In this episode of Retaili$tic, John Harmon, Managing Director of Technology Research, discusses the transformative role of AI (artificial intelligence) and other technologies in retail real estate management. Diving into the future of the mall, the conversation covers video analytics, RF (radio-frequency) analytics and GenAI (generative AI), highlighting their impact on leasing, operations and building management. Listen now to understand the importance of data-driven decision-making and uncover the potential for significant ROI (return on investment) for landlords through the implementation of smart technologies.
Takeaways
Chapters
00:00 This Week in Research: New Reports and Data
02:11 The Impact of Video Analytics on Malls
07:08 Understanding RF Analytics and Its Benefits
12:26 Generative AI: Transforming Leasing and Operations
18:35 Smart Building Management with AI and IoT
23:20 Conclusion: The Future of Data-Driven Malls
Read more on this topic with the full report from Coresight Research: RetailTech: Three Technologies Landlords Can Use to Take Malls to the Next Level
Welcome to Retaili$tic, the official podcast of Coresight Research for May 20th, 2025. This week, we are talking with Managing Director of Technology Research, John Harmon, about AI tools for real estate management. At first, Georgina Smith, our Head of Editorial, is here from the London office to highlight some of the research publishing this week on coresight.com. Hi Georgina.
Hi, Philip. This week, new research to look out for includes our insights on the key technologies that are transforming shopping malls into data powered environments that are more efficient, intelligent and profitable, which John will dive into in this episode of Retaili$tic. We will also publish our much anticipated report on artificial intelligence in digital commerce. From personalized messages that grab attention in seconds to virtual stores built in hours, not weeks, AI is completely transforming how we shop.
We will dive into how generative AI is redefining every step of the customer journey, from awareness and consideration to purchase and retention. How are retailers keeping up with shrinking attention spans? And what does it take to create seamless, personalized experiences, both online and in store? Stay tuned to Coresight.com to find out.
Other new reports will cover department store shopping trends, leveraging proprietary US consumer survey findings, as well as our take on recent retail industry news from management changes at Coles and OpenAI to the acquisition of Footlocker by Dick Sporting Goods. We will also update our earnings season coverage this week with insights and management commentary on the recent performance of Alibaba, Burberry, Under Armour, Walmart and more, as well as details on these companies' forward-looking guidance.
Finally, we continue to keep our finger on the pulse of tariffs. Keep updated with key developments and consumer sentiment as it relates to tariffs in our dedicated timeline infographic which we update regularly, and look out for our new reports to stay ahead of the curve. Thank you.
Thanks Georgina. We are delighted to have John back with us, our Managing Director of Technology Research. Thanks for joining us, John.
Mall operators are looking to boost performance and technology is playing a big role in that. Let's dive into three key tech areas making waves. Video analytics, RF analytics, and generative AI for leasing and operations. We'll also touch on how AI and Internet of Things, IoT, can make building management smarter. First up, video analytics. Malls have used cameras for security for years, but now those cameras are getting smart. How can video analytics help mall landlords.
Well, I mean, that's a great question. Video analytics mean using AI powered cameras to do more than just record footage. They can analyze what's happening in the mall in real time. And at our thesis of cameras powered by AI are just getting smarter and smarter. Computing power is the cost is decreasing. It makes cameras just be able to do more. And video as a picture is worth a thousand words. Video is really advancing at a rapid rate to analyze what's going on and interpret a customer behavior. For the landlord, this is huge. Cameras can count how many people enter a store or a mall, where they go, how long they dwell in certain spots, and even identify bottlenecks to customer traffic. And this helps paint a picture for landlords. For example, you might discover one wing of the mall gets much less traffic.
In one case, a mall found that a newly opened wing was being bypassed by visitors. Mall management responded by adding better directional signage and hosting special events in the wing and foot traffic increased by 35% within a month. That's really a dramatic improvement just from understanding traffic data. Video analysts can show shoppers are lining up to get into a store or a food court. So mall staff can react, maybe open another register or deploy crowd control. It's about using the camera to actively manage the mall practically in real time.
Cameras can also identify threats to customers or stores or mall visitors, or in the case of spillage and impediment in the hall, and really also determine customers' emotions, which can help store operators and mall operators determine how they're reacting to certain products or displays.
Yeah, that's interesting. An interesting insight. You know, the camera says, hey, this corridor is underused. And then they take steps bump traffic 35%. That's a tangible result. More foot traffic in a less used area means happier tenants in that part of the mall and potentially higher sales.
