DEEPINGSOURCE
ANALYZE · Reads the why

Read what happened in your space.
See the behavior behind it.

saai insight reads POS, weather, inventory, and spatial behavior together.

Request Consultation
ProductsCurrent product: saai insight· Analyze

Where it breaks

The outcome is recorded. The behavior before it is not.

  1. 01

    Only outcomes

    POS records the payment. What happened after someone walked in is not kept.

  2. 02

    Causes are guessed

    The reason sales moved gets filled in with experience and hunches.

  3. 03

    Hard to compare

    Conditions differ store by store, so the numbers do not line up.

Operating the space

Read the behavior before checkout, on one screen.

Visits, paths, dwell, and conversion sit on the same screen. Layout, comparison, drop-off points, and the effect of a change share one basis for the judgment.

saai insight

Today

Store Analytics Dashboard

Visits peak at 2–4 PM

Est. daily sales
0.0k
Total visitors
0
Avg. dwell time
0.0min
Shelf conversion
0%

Beverage shelf dwell time is 2x the usual. New-product promotion effect confirmed.

Layout

Display · layout

See where a zone stalls.

Comparison

Space variance

Compare stores on like conditions.

Drop-off

Entry · dwell · conversion

Find pauses and exits.

Validation

Before · after

Compare dwell and conversion after a change.

Turn on only the analyses you need

* Multi-store dashboard and agentic AI are scoped separately.

  • Live dashboard
  • Zone analysis
  • Visitor pathways
  • Heatmap
  • Funnel analysis
  • CSV export

Zone · fixture calls

Go down to the zone and the fixture, all on one screen.

From live flow to accumulated periods, then narrowed to zones and fixtures. Analysis proposes the candidates. The operator decides.

Period

Analysis dimensions

7 days

Cross time, people and behavior on the same map.

Gender
Age
Heatmap

1,576

store entries

1,672

zone transitions

2.9s

avg. dwell

Gender and age segments are virtual segments the simulation models. They are not measured personal attributes.

Decision steps

Bring hidden lunchtime snack demand to the entrance

01 / 07 · Observe

Short lunchtime visits repeat through the store.

Some fast-moving visitors reach the snack aisle at the back, then leave without a long stay.

Observed window

Weekdays 11:30–13:30

An illustrative view modeled on real convenience-store operating patterns. The data is synthetic.

3× speed

0s · 0 customers moving

Map tools

Fixtures

  1. 1Pass
  2. 2Enter
  3. 3Dwell
  4. 4Gaze
  5. 5Pick up
  6. 6Buy
SIDEWALKENTRANCEAmbient foodA · Wall aisleChilledE · Chilled aisleB · Aisle 1Instant noodlesC · Aisle 2Household goodsD · Aisle 3SnacksBeverage endcapDining counterPromoPOSstaff-1 · patrollingcashier-1 · cashierpasser-1 · passingpasser-2 · passingStaffPasserby

See how people are spread across the five zones at this moment.

Illustrative simulation in a virtual space. Values do not represent actual performance.

Data you can add

Add POS, weather, and inventory to narrow where to look.

Operational data you provide is joined with visit, path, and dwell signals from the space.

Connected dataResults · external variables

  • POS · sales−8%
  • Weather · temperature31℃
  • Stock · promotion2

Connect data provided by the customer

saai insightWhy · where

Dwell at the drinks shelf +100%

Comparison question

Where, and why, did it change?

Put outcomes and signals on the same period and store basis to inspect the cause.

* Sample screen · illustrative data

How we read numbers

We never compare raw numbers across different conditions

Store size, hours, and day of week change what a number means. We normalize to one basis before comparing.

Before normalizing

Revenue this month

₩32M

After normalizing
38 comparable storesThis store · 12th

12th out of 38 comparable stores.

Visits went up but dwell went down. What should we check alongside it?

Curious how this looks with your own space data?

Request a consultation

ONBOARDING

How onboarding works

  1. First visit

    We review the site and select and connect the existing CCTV with the best angles.

  2. Build & ship

    A compact AI analysis device is configured and shipped to your site.

  3. Second visit

    We install the device, verify it runs cleanly, and set up your custom dashboard.

Get started about three weeks after the first visit. Timing can change with site conditions.

From insight to action

saai insight shows why. saai agent proposes what to do next.

Turn an insight into an action card

Beverage dwell is up? saai insight surfaces the relevant signals. saai agent drafts options for quantity, supplier, and timing. You approve the next step.

See saai agent

Continue your review

ENTERPRISE INSIGHT

See what changed and why, every morning.

Start with the CCTV you already have. It takes about three weeks after the first visit.

Request a consultationSee saai agent