set the minimum from your actual usage, not by feel:

• Estimate weekly usage per bolt. If a print sells about 3 yards a week, that is your rate.
• Add lead time. If the distributor takes two weeks and you order weekly, your exposure window is about three weeks, so 9 yards.
• Add a safety buffer for a backing sized request, typically 3 to 4 yards.
• Reorder point is roughly 12 to 13 yards on a bolt like that. On a slow basic selling half a yard a week, the same math gives about 5.

Then set a floor for what is even worth keeping on the shelf. Below about a yard and a half, a bolt cannot fill most requests and is occupying shelf space that a full bolt could use. That is the moment for the remnant transfer described above.

Getting the record and the shelf to agree

None of these leaks is dramatic on its own. The generous cut, the unlogged sample, the remnant that left the record, the void that never reversed: each is a quarter yard here and ten yards there. Together they are the reason your January count always comes up short and you can never say quite why.

Every fix above is a habit, and habits hold when the system asks for the entry at the moment the fabric moves. QuiltCounter tracks yardage on each individual bolt, records sample pulls and remnant transfers as their own movements, keeps a running variance history per bolt so you can see which habits are costing you, and fires a reorder alert at the yardage you set rather than when someone notices an empty spot on the wall.