6 ways to automate stock counts

#1 - The basic check-in

Who it's for: Single-site businesses doing a periodic count, anyone who needs to know what is actually on the shelf without running a full audit.

What it is: A stock count is a record of what you physically have, logged against what the system says you should have. This version captures five things per line: the item, the expected count, the counted total, the variance, and a note. Each line is one stamped record. The person counting runs the canvas once per item or category, and the day's count is the list of completions.

In practice: Take a single-site hardware shop. On the first of the month the assistant counts the paint aisle. They open the canvas, type "Matt emulsion 5L", read "Expected count" off the system as 18, count the shelf and the stockroom and put "Counted total" as 15, log a variance of minus 3, and write "two tins damaged, one sold but not scanned" in the notes. They submit and move to the next line. By the end of the aisle there is a row per product, each with a number against the system and a reason where the two do not match.

Why it works: The variance is the point. A count on its own only tells you what is on the shelf. A count against the expected number tells you whether the system can be trusted, and the note tells you why it cannot. The moment the count is logged on the canvas, it is a server-side record the manager can pull later, not a number on a clipboard that gets thrown out at the end of the day.

Steps included:

• 1 text input (item)
• 3 number inputs (expected count, counted total, variance)
• 1 text input (notes)

#2 - With written guidance

Who it's for: Sites with rotating staff who all need to count the same way, so the count from a Saturday casual matches the count from the full-timer.

What it is: The basic check-in plus two guidance panels woven through the canvas. The first panel tells the person how to work a shelf: start at the front, check the back of house and any displays, group by SKU, and ask before logging a zero. The second panel tells them what to do when a variance is big: log it, and flag a short on a high-value or fast-moving line to the duty manager straight away.

In practice: Take a three-site convenience chain. Stock is counted by whoever is on shift, which on a given week might be four different people across the three shops. Before guidance, one person counted only the shelf and another included the stockroom, so the same product threw a different variance depending on who did it. The "How to count" panel now puts everyone on the same path: shelf, back of house, displays, group by SKU. The "What to do about variance" panel tells them not to sit on a short of cigarettes or spirits but to flag it to the duty manager there and then. The variances stop being noise and start being something the area manager can read.

What it adds to the basic check-in:

1. A "How to count" panel before the inputs that walks the person through working a shelf in order and asking before logging a zero.
2. A "What to do about variance" panel after the inputs that separates a small variance from a big one and names who to tell.
3. A consistent method across everyone who counts, so the variance reflects the stock, not who happened to be on shift.

Why it works: Written guidance sits inline at the moment the person is about to count. They read the method the first time they open the canvas, and it is there again the next time. It is not a briefing they half-heard on their first day. The most common cause of a false variance is a misplaced case, and the guidance catches it before a zero gets logged in error.

Steps included:

• 1 guidance note (how to count)
• 1 text input (item)
• 3 number inputs (expected count, counted total, variance)
• 1 text input (notes)

#3 - With a signature

Who it's for: Businesses that need a name against each count, so a queried variance has a signed record and a named person to speak to.

What it is: The basic check-in plus a signature captured at the end of the count. The person who counted signs, confirming they are the one who did the count and that the numbers are theirs. The five items and the notes stay the same; the signature adds a name and a timestamp to the record.

In practice: Take a multi-site off-licence group where spirits and high-value wine walk if nobody is watching. Every count on a controlled line ends with a signature. The supervisor counts the spirits shelf, logs the variance, and signs. When the area manager reviews the week's counts, every controlled-line variance has a name against it. A short that nobody signed off stands out immediately as one to ask about.

What it adds to the basic check-in:

1. A signature naming who did the count, captured at sign-off.
2. A record of when the count was signed, next to the numbers.
3. Accountability, so a queried variance points at a named person rather than nobody.

Why it works: The name goes on the record at the moment the work finishes, not from memory later. When something is found short the next week, the conversation starts with who signed it off, which is specific instead of general.

Steps included:

• 1 text input (item)
• 3 number inputs (expected count, counted total, variance)
• 1 text input (notes)
• 1 signature (count sign-off)

#4 - With photo evidence

Who it's for: Teams wanting visual proof alongside the count, so a recount or a queried variance has something to look at.

What it is: The basic check-in plus a photo of a high-value shelf section or storage bay taken at the moment of count, with the count sheet clearly visible in the frame. The person captures one specific, telling spot rather than a vague wide shot. A variance is no longer just minus 7 with a note. It is minus 7 with a note and an image of exactly what that shelf section looked like at the moment of count.

What it adds to the basic check-in:

1. A photo of a high-value shelf section with the count sheet in frame, captured at the time.
2. Proof that holds up to an inspector or auditor, not just a ticked number.
3. A visual record kept alongside the numbers, so a recount can be judged from the image.

Why it works: A number is data. A photo is context. A variance of minus 7 on a high-value line could be a miscount, theft, wastage, or genuine sell-through, and the number alone cannot tell you which. The photo, taken at the moment of the count on the same device, shows the state of the shelf and lets the person reading it decide without a return visit.