Certainly understanding traffic patterns better makes for happier landlords because landlords base the rents that they charge based on traffic. If they have a good handle on it, then they can price their space more appropriately.
Any other, and you mentioned some of the safety things beyond what you mentioned already, any other real world results from video analytics come to mind?
But we have a couple other examples. In one example, a mall used analytics and found that they saw a big spike in customers at lunchtime, but then a sharp drop off in traffic after 2 p.m. That meant that customers are coming for lunch and then leaving the mall. So the mall collaborated with its food court and retailers to run mid-afternoon promotions, like stay-in-shop deals after lunch. And the result was longer visit durations and a 20 % increase in sales during that afternoon period.
So these technologies can really affect sales for stores and landlords. Beyond just revenue, video analytics help with safety and security as well, identifying overcrowding, suspicious behavior such as loitering. But from a pure performance standpoint, it gives landlords a lot of data to optimize the mall layout and the tenant mix. If zone A in the mall has doubled the foot traffic of zone B, it could justify higher rent in zone A or the landlord could decide to reposition stores to even out the traffic flow.
And mall operators are increasingly sharing these insights with their tenants. A store owner might learn, for example, that most people pass their store from between 5 and 6 p.m., which could inform staffing or window displays. Overall, it really turns the mall from a collection of stores into a data-driven ecosystem. And this 20 % revenue increase that was seen due to advanced analytics is really hard for a landlord to ignore.
Yeah, definitely 20 % revenue bump just by leveraging data. That's a pretty good ROI. That leads into the second tech, RF analytics. By RF, we mean radio frequency methods, things like Bluetooth, anything that's using the radio frequencies, essentially using signals from shoppers' mobile devices. So how does that work and how is that different from video analytics?
Right, so consumers emit a lot of radio waves. If we look at our smartphone, it's got three radios in it. It has a cellular phone, the cellular radio, the Bluetooth radio, and a Wi-Fi radio. And all of these radios transmit information about the customer's cell phone.
So in a sense, the mall's Wi-Fi network doubles as a sensor network. If 100 people walk by a mall, but 50 go inside, Wi-Fi tracking could capture that conversion funnel from passerby to visitor. If the mall installs the correct devices, they can pick up these signals that identify our phones. Each of our phones has a unique MAC address, media access control. It's a unique address used in networking, but it doesn't identify our personal information. So malls can identify us going through the mall or going through the store. It can map out our journey with multiple access points. You can follow a customer throughout their whole journey throughout the mall, which may not be possible with cameras.
Because cameras, if something's blocking the way, camera goes out, they're not placed in the right position, it could lose our journey. So it's a really big advantage of RF analytics and measuring our whole journey. Cameras also face an issue with customer privacy. The camera sees your face unless the software blocks it out. And this has become an issue in some geographies, some stores, as we know are collecting biometric information using facial recognition software, which has led to some lawsuits in some areas. But the RF analytics just know you got a phone and where it goes and how long you stay. It doesn't know who you are. And malls love this kind of data. It helps map out dwell times, how long people stay in a given location, what their journey is in the store, but even conversion rates.
These RF detectors can have a accuracy of one meter. It can tell if a consumer stood in front of a display, a retail media device, and if the consumer subsequently passed by and spent time at the cashier. So these devices spin off a lot of data, don't need line of sight like a camera does. And as long as your phone's got a radio, the system can log it. They can also track other devices we've got, like fitness wearable devices or even our wireless headphones.
But I have to ask about accuracy and privacy. Not everyone connects to public Wi-Fi by phones. So the phones try to protect privacy now, right? They have some kind of privacy security on them.
Increasingly, our phones protect our privacy. For example, on iPhones, the phone gives you the option on certain apps to request that your information not be tracked. And there similar functions on Android devices, it turns out. About 10 % of iPhone and Android owners are sharing their privacy information. So it really limits the ability of some devices to collect information. But yet, all of our devices emit radio waves and do have this identifying information but it's not personally identifying information. So it can be collected. As soon as you leave the store or the mall, in about six seconds, it's gone. And the retailer or mall owner owns the data. They have the data, but they don't have your personal information. They'll never identify you ever again.
So that combination sounds powerful, almost like what online retailers have, but in the physical world, it reminds me of how e-commerce tracks page views and conversion funnels. In fact, one article likened these sensors to recreating web analytic KPIs in stores, like page visits, time on site, exit paths, et cetera, for mall owners. That means data-driven decision-making rather than gut feel.
