If ChatGPT Can't Find Your Community, Renters Won't Either

Hello everyone and welcome to today's webinar. If ChatGPT can't find your community, renters won't either. We're excited that you're joining us for today's 1-hour session. My name is Michael Kger. I'll be moderating the event. And today we're exploring the latest disruption to the multifamily industry, the growing reliance of apartment hunters on large language models or LLMs.

As you all know, change is the only constant in real estate. And as the supply of apartments outpaces demand in many regions, vacancies are creeping up while rents are going down. To keep occupancy rates and revenue high, communities must first and foremost be visible to their audience. Relying on what's worked in the past, internet listing services and SEO is just table stakes. The nature of how people look for information online is changing. AI powered search is used by more than half of consumers today with the majority citing it as their preferred digital source.

LLM such as ChatGPT and Claude deliver explicit answers in lie of the ranked results of conventional search engines. And for owners seeking to expand their market reach, LLM offer huge opportunity if their communities are visible on AI search. In this webinar, two experts will discuss how communities can assess and improve their AI visibility. We are thrilled to be joined by our speakers. Nicole Jones is chief marketing officer at Veris Residential based in Jersey City and Austin Lo is founder and CEO of Peek based in New York. But before we get started, just a few housekeeping items.

The slides will advance automatically throughout the presentation. And to enlarge the slides, click the enlarge slides button located top right corner of your presentation window. If you need any technical assistance, just click on the help widget located on the bottom left corner of your console. And we encourage you to submit questions at any time throughout the presentation using the Q&A widget at the bottom of your screen. To kick things off, I'd like each of our speakers to provide a quick introduction of themselves and their work. Nicole, would you please start?

Hi everyone. I am Nicole Jones. I am the CEO at Veris Residential. We are a publicly traded multifamily REIT based in Jersey City. We own and operate primarily class A buildings in New Jersey and Boston. I oversee marketing and communications for the company including all corporate and PR.

Um I've spent the last 20 plus years in the real estate um industry and specifically I've been focused on multifamily marketing. It's nice to have Thank you Nicole and Austin. Hi uh my name is Austin. I'm founder and CEO of Peek. We are an AI powered leasing and marketing platform for the multifamily real estate industry. We started this journey in trying to optimize leasing and making it easier for renters to find a place to live with unit level 3D tours.

Um and as the rental search process has gravitated and started uh moving towards AI search uh we recently launched peak discover uh which is our hat discoverability product uh focused on helping and operators determine how visible they are eventually to optimize uh their appearance in front of vendors uh in these AI engines. Thank you Austin. Now let's jump into our discussion. Nearly every industry has been affected by AI powered technology which has also affected the behavior and online search habits of consumers. According to a 2025 McKinsey article, half of consumers of all ages seek out AI powered search engines and apps such as ChatGPT, Gemini, and Claude. The article estimates that brands are unprepared for this chose those brands that are unprepared for this change may see a decline in site traffic from conventional search channels of between 20 and 50%.

That's pretty dramatic. How have our changing search habits affected the multifamily industry? Well, let's start with Nicole. What do you say? So, when I started in marketing 20 plus years ago, we were placing ads in newspapers. uh things that our renters know today such as ILS's like apartment guide, apartments.com, those were printed books that people would pick up at the grocery store and that's how they, you know, would search was sifting through sort of some sort of printed material.

That has obviously changed to um Google really dominating the search um through PPC where companies like mine, we spend quite a bit of money uh both to allow our um our properties to show up organically which takes a different kind of work which is SEO um to actually paying Google to show renters our results which can get quite expensive when you are in a very crowded market um and when you are you know advertising things luxury apartments. That's very expensive. Um, and what we've seen over the last few years, but specifically the last few months is a lot of renters are coming through the LLMs. Um, which is obviously the the thing that we're here to discuss today and it continues to grow. Um, it continues to accelerate. Um, I mean, from the time we started prepping for the session and this discussion, the numbers have grown even just in in those few weeks.

Um, so we know that our, you know, our renters in particular are very savvy. They're sort of ahead of the technology curve, if you will. Um, I wouldn't say it has anything necessarily to do with demographics, but there are people who are probably using AI in the professional um, careers and so they're very comfortable having conversations um, you know, with the LLMs to change how they search. So, it just it really has um accelerated the the conversations that they're having with the LLMs are really nuanced and tailored and personal. Um and we're just we're seeing renters get to us um via those conversations. Are there personal reasons why people prefer the LLM search versus the traditional searches?

I don't. It's really anecdotal what we have at this time, but it seems that they can have that personalized conversation. And um we have a lot of renters who are moving to uh I know I mentioned we're in Jersey City, but a lot of our properties are right across the river from New York City. So people may be moving to New York City for a job, but they really have no idea where they should live. And you know, previously you search apartments New York City into Google and you get, you know, just inundated with options and it doesn't help you narrow down um where you should live. Now they can have that conversation with the LLM of their choice and say, "Hey, here's where my job is going to be.

Here's, you know, I don't know if I'm ready to live in New York City, but I need to be in the city three days a week. Help me figure out what kind of, you know, what options for neighborhoods." Then they can start having I want to school for my kids. That kind of thing, right? Yeah. Yeah. And then it's, you know, I've got a dog who doesn't do well with, you know, the city or, you know, I want to take my kids to the park after after uh work.

