When AI Handles the Errand, the Centre Must Own the Routine

AI shopping agents will make a large part of retail feel less like shopping and more like background admin.

That matters for shopping centres.

The early agentic-commerce story is usually told as a technology story: AI assistants searching, comparing, checking reviews, finding stock, assembling baskets and completing purchases for the customer. Bain has already framed the risk for retailers as disintermediation at the top of the funnel, with third-party agents influencing discovery before a customer reaches a brand or retailer. JLL has made a similar property-market point: AI-mediated shopping will affect which retail assets can still create demand and which become easier to bypass.

For centre owners and leasing teams, the useful read is more practical. AI changes the value of a physical visit.

Some customer missions are mostly friction: find the acceptable product, get the lowest sensible price, arrange delivery, reorder when supplies run low. These are the missions most exposed to automation. Printer cartridges, pantry staples, pet food, household consumables, basic electronics accessories and repeat purchases can all drift into the background when an agent can handle search and replenishment.

The centre then needs to be clear about the missions it owns.

The exposed mission is the easy errand

The weakest physical-store logic sits where the product is simple, the choice is easily compared, the customer already knows what they want and delivery solves the timing problem.

That kind of trip has little emotional weight. The customer may value the outcome while feeling no attachment to the process. If AI removes the process, the lost visit is small.

Commodity categories still belong in centres when they anchor local routines. Grocery, pharmacy, discount variety and everyday goods are strongest when they are tied to immediacy, trust, fresh choice, prescriptions, advice, top-up behaviour or another trip purpose.

The risk is different: weak categories need a stronger reason to be physically useful. Price and availability alone are thin defences when a digital agent can compare both in seconds.

The durable mission needs presence

The stronger centre missions involve the customer’s body, time, trust or social routine.

A customer still needs a place for a haircut, GP visit, physio appointment, eye test, beauty treatment, pilates class, restaurant meal, child activity, phone repair, product fitting, gift browse, return, pickup, coffee meeting or quick dinner solution.

These missions carry a physical requirement. They also create adjacent behaviour. A gym visit can become coffee. A GP visit can become pharmacy and groceries. Childcare can become breakfast, fresh food and an afternoon top-up. Dining can extend the daypart. A beauty service can support product trial and replenishment.

That is where tenant mix becomes more defensible.

Retail Insider’s coverage of JLL Canada’s 2025 retail report noted that dining, entertainment, fitness and wellness services accounted for 51% of new-store announcements in the first quarter of 2025, with dining alone contributing more than 40%. Deloitte’s store-reinvention work points in the same direction: stores are evolving into hubs for cross-channel convenience, product engagement, brand connection and human experience.

The signal is consistent. Physical retail keeps its power when the store does a job the screen cannot complete on its own.

Footfall needs a purpose attached to it

A busy centre can still be exposed if many visits are attached to low-differentiation errands.

The better diagnostic is mission quality:

• Which visits are urgent or immediate?

• Which visits involve advice, service, fitting or care?

• Which visits create cross-shopping?

• Which tenants generate weekly repeat behaviour?

• Which dayparts are supported by food, wellness, medical, childcare or entertainment?

• Which parts of the centre reduce friction enough to become habitual?

This is where the old traffic-first conversation starts to look blunt. A visit count tells the owner that people arrived. Mission data explains whether the trip is resilient, replaceable, social, routine-forming or vulnerable to automation.

Coniq’s recent writing on shopping-centre loyalty makes a related point: loyalty is moving from a marketing programme toward customer intelligence infrastructure. The valuable question is how engagement changes behaviour — visit frequency, cross-store discovery, spending and tenant performance.

For centre strategy, that means repeat behaviour matters more than broad claims about experience. A yoga class every Tuesday, a trusted pharmacist, a strong food operator, an easy click-and-collect pattern and a reliable children’s activity can be more commercially useful than a one-off activation.

A practical tenant-mix test

A simple test for each category is: can the customer complete the mission through an AI agent without visiting a place?

If the answer is yes, the physical store needs one of five defences:

1. Immediate resolution — the customer needs the item, repair, prescription, return or service today.

2. Human judgement — the purchase needs advice, diagnosis, fitting, reassurance or trust.

3. Sensory proof — the customer needs to touch, taste, trial or compare in person.

4. Bundled routine — the trip is linked to grocery, childcare, gym, medical, dining, work or school movement.

5. Social value — the visit is part of a family, friend, community or leisure routine.

If the store cannot claim one of those roles, its physical logic is weaker.

This test is especially useful for neighbourhood and sub-regional centres. Their advantage is rarely spectacle. It is routine density: several practical missions solved in one easy local trip.

Grocery plus pharmacy. Childcare plus coffee. Medical plus fresh food. Fitness plus lunch. Click-and-collect plus a top-up shop. Dining plus entertainment. A centre that stacks these missions well becomes harder to substitute because the customer is no longer making one isolated purchase.

They are solving part of the week.

AI also changes centre operations

AI will also help operators run better places.

ICSC has written about agentic AI moving into retail and real-estate workflows, including merchandising, marketing, inventory, procurement, customer care and even event planning. In a centre context, the value sits in sharper operating questions:

• Which tenants pull customers across categories?

• Which missions are rising or weakening?

• Which events change actual visitation rather than app engagement?

• Which dayparts need support?

• Which offers create repeat behaviour?

• Which customer groups are leaking to competing centres?

• Which parts of the physical journey add avoidable friction?

The technology is useful when it improves those decisions. A weak centre with better automation remains weak. A strong centre with better behavioural intelligence can lease, market and manage with more precision.

What Loculyze is watching

For Loculyze, agentic commerce becomes an asset-resilience question.

The centres to watch are the ones with a high share of presence-led missions: food, health, wellness, beauty, services, entertainment, repair, childcare, community use and fulfilment-linked convenience. The exposed centres are those relying heavily on categories where the customer mainly wants the transaction completed with minimum effort.

The distinction will show up in leasing demand, daypart coverage, repeat visitation, tenant productivity and the ability to defend local relevance as more buying moves into automated channels.

AI will make some shopping invisible.

Physical retail now has to be more deliberate about the parts of shopping that still deserve a place.

AI Accelerated Research by Locyra

Meet Locyra — Loculyze's AI property analyst with a retail property lens. Each week she reviews public retail-property reporting, selected industry commentary and Loculyze's internal property frameworks, then turns them into clear, practical takeaways for investors, owners, retailers and asset teams. She is built to do what good analysts do best: find the signal, explain why it matters and help the market make sharper decisions.

Source note

This article draws on public retail and property commentary including Bain’s article on agentic AI in retail, Deloitte’s note on the evolution of retail stores, ICSC’s discussion of agentic AI in retail and real estate, Retail Insider’s coverage of experience-led new-store openings, Coniq’s article on shopping-centre loyalty and commercial impact, and selected real-estate research on AI shopping agents and property demand. Locyra is reviewing public retail-property information and Loculyze's internal property frameworks. It does not use confidential client data.