Something quiet is happening at your front door. The shopper walking in today did not stumble across your store. Odds are they asked an AI assistant what to buy, compared three options before leaving the house, and showed up already knowing what they want. The browsing happened on a screen. The buying decision is happening in your aisle.
That shift changes what you need to measure. As AI shopping agents take over discovery and comparison, the store visit becomes the deliberate, high-intent final step. Which means the in-store experience metrics that prove your store turns those intentional visits into sales matter more now than they ever have. Here is what is changing, and the exact numbers to watch.
AI is deciding what to buy. Stores are where it happens.
Agentic commerce is no longer a forecast. A report added on August 17, 2026 by Research and Markets found that AI usage is now concentrated at the start of the buying journey, reaching roughly 62% for product comparison versus about 23% at checkout and 19% for post-purchase tasks. In plain terms: shoppers are letting AI research and shortlist, then finishing the decision themselves.
The traffic patterns back it up. That same analysis reported that generative AI referral traffic to U.S. retail websites climbed by as much as 4,700% year over year in July 2025, while traditional organic search traffic declined. Discovery is moving into AI interfaces fast.
Here is the part that surprises people: this is good news for physical stores. A joint report from ICSC and McKinsey, "Shopping in the Age of AI," found that store visits are becoming less frequent but more intentional as shoppers research, compare, and refine choices before they leave home. The store is increasingly the final step in the purchase, not the place where browsing starts. McKinsey estimates agentic commerce could drive up to $1 trillion in U.S. retail revenue by 2030, and that many of those decisions still get finished in person.
Physical stores are getting more important, not less
If you assumed AI would hollow out brick-and-mortar, the data says the opposite. A 2026 Colliers report, summarized by Retail Brew, laid out the case clearly:
- 85.1% of U.S. retail sales still flow through physical stores.
- 71% of retailers say they will expand their brick-and-mortar footprint this year.
- 1 in 4 online orders is now fulfilled by a physical store.
- The share of consumers using AI agents is expected to jump from 19% today to 46% by the end of 2026.
"One of the biggest misconceptions around AI is that it reduces the importance of physical retail," said Nicole Larson of Colliers in the report. "In reality, retailers are expanding footprints and redesigning stores as physical locations are becoming even more critical to fulfillment, customer engagement, and operational efficiency."
The ICSC and McKinsey research adds a useful frame: stores now have to deliver one of two things clearly, either convenience (speed, easy navigation, in-stock reliability, cited by 37% of consumers as a top driver) or discovery (experience, service, and the kind of in-person moments an algorithm cannot replicate). Either way, the store has to perform. And you cannot manage what you cannot see.
The new problem: high intent is invisible without measurement
When a shopper arrives pre-sold, your store has one job: do not lose the sale. That sounds obvious, but most stores have no idea whether they are converting these intentional visits or quietly leaking them.
Think about the trap. If AI-driven shoppers arrive ready to buy, your conversion rate should be climbing. But a strong intent environment can hide a weak store. Sales tick up because more of the people who come in already decided to purchase, so an understaffed floor, a confusing layout, or a slow checkout gets masked by the tailwind. You feel fine while you leave money on the table.
Foot traffic alone will not tell you. Neither will your point-of-sale system on its own. Your POS knows who bought. It knows nothing about the intentional shopper who walked in, could not find help, and walked out to buy from a competitor the AI also recommended. The gap between "people who came in ready" and "people who actually bought" is exactly where the AI-era opportunity lives, and it is invisible unless you measure it.
The in-store experience metrics that matter in the AI era
You do not need a dozen dashboards. In an environment where visits are fewer and more deliberate, a handful of store-level numbers tell you almost everything. Here are the ones to prioritize.
1. Foot traffic (the denominator for everything)
Counting how many people actually enter is the foundation. Without it, every other metric is a guess. Fewer, more intentional visits make an accurate count more valuable, not less, because each visit now carries more purchase intent and more revenue weight. Missing 15% of your door count means misjudging every decision built on top of it.
2. Conversion rate (the AI-era scoreboard)
Conversion rate, the share of visitors who buy, is the number that proves whether high-intent traffic is turning into sales. It is the closest thing physical retail has to an efficiency score. Yet by Dor's own data, 91% of clothing retailers do not know their conversion rate. In a world where shoppers arrive ready to purchase, flying blind on conversion is the most expensive habit in retail. A 1% lift in conversion can translate into roughly 10% more revenue, with no additional marketing spend.
