CH 00

Prove every offer with the cameras you already have.

See how many shoppers reach each offer display, how long they stay, and how that compares with the weeks before. Numbers you can show whoever paid for the display.

Request a pilot See a sample offer report

RUNS ON A PC IN YOUR STORE · NO FACE RECOGNITION · NO VIDEO UPLOADED

AISLE 2 END-CAP · ZONE

CH 05AISLE 2 END-CAPSAT 12 SEP 2026 · 18:40

+27%

more shoppers at the Saturday offer than on the two Saturdays before

BEFORE 414 A DAY · DURING 526 · STORE +2%

38% OF VISITORS STEPPED IN · 19 S PER VISIT · 44% STAYED 8 S+

SAMPLE

CH 01ENTRANCE

1,384 came in today

SAMPLE · SAT 12 SEP

CH 02FLOOR

37 in the store now

SAMPLE · SAT 12 SEP

CH 03AISLES

58% stepped into fresh produce

SAMPLE · SAT 12 SEP

CH 04ROUTE

Entrance → Aisle 2 offer → Checkout · most common route on this camera

SAMPLE · SAT 12 SEP

A sample store with example numbers — not a customer result.

Sales show what sold. Cameras show how many reached the display.

When an offer sells slowly, sales data can’t tell you why. Maybe shoppers never reached it. Maybe they reached it, stayed and walked on. The first means move the display. The second means change the offer. Counting the shoppers who reach the display, and how long they stay, tells you which.

Mark the display a week early. Read its lift every day.

Draw the display on the camera view at least a week before the offer, then set the offer dates. Retailytics counts how many shoppers step in, how long they stay and how many stay 8 seconds or more. It compares the offer days with the same weekdays before, next to the change in store visitors, so a busy week isn’t mistaken for a good offer.

CH 05AISLE 2 END-CAP · 2-FOR-1 SNACKS

BEFORE 414 A DAY · DURING 526 · STORE +2%

AISLE 2 END-CAP · STEPPED IN

SAT 29 AUG402

SAT 5 SEP426

SAT 12 SEP526 +27%

STORE

BEFORE1,361 A DAY

DURING1,384 +2%

SAMPLE · SAT 12 SEP

CH 05AISLE 2 END-CAP

HIT RATE38%: Share of the store’s visitors who stepped into the section. Needs a door line at every entrance.
TIME SPENT19 S: Average time per visit inside the section.
ENGAGED44% · 231 VISITS: Share of visits that lasted 8 seconds or more.
LIFT+27%: Change in daily visitors to the section against the same weekdays before the offer. It counts people, not sales.

SAMPLE · SAT 12 SEP

A new product pulls shoppers to its shelf, or it doesn’t. Now you’ll see which.

CH 06AISLE 4 END-CAP · COLD-BREW COFFEESHELF VISITORS

DAY 01118
DAY 02121
DAY 03126
DAY 04124
DAY 05133
DAY 06139
DAY 07151
DAY 08147
DAY 09150
DAY 10158
DAY 11162
DAY 12166
DAY 13171
DAY 14176

SAMPLE · DAY 14

CH 06COLD-BREW COFFEE

12 S PER VISIT · 21% STAYED 8 S+ · RISING

SAMPLE · DAY 14

Mark the launch shelf. From the day you mark it, Retailytics counts how many shoppers reach it and how long they stay, and shows whether that rises or fades, so you can move it, back it with an offer or give it more time while the decision still counts.

It measures the shelf, not the pack: if the launch shares a bay, you see the bay.

Every corner on camera, ranked by the shoppers it draws.

Hit rate is the share of your visitors who step into a section. Put the next display where shoppers already go, and see which corners they skip.

Staff for the rush you actually get.

Hour-by-hour and weekday patterns from your own door counts, not a Saturday someone remembers.

CH 03AISLES

Fresh produce 58%
Aisle 2 end-cap 38%
Aisle 4 end-cap 13%

SAMPLE · SAT 12 SEP

CH 01ENTRANCE

SAT 17:00–18:00 · 164 ENTRIES

SAMPLE · LAST 4 WEEKS

Follow the route, not the person.