Exactly. We're essentially instrumenting malls to gather data like a website would, and mall owners are using this data in strategic ways. We mentioned before addressing the tenant mix. Let me talk about that a bit more. Suppose the analytics show that an area around a luxury fashion tenant doesn't get traffic in the mornings, but it booms in the late afternoon, whereas the food court is opposite to the store. The landlord might use that info in selecting and placing new tenants.
Maybe introducing a morning attracting use like a coffee shop near the luxury wing to balance out the traffic flows. In fact, landlords now use traffic data to set and justify rent levels, which is the data is collected by video or RF analytics. If one mall corridor sees thousands more visitors than another, the space is able to collect more rent than than another corridor.
Conversely, landlords can improve weaker areas by adjusting what the store configuration is. That's called tenant mix optimization. And in a very analytical approach, the data can even feed into marketing. Say the mall's app can send a coupon to the shopper's phone when they're in a low traffic zone, which might nudge them to explore a quieter wing. This kind of real-time engagement powered by location data can lift overall visit quality and tenant sales. And we're looking at the future where a mall is running as data-driven as an online platform. In addition, with the boom in retail media networks, the mall, the cameras and other devices can measure how much time customers spend in front of certain screens and other retail media devices.
Cool.
Let's pivot to the hot topic of the year, AI, specifically generative AI. We've talked about measuring what's happening in the mall, but what about behind the scenes like leasing, asset management, even automating some office work? You're the tech research guru at CoreSight. How do you see generative AI helping mall landlords and operators?
Well, I think generative AI was the topic of the year last year and this year and probably for a couple more years to come. Large language models like chat GPT, GPT-4 can be a real game changer in the back office of real estate. Let's think about the leasing process. A mall's leasing manager deals with dozens or hundreds of leases, and they're lengthy and complex. Generative AI can quickly summarize lease documents, extract key terms, or even the highest risks of opportunities across them. One thing that happens is a landlord may think, wow, we offered these terms to a tenant 10 years ago, these specific leasing terms, what were they? And due to the facility of large language models to interpret unstructured data like lease documents, PDFs, they can quickly comb through these documents to find that information or to find which leases are about to expire with the renewal clause in seconds.
AI and generative AI democratizes the use of AI and the use of data. It enables everyone to search through large amounts of data and find the insights and relationships that they're looking for. And this really speeds up decision-making, identifying which tenants you may want to start renewal talks with well in advance.
On the flip side, when pitching to new tenants, generative AI can help create personalized marketing materials. A landlord could input data about a vacant space, its size, foot traffic stats, the demographics of the mall's visitors, and asked AI to draft a compelling lease for, a specific retailer. It could also produce a tailored brochure, even a mock press release about the brand opening in the mall, highlighting how the mall's data aligns with the brand's target customers. It can really automate the first draft of a lease agreement or a marketing document, which the leasing team can then quickly refine. It offers just a much faster turnaround in a very data rich story to address prospective tenants.
It's also good at scenario planning. As I mentioned, the ability to talk to your data, to ask questions of your data without being a data scientist. Model owners can ask what if questions. What if we convert 10 % of our retail space to entertainment? What if we adjust store X's brand and bring in a new anchor store? Genitive AI is good at analyzing data that's already there. Machine learning is better at making projections and predictions. But it really can sift through the data such as sales data, foot traffic, industry data to give humans a picture of the outcomes. It's like having your own analyst on call 24 seven to brainstorm with.
Yeah, it sounds like having a really fast researcher or assistant that could synthesize information. I imagine asset managers would appreciate that. They could use AI to flag anomalies in financial reports, for instance, or help with capital planning. I've even heard of AI being used to scan through maintenance requests and suggest optimizations. What about actual operations or tenant support?
Well, again, AI excels at finding relationships among data of all kinds. One application of generative AI is chat bots for customer and tenant service. Some also have hundreds of tenants. Generative AI can collect all the tenant information and ask common questions from store managers, like when is the next marketing event? How do I request extra cleaning after an event?
Many enterprises have created corporate knowledge bases where they take in all of their procedures and rules and guidelines and information, which can be searched quickly by users. Rather than calling the management office, a tenant can get answers from this kind of a chatbot trained on the mall's procedures and policies quickly. But also for consumers, malls can have AI chatbots that help them find store information, mall hours, even recommendations for styles or what stores to visit.
And they can answer a lot of the regularly occurring questions, freeing up call centers to focus on customer service. On the operations side, AI can also help maintenance teams by analyzing maintenance logs. can predict which escalators or air conditioning units might need service soon. So the mall owners can provide maintenance proactively rather than reactively.
