Customer implementation story

UzCarlsberg: photo-led shelf audits

A shelf photo becomes evidence for the audit, facing count and share-of-shelf calculation. Planned and recorded visits give managers a connected view of merchandiser coverage.

Book a working session All cases

Client: UzCarlsberg
Industry: Brewing and beverage distribution

Mobile outlet audits
Shelf-photo recognition and automatic facing counts
Share-of-shelf calculation from audit data
Route tracking and monitoring of planned visits
Territory coverage based on merchandiser visits

Evidence level

Evidence boundaries

Evidence level: Verified mechanism
Implementation scope: Mobile outlet audits · Shelf-photo recognition and automatic facing counts · Share-of-shelf calculation from audit data · Route tracking and monitoring of planned visits · Territory coverage based on merchandiser visits
Operating model: Not disclosed
Measurement window: The public source does not state one measurement period; this page presents the verified implementation mechanism.
What is verified: The implemented mechanism is shown. Outcomes without a verified period, baseline and method are not published.
Country: Uzbekistan

Shelf evidence you can review

Connect the shelf photo, facing count and share of shelf.

Explore five views of the workflow implemented at UzCarlsberg: mobile audits, shelf photos, recognition, share-of-shelf calculation and merchandiser routes.

Choose a view or let the walkthrough advance automatically.

Shelf photoAudit evidence

AI Shelf Recognition

This illustration explains the recognition workflow. It is not a screenshot of the customer’s system.

01 · Audit

Run a consistent mobile audit at the outlet

What is recorded: The result is recorded at the outlet, not after the employee returns.
How it works: The merchandiser follows the audit in the mobile app and submits the data from the outlet.
What you can check: Which audit was completed and what evidence supports its result.

01 · Audit

Run a consistent mobile audit at the outlet

What is recorded: The result is recorded at the outlet, not after the employee returns.
How it works: The merchandiser follows the audit in the mobile app and submits the data from the outlet.
What you can check: Which audit was completed and what evidence supports its result.

02 · Photo

Keep the shelf photo as the basis for the report

What is recorded: The photo records the actual shelf display at the time of the visit.
How it works: The image stays in the system alongside the audit result.
What you can check: Whether the report can be checked against the original shelf photo.

03 · Facings

Count product facings from the shelf photo

What is recorded: A manual count can differ from what is entered in the report.
How it works: Photo recognition counts facings - the visible product fronts on the shelf.
What you can check: Which facings were recognised and what image supports the count.

04 · Shelf share

Calculate share of shelf from the same audit

What is recorded: The assessment needs to reflect the display recorded at the outlet.
How it works: Share-of-shelf calculation uses the mobile audit and recognition data.
What you can check: How the display is represented without manually transferring the count.

05 · Routes

Compare planned visits with recorded outlet coverage

What is recorded: Fragmented tracking leaves managers without a current view of routes.
How it works: The system tracks planned visits over time and builds coverage data from recorded visits.
What you can check: Which outlets merchandisers reach regularly and which routes need review.

A connected evidence trail

From the visit to the calculation - with the photo still in view.

Each step uses the preceding data and stays connected to the original shelf image.

Visit

The merchandiser opens the mobile audit.

Photo

The actual shelf display is recorded.

Recognition

The image provides the basis for the facing count.

Calculation

Share of shelf is calculated from the audit data.

A mobile workflow

Complete the audit where the shelf is.

One sequence keeps the visit, photo and result together.

01
Open the auditStart the audit for the outlet.
02
Check the displayFollow the agreed audit sequence.
03
Take a photoRecord the shelf as it appears during the visit.
04
Submit the resultSave the audit in the system from the outlet.

Routes and visit coverage

Compare planned visits with the visits recorded.

Territory coverage comes from merchandiser visits, not from a separate manually compiled report.

01Outlet AVisited

02Outlet BVisited

03Outlet CPlanned

04Outlet DNeeds review

05Outlet EPlanned

PlannedWhich visits are scheduled.

RecordedWhich outlets were actually visited.

CoverageWhere the route needs attention.

What changed in the process

From manual counts to a shelf audit backed by a photo.

Before

The count and report were separate

Merchandisers counted facings by hand and transferred the results. Routes and visit coverage were tracked separately.

After

The photo connects evidence and calculation

Mobile audits, photos, recognition, share of shelf and recorded visits form one connected workflow.

Confirmed implementation scope

What the UzCarlsberg case brings together

01Mobile outlet audits
02Shelf-photo recognition and automatic facing counts
03Share-of-shelf calculation from audit data
04Route tracking and monitoring of planned visits
05Territory coverage based on merchandiser visits

Questions for your process review

What to check before a pilot

Mobile audits

Is the visit result recorded at the outlet or after the employee returns?

Photos

Can a report be checked against the original shelf photo?

Facings

How do you check a manual count against the actual shelf display?

Shelf share

Which data is used to calculate share of shelf?

Coverage

Which planned visits are supported by a recorded outlet visit?

Evidence and its limits

The implemented workflow, without an unverified percentage.

The public source does not provide a consistent measurement period, baseline and methodology. This page therefore does not repeat the earlier percentage claim about merchandising time.

Measurement period: The public source does not state one measurement period; this page presents the verified implementation mechanism.
What the public evidence supports: The implemented mechanism is shown. Outcomes without a verified period, baseline and method are not published.

Book a working session All cases

UzCarlsberg: photo-led shelf audits

Evidence boundaries

Connect the shelf photo, facing count and share of shelf.

Run a consistent mobile audit at the outlet

Keep the shelf photo as the basis for the report

Count product facings from the shelf photo

Calculate share of shelf from the same audit

Compare planned visits with recorded outlet coverage

From the visit to the calculation - with the photo still in view.

Visit

Photo

Recognition

Calculation

Complete the audit where the shelf is.

Compare planned visits with the visits recorded.

From manual counts to a shelf audit backed by a photo.

The count and report were separate

The photo connects evidence and calculation

What the UzCarlsberg case brings together

What to check before a pilot

Is the visit result recorded at the outlet or after the employee returns?

Can a report be checked against the original shelf photo?

How do you check a manual count against the actual shelf display?

Which data is used to calculate share of shelf?

Which planned visits are supported by a recorded outlet visit?

The implemented workflow, without an unverified percentage.

Related evidence