Each shopper is a numbered dot: no face recognition, no name, and a new number every visit. In each camera’s view you see the routes people take, and how many leave without reaching the checkout.

CH 04ROUTECAM 01

SAMPLE · SAT 12 SEP

CH 04CAM 01

Entrance → Aisle 2 offer → Checkout · most common route on this camera

SAMPLE · SAT 12 SEP

Your footage stays in the store.

ON-PREMISE

Detection runs on a Windows PC in your store, on the same network as your cameras.

RUNS ON A PC IN YOUR STORE

NO IDENTITIES

No face recognition and no age or gender guessing. Shoppers are numbered dots, and Retailytics saves no video or images; your NVR keeps recording as it does today.

NO FACE RECOGNITION

OFFLINE

Runs without an internet connection. Licences activate from a key file, with no account and no activation server.

NO VIDEO UPLOADED

CAMERA
STORE PCNO UPLOAD
YOUR SCREEN

YOU CHOOSE WHAT TO EXPORT

It reads the streams your NVR already records.

Paste each camera’s RTSP address (whoever installed your CCTV can find it), then draw the doors and the offer display. The answers fill in while the cameras run.

CAM 01

AISLE 2 END-CAP · ZONE

SAMPLE · SCHEMATIC

SETUP

RTSP IP CAMERAS · NVR / DVR WITH RTSP · VIDEO FILES · WINDOWS PC · NVIDIA GPU OPTIONAL

Dashboard languages: English · বাংলা · العربية · Español

Questions buyers ask first

Do I need new cameras?

Usually not. Retailytics reads the RTSP streams your IP cameras or NVR already provide, or recorded video files. It needs a camera that can see each door you want to count and each display you want to measure. If none can, one may need re-aiming or adding; check your local CCTV rules first.

Does any video leave the store?

No. Video is processed on the store PC, and Retailytics never saves or sends it. It keeps counts, times and numbered routes, and those leave the store only in the PDF and CSV reports you choose to export.

Does it recognise faces, age or gender?

No. It detects the shape of a person, not who they are, and follows each one as a numbered dot while they’re in view. It doesn’t estimate age, gender or mood, and it can’t tell when the same person comes back.

Does it count staff and promoters?

Yes. It counts everyone a camera sees. Keep door lines and offer displays away from staff routes where you can, and compare days with similar staffing.

When should we start before an offer?

At least a week before it goes up. Lift compares the offer days with the same weekdays before it, so those days need to be measured.

Can it follow a shopper from one camera to the next?

No. Routes and checkout reach are measured within each camera’s view. To see an entrance-to-checkout route, one camera needs to see both.

How accurate are the counts?

It depends on camera angle, lighting and how crowded the store is, so we don’t quote one figure. In a pilot, count one door by hand for an hour and compare. Every report also shows its coverage: how much of your trading time was actually measured.

Do we need to tell shoppers?

Probably. Using CCTV to measure how a store is used can be a new purpose under data-protection rules in several countries, including South Korea. Check the rules where you trade and update your CCTV notices if needed.

What does the store PC keep, and for how long?

Counts, times and numbered routes for each session, plus zone and campaign results. No video or images. You choose how long records are kept, in Settings.

What PC do I need?

A Windows PC on the same network as your cameras, switched on during trading hours. An NVIDIA graphics card lets one PC keep up with more cameras. Tell us your camera count in the pilot request.

Can head office see every store?

Today each store runs its own dashboard, and reports export to PDF and CSV for head office. Tell us how you’d want a multi-store view to work.

Which languages does the dashboard speak?

English, Bengali, Arabic (right to left) and Spanish. Korean isn’t available yet; tell us if your stores need it.

YOUR STORE

Prove your next offer.

Tell us about your stores, your cameras and your next offer or launch. We’ll plan a pilot that starts measuring a week before it goes up, so there’s a real “before”, and you judge Retailytics on the report you’d send to head office or a supplier.

Request a pilot

Licences that work offline

MONTHLY: Month to month, for a season or a campaign.
YEARLY: A year at a time, renewed with a new key.
PERPETUAL: A key with no end date.

Every licence activates offline with your installation ID. Ask us what each includes. Request pricing