Steps included:

• 1 text input (item)
• 3 number inputs (expected count, counted total, variance)
• 1 text input (notes)
• 1 photo of a high-value shelf section

#5 - With Poppi checking the photo

Who it's for: Multi-site groups where photos get taken and filed but nobody reviews them. Head office that cannot look at every site's photos in real time.

What it is: A photo-checked count is the basic check-in plus a photo of a high-value shelf section that Poppi (AI) reviews the moment it's saved. Poppi answers one question about that photo, set by you: is the count sheet clearly visible and the shelf section identifiable? If the answer is no, Poppi posts what it spotted to the team chat, so the problem gets fixed before the count moves on.

In practice: A five-site retail group counts on the first Saturday of the month. Each site takes a photo of a high-value bay after counting. Poppi reads each photo: shelf clear, count sheet visible, items identifiable. Most pass and nothing changes. On one photo the count sheet is obscured by glare and items are blurry. Poppi answers no and posts the reason to the team chat ("The count sheet is not clearly visible, and shelf items are blurry"). The site retakes the photo while the count is still in progress, and it passes on the second try.

What it adds to the basic check-in:

1. A photo of a high-value shelf section that gets checked the moment it's saved, not just stored.
2. A team chat message with Poppi's reason the moment a photo fails the check.
3. The manager stops being the only person who ever looks at count photos.

Why it works: The check happens in the seconds between the photo being taken and the count moving on. That is the only moment the problem is still cheap to fix. A manager reviewing photos the next day can only record that the photo is unusable; Poppi catching it at the time gets it retaken while the area is still in view.

Steps included:

• 1 text input (item)
• 3 number inputs (expected count, counted total, variance)
• 1 text input (notes)
• 1 photo of a high-value shelf section
• 1 Poppi decision (judges the photo against your question)
• 1 Poppi action (posts to the team chat if the photo fails the check)

#6 - With a team alert if the count doesn't finish by the deadline

Who it's for: Teams where a stock count is due by a set time and sometimes gets left unfinished, delaying reconciliation or holding up the next day's work.

What it is: A deadline alert is the basic check-in plus a Poppi (AI) action set to the workflow's end time. If the stock count isn't finished by the time it is due, Poppi posts a message in the Pilla team chat so the outstanding items get completed before the deadline passes. It watches the end time, so it catches the count that quietly got abandoned half-done, not just the one that was never started.

In practice: A five-site retail group counts stock on the first Saturday of the month, with a 3pm deadline so the office can reconcile by Monday. On a busy Saturday with customer traffic, it is easy for a smaller site to leave the last few sections uncounted. With this version, if a store hasn't finished by 3pm, the team chat gets a message, so the shift lead catches the unfinished areas that afternoon rather than the manager discovering on Monday that one site's count is still pending.

What it adds to the basic check-in:

1. A message in the team chat if the count isn't finished on time.
2. A catch for the count that got abandoned half-done, not just the one that was never started.
3. A record of when the count should have been done, next to when it was.
4. The manager finds out the same day, not when they sit down to reconcile two days later.

Why it works: The alert is tied to the count deadline, so an unfinished count raises its own hand. Nobody has to check the status board or chase each site; the deadline does the nudging.

Steps included:

• 1 text input (item)
• 3 number inputs (expected count, counted total, variance)
• 1 text input (notes)
• 1 Poppi action (posts to the team chat if the count isn't finished by deadline)

How to pick the right version

You don't need to know our product to choose. Every version here is the basic check-in plus one addition, so pick the additions your team actually needs.

Do other people run the count?

If it is just you and you know how to count a shelf, the basic check-in (#1) is enough. The moment rota staff or new starters count, written guidance needs to be on the screen so everyone counts a shelf the same way: #2.

Do you need a name against the count?

If knowing it was done is enough, skip this one. If you want who counted it on the record, #3 adds a signature.

Do you need photo proof?

A ticked count says the work was done; a photo shows it. If you want visual proof of a high-value section, #4 adds a photo of one telling spot like the spirits shelf with the count sheet visible.

Does anyone actually look at the count photos?

If a manager genuinely reviews every photo, #4's record is enough. If photos get taken and filed unseen, #5 has Poppi (AI) check each one the moment it's saved and tell the team chat when something's wrong.

Does the count have to be finished by a set deadline?

If a team sometimes leaves the count unfinished, #6 posts a message to the team chat when the deadline passes with the count still pending.

Need more than one addition? Open the version with the addition that matters most in the playground and add the others as steps. That is how the product works anyway: every option here is one step added to the same record.

Related workflows

• 4 ways to automate daily cash counts
• 4 ways to automate deliveries received
• 4 ways to automate damage reports
• 4 ways to automate tool and asset checkout
• 4 ways to automate lone worker check-ins

Conclusion

A stock count is a record of what you physically have, logged against what the system says you should have, with the variance and the reason captured per line. Every version above is the same basic check-in plus one addition: written guidance, a signature, a photo, Poppi checking the photo, or a deadline alert. Pick the ones your team needs and combine them in the playground.

→ Build your own stock count on Pilla.