Count inventory while your Shopify store is open: how to

Yes, you can count while orders keep coming, if you apply the count as a change and not as a total. Say Shopify showed 12 when you counted a line, and you found 10. Apply that as "minus 2". Every sale or delivery that lands before you finish stays in. Write "10" instead and you erase them.

The rest of this article shows why, with numbers you can redo.

A count is a photo, and the shelf moves after the shutter#

A stocktake (some people say stock take) records one moment. You count a shelf at 10:00. A customer buys two at 10:40. You press apply the next morning. The shelf has changed since your count, and your count has not.

Shopify's own guide to counting with POS Quick Count recommends counting when the store is closed or quiet. It also says that submitting a count adjusts inventory to match what you counted, and that if a sale lands between counting and submitting, you "may want to recount that item or make a manual adjustment". That is honest advice for a store with doors and opening hours.

An online store has no closing time. Orders arrive at 3 a.m. For these shops, "count when nobody buys" is not a plan.

Two ways to apply a count#

Overwrite. The tool takes your counted number and writes it as the new total. Simple, and correct only if nothing moved between your count and the write.

Change. The tool remembers what Shopify showed at the moment you counted that line. This is the baseline. When you apply, it reads the quantity Shopify holds now, takes back only the counting correction, and leaves everything else alone.

quantity to write = live now - quantity when counted + counted

Read it left to right. Start from what Shopify holds this second. Subtract the quantity Shopify held when you counted. Add what you counted. The gap between those last two numbers is the correction your count found. Everything that happened in between is already inside "live now", so it stays.

Worked example: counted 10, sold 2, applied the next morning#

Shopify showed 12 when you counted the line. You found 10, so the count says minus 2. That evening two units sell and the orders ship. The next morning Shopify shows 10 and the shelf holds 8.

Now the two methods side by side.

Overwrite writes 10. Shopify now says 10 and the shelf holds 8. Shopify is 2 too high, and nothing flags it.

Change writes this:

10 - 12 + 10 = 8

Shopify says 8. The shelf holds 8. Correct.

Same count, with a delivery#

Make it harder. Five units arrive after your count and are booked into Shopify. The baseline is still 12, you still counted 10, and two units still sold.

Live now:        12 - 2 + 5 = 15
Change method:   15 - 12 + 10 = 13
Shelf:           10 - 2 + 5 = 13

The change method gives 13 and the shelf holds 13. Overwrite writes 10, which leaves Shopify 3 too low.

When the result would go below zero#

Suppose the baseline was 12 and you counted only 1, so 11 are missing. Then six sell. Live now is 12 minus 6, which is 6.

6 - 12 + 1 = -5

You cannot hold minus 5 units. More left the shelf than your count said was there, so something is off: a miscount, or sales that shipped from another place. A good tool writes 0 and tells you to look at that line. Do not read a clamped zero as a clean result.

Whose fault is the gap? Receipts, returns and transfers break counts too#

Sales get the blame, but any change to the quantity after your count has the same effect. A delivery booked into Shopify, a return put back on the shelf, a transfer received from another location, a manual edit. The change method keeps all of them, because they are part of "live now".

It cannot keep a movement nobody recorded. Here is the pattern that goes wrong. A delivery of 5 sits on the floor of the section you are counting, still sealed in its box, and nobody has booked it in Shopify. You open the box, shelve the units and count them. Your count includes 5 that Shopify has never seen. Later someone books the delivery, and Shopify adds 5 again. Now the line is 5 too high.

Two habits fix this. Book in or set aside every delivery before you count the section. Or keep boxes sealed and out of the section until the count is applied.

One more trap sits in the word "quantity". Shopify's help center says the available quantity leaves out committed units, which are units in orders you have not fulfilled yet (inventory states). Those units may still be physically in your building, packed in a box by the door. A tool that compares against available expects those packed units to be left out of your count. Ask which number yours uses before the first scan.

Count in sections so each one is short#

A long count is a risky count. Every hour a line stays open, more can move under it. So cut the catalog into pieces you can finish.

Pick the cut that matches how you already think about stock. By collection, by supplier, by tag, by shelf or bin. A clothing shop might count one collection a day. A shop that buys from three suppliers might count one supplier's goods at a time. Our stocktake guide covers how to choose sections, and the page on cycle counts covers a rolling schedule.

Start with the fastest sellers, because they move the most in an hour. Apply them first, too.

A small count has a second benefit. When it disagrees with Shopify, the list of suspects is short.

Rules for a count that stays open for a day or more#

Some counts run past one sitting, especially with thousands of variants. These rules keep a long count honest.

1. Count one section per sitting, and finish each line in that sitting. If you come back to a line hours later and change the number, you are mixing two moments.
2. Book deliveries before you count their section. See the pattern above.
3. Pack and ship, or set aside, the orders for items in the section. Then the shelf and the available number mean the same thing.
4. Apply the same day if you can. A count applied at 5 p.m. has had hours to drift. A count applied three days later has had days.
5. Do not delete a nearly finished count because you are afraid to apply it. A team spends a day counting, finds the apply step scary and sees only two exits, apply or delete. The day is gone. Save the count without changing Shopify instead. In Binly that is Mark as reviewed, and Reopen puts it back to counting later.
6. Write the start and end time on the count. When you investigate a gap, the first question is what happened between those times.

After applying, check a few lines in Shopify#

A count is only done when Shopify shows the right number. Pick five lines, including two with big differences, and open them in Shopify.

Do not expect Shopify to equal your counted number. It should equal your counted number plus whatever moved since. In the first example you counted 10 and Shopify shows 8, which is right, because two units sold. Binly runs this check itself after you apply and says "Checked in Shopify" with how many changed lines now match. A manual look at five lines is still worth doing the first time you trust a new method.

What no tool can fix#

The change method repairs timing. It does not repair a count that was wrong when you made it, a unit that left the building without a record, or a box that moved between sections while you walked. Each of these can leave a gap after a careful count, and our guide to why Shopify inventory does not match the shelf goes through each one.

Two limits, stated plainly. First, recounting a line late can cancel the protection. Say the baseline was 12 and you counted 10 at 10:00. Two units sell, so live is 10. At 3 p.m. you recount, find 8, and type 8. The tool measures that against the 10:00 baseline.

10 - 12 + 8 = 6

The shelf holds 8, so Shopify is 2 too low. The 8 you typed already contains the two sales, and the formula takes them off a second time. If you must recount a line after sales, write a note on the count and check that line in Shopify afterward.

Second, stock in transit is invisible to a count. Units on a truck are in neither your section nor your shelf, so nothing you scan can account for them.

Tomorrow, pick one section, count it start to finish in one sitting, apply it, and open five of its lines in Shopify. Then you know how your own store behaves.