Retail KPIs Every Retailer Should Track, from the Store Floor to Omnichannel

A retailer without measurable targets is guessing. Most store KPIs fall into three groups: sales, inventory and customers. The right mix differs by store format, but a core set applies almost everywhere, and AI video analytics can now measure the customer-facing ones automatically.

There is a second reason to measure the store carefully. Online, every view and click is recorded. In a physical store, the billing counter records the sale and little else. Closing that gap is what lets retailers run stores, websites and apps as one business, which is the idea behind unified commerce. The last section of this guide covers it.

 

The retail KPIs sheet

KPI	Formula	What it tells you	How video analytics helps
Conversion rate	Invoices ÷ footfall × 100	The share of visitors who buy	Automatic footfall count gives the denominator
Footfall	Visitors counted per hour, day or week	Whether campaigns, displays or a new store are drawing people in	Hands-free counting, comparable across stores
Dwell time	Average time in store or zone	Engagement, and whether visitors found what they wanted	Time per visitor and per zone
Repeat visit rate	Repeat visits ÷ total visits × 100	How well you retain customers	Opt-in loyalty alerts when members visit
Customer mix	Share of visitors by age band and gender	Whether assortment and campaigns suit who actually visits	Demographic analysis by hour
GMROI	Gross margin ÷ average inventory at cost	Profit earned per rupee invested in stock	Indirect: better buying decisions
Sell-through	Units sold ÷ units received × 100	How quickly stock moves before markdown	Indirect: demand signals by store
Shrinkage	(Recorded − actual stock) ÷ recorded × 100	Losses from theft, error and damage	Alerts for restricted-area entry and suspicious activity

 

Worked example. If 100 people enter a store and 40 invoices are raised, the conversion rate is 40%. One invoice can cover a group, so treat the figure as a trend to track rather than an exact count of buyers.

Customer retail KPIs in more detail

Footfall

Visitor count is the base metric because most others depend on it. Counting it consistently shows whether a holiday campaign, a new window display or a loyalty launch actually brought people in. TORK’s footfall analysis removes the manual counting and lets you compare invoices against visitors.

Dwell time

A visitor who stays long and buys has found what they came for. One who stays long and leaves empty-handed probably wanted something they could not find. Read dwell time alongside conversion rather than on its own.

Loyal customer visits

The widely quoted 80/20 rule of thumb says most revenue comes from a minority of customers, so repeat visits deserve their own KPI. Many stores track loyalty by phone number at billing. Opt-in recognition adds a signal at the door, so staff can respond while the customer is still in the store.

 

Using these retail KPIs for merchandise planning

Buying decisions are where margin is made or lost. Over-buy and the season ends on heavy markdown; under-buy and sales walk out. Four store-level signals help planners buy closer to demand, and they match what TORK reports under merchandise planning:

• Footfall trends. A sustained drop may mean customers dislike the range. Comparing stores also guides stock rotation.
• The age and gender mix of real visitors should shape the assortment.
• Dwell time against conversion. Long dwell with no purchase suggests unmet intent worth investigating.
• Loyal customers who visit but do not buy. Drill into what they were looking for and what is missing.

Less overstock, less waste

Unsold stock costs money, and in fashion it also carries an environmental cost because overproduction ends up discounted, stored or discarded. Store-level data on what customers visit and buy helps planners order closer to real demand, which means fewer markdowns and fewer unsold garments. Personalised suggestions that match a customer’s style can also reduce purchases that are returned.

Forecasting with history, not guesswork

Reliable forecasts combine past sales and trends with current supply-chain conditions. Dashboards that store historical footfall and customer data enable planners to test scenarios before placing orders.

From store KPIs to omnichannel: closing the in-store data gap

Shoppers now move between stores, websites, mobile apps and social channels, sometimes within a single purchase. This blend is often called phygital retail. Retailers have invested heavily online, where analytics is routine, while the store often remains the least measured channel.

Why omnichannel often fell short

Omnichannel promised a smooth journey across channels. In practice the channels often competed. Store regulars missed personal service online, and online shoppers noticed price differences against the store. Retailers fell short on the two things customers notice most: fair pricing and personalised service.

What unified commerce changes

Unified commerce puts stores, e-commerce and social channels on one platform that handles inventory, pricing, orders and customer data together. The aim is a single, 360-degree view of each customer, so a sales associate in store can use what the brand already knows. That view is only as good as the data going into it, and the store is where most of it is missing.

Every web metric has a store equivalent (retail KPI)

Online signal	In-store equivalent
Page views and sessions	Footfall
Time on page	Dwell time
Bounce rate	Visitors who leave without buying
Conversion funnel	Invoices ÷ footfall
Returning visitors	Repeat and loyal-customer visits
Click heatmap	In-store heatmap of busy zones

 

This is what in-store customer analytics adds. Video analytics turns store cameras into the physical-world version of web analytics, so the store can be managed with the same discipline as the website.

Putting it to work

• Run the store funnel like a web funnel. Track footfall, dwell and conversion by hour, zone and store, then fix the biggest drop-off first.
• Personalise in store, with consent. A loyalty member who has opted in can be recognised on entry, so staff suggest items that suit their purchase history.
• Align offers across channels. Compare which promotions draw visitors in store with those that work online, and keep pricing consistent.
• Plan stock and staff from real demand. Customer mix and busy periods inform assortment and rosters. See 5 ways to increase retail store productivity.

 

Handle customer data with care

Joining store and online data raises the stakes on privacy. Tell customers what you collect and why, use consent for any identification feature, collect only what you need and set retention limits. India’s Digital Personal Data Protection Act, 2023 treats facial images as personal data, so check your design with legal counsel. Our practical guide to retail video analytics covers responsible deployment in more detail.

What the data can deliver

One Softweb retail client reported a 35% increase in in-store footfall alongside an 18% increase in sales after adopting TORK for Retail. Read more about our approach to retail customer analytics with AI.

Where to start

Pick one store, two or three KPIs from the cheat sheet and one change to test. Measure before and after, then extend what works to more stores and channels. To discuss your own KPIs or a pilot, get in touch with our team.