Retail Store Audit App With Photo Proof: How Scored Audits Replace the Clipboard

What the app does that the clipboard cannot

A clipboard audit records that a person believed a thing was right at some point on a day. It cannot record when, cannot show what the person saw, cannot weight one item above another without a calculator, and cannot make anyone fix anything. The photograph of the completed sheet, posted to a group with "Audit done sir, 92 percent", proves only that a sheet was filled.

The app records who, where, when and what, item by item. Each item is a live photo taken in the app at that moment, a yes, no or scored answer, and a comment. The score is computed from weights the brand set, the fails become tickets with owners, the PDF goes to the store, the region and whoever owns each fix, and the same audit in 200 stores lands on one ranking. The audit stops being a document and becomes a dataset.

The app also removes the argument. When the store manager says "Woh mannequin toh subah theek thi" and the area manager says it was not, the photo with the timestamp settles it in a second, and the conversation moves to the fix.

Anatomy of a scored audit item

Why the photo must be live

The whole value of photo proof rests on the photo being taken in the store, at that moment, by that person. The moment an app allows a gallery upload, the audit is back to a claim: last Tuesday's window, or the flagship's window, will do for this Tuesday. The pattern is common enough that it has a name on the floor: "purani photo chalegi?" The methods for stopping it, and what a backdated photo looks like in the data, are at /learn/how-to-stop-backdated-store-photos-and-fake-checklists/.

Three locks make a photo evidence. The camera opens inside the app and the gallery is not an option. The GPS position rides with the photo and is compared to the store's registered location, with a tolerance for a basement showroom that loses signal. The time is the server's, not the phone's, so a phone with its clock set to yesterday still submits today. A fourth, optional lock is the audit question on the item, because a person who has to answer "is the brown coat on mannequin 2" has to look at mannequin 2.

Scoring: weights, fatal items and the pass mark

A good audit score is one the store manager can predict before the area manager arrives. That requires three design choices to be public inside the chain.

• Weights by what the customer sees first. Facade and window carry the most in a showroom because they decide whether a walk-in happens at all; grooming and the billing counter carry the most in the last three metres of the visit; back of house carries the least. A typical split gives the front 40 percent, the floor 35 percent, people 15 percent and back of house 10 percent, and the split is a brand decision, not a vendor default.
• Fatal items outside the arithmetic. A fire exit blocked by cartons, an expired licence or a demo unit with an exposed wire should cap the audit at zero however good the window is. Fatal items are few, listed, and never argued.
• A pass mark that means something. A pass mark of 80 with a network average of 91 tells nobody anything. Set the mark where the top third of stores land in the first month and raise it each quarter.
• Scored levels with definitions. A 0 to 5 scale where 3 means "in place, minor deviation from the docket" and 1 means "in place, wrong campaign" is worth more than a yes or no, because the trend from 3 to 4 is visible.

Building your first scored audit in seven steps

1. Walk one store with the display standards document and write down every checkpoint a customer could notice, in the order a customer meets them: facade, window, entrance, floor, displays, trial rooms or demo bay, billing counter, staff.
2. Cut the list to 25 to 40 items. A 90-item audit takes two hours and is done badly; a 30-item one takes 40 minutes and is done every fortnight.
3. Write each item as a question a photo can answer, with the docket reference in the text. "Window good" becomes "Window dressed to docket W39 with campaign backdrop".
4. Set weights and fatal items with the VM head and the operations head in one room, and publish them to the stores. A score nobody can predict is a score nobody respects.
5. Decide the cadence per role: store manager self-audit weekly, area manager audit monthly, VM manager audit at every docket change, HQ surprise audit quarterly. The same template, different auditors, so the gap between self-score and area-manager score becomes a number.
6. Wire every fail to a ticket with an owner and a deadline, and every fatal to an immediate escalation. An audit with no consequence is a survey.
7. Run it in ten stores for a month before the network, and change the wording of any item that scored the same in every store. An item that is always 5 is not measuring anything.

What to do with the score

The audit score is most useful in three reads. The first is the ranking: the same audit, the same weights, every store on one list, with the bottom ten named for the area managers' Monday call. The second is the repeat miss: a store that fails the same item three audits running has a training problem or a supply problem, not a discipline problem, and the pattern is visible only when audits are data. The third is the gap between self-audit and external audit, store by store, which tells you which store managers see their own store honestly.

The score is also the input to the store visit. An area manager who has read the last three audits and the open tickets before leaving the office can spend the visit coaching rather than inspecting, described at /learn/store-visit-report-app-for-area-managers/. And the score becomes one half of the store's overall grade once it sits next to a measure of selling, which is the argument at /learn/how-to-measure-in-store-execution/.

The measured way: BorentisOps

BorentisOps at /solutions/products/borentisops/ runs scored audits exactly as described: a no-code builder for sections, photo fields, yes-or-no and 0 to 5 audit questions, weights, fatal items and conditions; a live photo with GPS and server time on every item and no gallery uploads; a ticket with an owner and an SLA raised on every fail; a PDF with photos, scores and actions in one tap to the store, the region and the fix owners; and the same audit across the network on one ranking with trends. Drishti, the visual merchandising AI, can score the window and display items from the same photo so the auditor's judgement and the model's sit side by side.

The audit score lands on the store scorecard next to the sales-conversation score from Borentis Floor, so a store that passes the audit and loses the walk-ins is visible as one row. It is priced per store, works offline and shows item text in Hindi, English and Hinglish.