3. Traffic-to-transaction by hour
Intent does not arrive evenly across the day. AI-assisted shoppers plan their trips, which can reshape when your store gets busy. Tracking conversion by hour, not just as a daily average, shows you the windows where visitors show up ready but staffing or service is thin. Retailers are mis-staffed roughly 86% of the time according to Dor's research, and that misalignment is pure lost revenue when the shopper walking in already wants to buy.
4. Capture rate
Capture rate is the percentage of nearby or passing shoppers who actually enter your store. In the AI era, discovery may happen on a screen, but the physical decision to walk in is still shaped by your windows, signage, and entrance. If AI is sending qualified shoppers toward your category, capture rate tells you whether they are choosing your door or the one next to it.
5. Dwell and repeat visits
How long people stay, and how often they come back, signals whether your store is winning on discovery and experience, the differentiator AI cannot copy. Rising dwell alongside steady conversion usually means your in-person experience is doing real work. Falling dwell is an early warning that something at the shelf or the service desk needs attention.
How to respond: turn AI-era intent into sales
Knowing the metrics is step one. Here is how to act on them.
- Measure conversion before you spend another marketing dollar. If shoppers are arriving pre-qualified by AI, your fastest growth lever is closing more of the intent already walking in. Lifting conversion a few points is cheaper and more durable than buying more traffic.
- Staff to your real, current peaks. Pull the last several weeks of hourly traffic and put your strongest associates on the floor when intentional shoppers actually arrive. Do not schedule to last year's rhythm.
- Compete on the two things AI hands you. The ICSC and McKinsey framework is a gift: pick convenience or discovery and be excellent at it. If you are a convenience play, obsess over in-stock reliability and speed to purchase. If you are a discovery play, invest in service and experience, then watch dwell and conversion to prove it is working.
- Treat the first ten feet as a conversion tool. When visits are fewer and higher-intent, the entrance does more work than ever. Refresh displays on your busiest days and make it effortless for a pre-sold shopper to find what they came for.
- Benchmark location by location. AI-era intent will not lift every store equally. Comparing conversion and capture across sites shows you which locations are converting the new demand and which are letting it slip.
Where people-counting data closes the gap
Every move above depends on one capability: accurately counting the people who enter your store, in real time, and pairing that count with your sales data. Market reports tell you the category is hot. Only your own door count tells you whether the intentional shopper made it inside and bought.
This is the gap Dor was built to close. Dor's thermal sensor counts everyone who enters automatically, with no clicker, no cameras, and no IT project. It is battery-powered, installs in minutes with peel-and-stick mounting, and starts delivering data the same day. Because it uses 100% anonymous thermal sensing rather than cameras, there is no personal data collected and nothing to slow down deployment.
From there, Dor turns your door count into decisions. It connects to your point-of-sale system to reveal your conversion rate by hour, day, and location, shows your real traffic patterns for smarter staffing and capture decisions, and lets you compare stores side by side. So when the headlines say AI is sending more intentional shoppers to stores, you will not have to wonder whether you are winning them. You will know. For the full foundation, our guide to foot traffic analytics walks through the basics.
The bottom line
AI is not emptying stores. It is filling them with shoppers who already know what they want, then leaving the close to you. That is the best kind of traffic a retailer can get, and the easiest kind to waste. The stores that win the AI era will not be the ones with the flashiest technology. They will be the ones that can see their own front door: how many intentional shoppers walk in, how many buy, and exactly where the intent is slipping away.
If you want to know whether the AI-era shopper is converting in your store, book a Dor demo or see pricing and start counting what matters.
Sources
- ICSC and McKinsey & Company, "Shopping in the Age of AI: Redefining Stores for a New Era" (April 27, 2026): ICSC, Axios
- Retail Brew, "AI is making physical stores more important, not less" (May 19, 2026), summarizing a 2026 Colliers report: retailbrew.com
- Research and Markets via GlobeNewswire, "AI Shopping Agents and Agentic Commerce 2026: Adoption Trends and Execution Limits" (August 17, 2026): globenewswire.com
- Dor (Dor Technologies), retail people counting and conversion statistics: getdor.com/industries/retail