That's classical AI, but generative AI adds a twist to it. Its ability with language can produce a report every morning in natural language summarizing all open maintenance issues and their priorities. Rather than a manager having to sift through spreadsheets, they can get a narrative like, good morning, five escalators have issues. The one near Store X had three incidents this week, suggesting a high priority to check out that escalator. And it makes the consumption and analysis of data more user-friendly. But yet, generative AI still isn't a magic bullet. It excels at producing content, finding patterns in data.
But there are limitations to generative AI, such as hallucination, producing incorrect results. So AI still at this point does need human oversight, especially in things that are sensitive, such as lease terms or financial projections. But it really is an accelerator or a co-pilot for all of us. In one report, generative AI was mentioned as streamlining leasing and asset management tasks, as I mentioned before, from drafting documents to analyzing risks which frees up humans to do more interesting higher level work.
We're seeing huge investment in the AI space and real estate can clearly be a beneficiary of this investment. The numbers for AI are staggering. It's clear it's not just a buzzword, but it must be driving ROI for the amount of money that's getting poured into it. if a leasing team can save weeks of work on analysis or just get materials out faster, that can translate into real dollars, not to mention better decisions on tenant mix, as you noted earlier. Now let's talk about the mall building itself. These are massive properties, energy bills, cleaning, security, all that. How are AI and IoT improving building management for landlords?
Well, this has to do with smart building technology, which is another exciting area. Malls are like mini cities. They consume a lot of energy and have a ton of equipment to maintain. Combining AI and IoT, the Internet of Things, can turn a mall into a smart self-repairing system. For example, having IoT sensors on climate control at the mall, measuring temperature, humidity, occupancy in different zones.
These algorithms can use the data to adjust heating, ventilation, and air conditioning all in real time, which improves comfort for customers and minimizes wasted energy. For example, if the food court is packed at noon, the system might boost the air conditioning there, but dial it down in empty corridors to save energy. And these micro adjustments really add up. Mall owners can see the ROI of this kind of control immediately as their energy bills drop, which really justifies the investment.
This is doing things like the Google Nest thermostat for consumers, which uses AI to learn your habits and your schedule and automatically saves energy costs by adjusting temperatures. In a million square foot mall, this can translate to enormous savings. For example, Google applied its own eye to its data centers cooling systems and cut energy usage by 40%. A data center is not a mall, but the concept is similar.
They're both big buildings, lots of HVAC high voltage air conditioning equipment. And if AI can gain 40 % efficiency improvements in a setting like that, they're also potential huge savings for a mall. Microsoft was a leader in this, again, with IoT and machine learning, and reported an 80 % 90 % reduction in faults, which are HVAC equipment failures or false alarms, by using AI to predict and filter maintenance issues.
AI can not only identify and triage events when they occur, but again, going through the data can predict when a certain air conditioning unit is likely to go down or need maintenance, which enables the landlord to address it before it's a problem. And fewer faults mean less downtime or repair costs.
And from all, it can keep their escalators, elevators, lightning, air conditioning more reliably up and running with fewer outages, which leads to better customer satisfaction.
Those are impressive statistics. A 40% energy reduction at a Google server farm is probably a huge number, but at a mall, it's still 40%. If you could reduce your monthly energy bill by 40%, that's a great return on deploying these technologies, then let alone the 80 or 90 % fewer maintenance faults. It highlights that investing in these technologies isn't just tech for tech's sake. There's real ROI and cost savings. I imagine it also plays into sustainability, is a growing concern for large properties.
This really paints a picture of the mall of the future, where every decision from leasing to temperature control is aided by data and AI. It's crazy. Before we wrap up, let's synthesize this for our listeners, many of whom are mall operators or landlords. What's the big takeaway about these technologies?
Data and tech aren't replacing the human touch or strategic thinking, but they're amplifying it. It's sort of akin to having superpowers for insight and efficiency as a landlord. It's interesting that the data that these malls are collecting will probably one day be the source data for programming the matrix when everybody plugs in their Neuralink and their VR glasses or whatever they connect. And off they go. They'll be walking through malls that are created in the digital space based on the data that we're collecting today.
So this has been a really enlightening discussion, John. Thanks for breaking down these concepts with concrete examples and analysis. Mall operators have a lot to chew on here, from counting footfalls to drafting AI-powered lease proposals. I appreciate the practical perspective you brought grounded in research and real-world results. We've got a report that's coming out if people want more details, correct?
Another stellar visit from John Harmon, our Managing Director of Tech Research. Thanks, John. I can't wait till we can get you back on the show.
Have a wonderful day and we'll see you next week.