Sure. I I think that really like it's hits the nail on the head like the like the big shift is not only in terms of like the search volumes and and it's how we got into this was that we were seeing more and more of our traffic on the 3D tours come from engines like chatb and claude but also just people are using it um in traditional search like people are looking for indexes of information which they would dive more deeply into. Um, but what we're seeing and and what you see from even like large like public data sets on how people are engaging with these large language models is that people expect answers. Remember like when ChatGPT first came out and everybody had that like there was all this kind of hollow blue over hallucinations. Oh my god, it's telling me the wrong thing. Whereas like the expectation was that okay if I converse with this large language model I'm going to get the answer going for me.

It's going to find the right answer for me. It's going to guide me to the right answer. And like as Nicole mentioned, like when it comes to finding a place to live, it means that people are actually engaging conversations about okay, I need my commute to be this or if I need XYZ qualities in the place that I live or this is what I'm looking for. And I think what this means is that the same techniques that optimized for SEO to get you placed in or placed highly in Google, it's a decent start. It'll it'll make sure that you're doing okay in terms of AI visibility, but what's required to truly outperform or do well and kind of gobble up the lion share of this traffic from these LLMs is fundamentally different because it's just a different game. So that's really interesting, Austin, but let let me step back a minute and lay out the groundwork for how renters currently find multifamily communities.

Nicole, how do people find your properties? Um, so there are there are obviously a number of ways. There are still the traditional um means of finding us, as I mentioned, ILS's. So those are your your Zillow, your apartment guide, apartments.com. Those continue to be a source. um they're still an important source um where people go to search out um usually those are people who kind of know everything they want in an apartment.

They like to be able to filter sort of the Airbnb model. Here's what I know that I want. Help me filter in and out the choices that do or don't have the things that meet my requirements. Um there's still the conventional search engines. You know, you still have to optimize for the Google search. um you still have to appear both organically and still have to do the PPC.

We we have not seen a decrease just yet in our PPC budgets. Um you know hopefully at some point that that will be a result of people kind of moving more to the to the LLMs. Um, and then obviously as we are talking about today, LLMs are starting to bring a different renter um to us who is probably much more qualified because they've had such an indepth um personal conversation with the LLM about their um their actual needs that by the time they get to us, they are pretty um set in their decision. So they're pretty far down the decision the decision pipeline which is really encouraging for us because our goal in marketing is always to drive the most qualified traffic to our sites. Thanks Nicole and Austin. What have you observed in renter search behavior?

So I I think for us we um we actually found a large aggregated data set of just general conversations with large language models. And the striking thing and to go back to my previous point about people expecting answers is that people are treating ChatGPT or Claude almost like a real estate agent like as if you were or as if they were buying a house. You're getting very specific with their preferences. It's not just your traditional kind of Google search where renters would be putting in apartments in this neighborhood and then as Nicole mentioned like going onto these platforms and then doing the filtering there like They're expecting these LMS to actually kind of do the filtering for them to narrow their preferences down to say, "Okay, my budget is this," or, "Oh, I don't want to live south of this street, or I want something that has these amenities, or I want a layout that has uh that has an island kitchen, or I don't want carpet anywhere in in my apartment." like these are the things that that renters are putting in because it has that human conversational feel. They're treating it as okay, I have this agent that is there to kind of do my bidding or to go like find what I what I truly need. which I if I if I may say I think that that is um to to Austin's point there that is one of the things like we've always been filtering for a set of features.

So pools, gyms, hardwood flooring, washer and dryer, these you know parameters that are pretty strict and that's you know how the ILS's um you know present themselves. Even on our own websites, there are fields that populate these things so that people are able to filter and it's not just sort of an endless um sea of options um because we're trying to help the renter narrow down. the LLM's people are able to ask for the nuanced thing that maybe either we didn't think would be of interest to a renter or traditionally hasn't been of interest to a renter or the thing that is um you know could be considered like specific to that person that we just simply can't filter for because they're the only person who wants that one specific thing. Well, now they can have those conversations. Yeah, I was going to say it it helps extract structured information out of the unstructured data in the listing. Exactly.

But, you know, it's I know we'll get to it in a bit, but it's it's pulling that information from somewhere, right? And people do trust the conversations that they are having um for the most part with these um with the LLMs to say like again, which building actually has both a dog park at the building, but also has a neighborhood dog park within two blocks. And by the way, I want a playground on the way because my kid and I are going to stop and play at the playground. Then the dog's going to go to its playground, but for quick, you know, out breaks for the dog, we'd like it to be a building that has a dog playground within it. So you can have those conversations whereas you wouldn't have been able to filter for that specific nuance previously. I think Yeah, go ahead.

I I think that's like that actually goes to like one of the big differences between traditional like SEO and even like how SEO tools have been oriented is that in the SEO world you're not focused on those keywords about these all these little things or these kind of longtail topics or other kind of variables in the search process because search by nature is so topheavy and renter search behavior is so topheavy and that people aren't going to be searching for an apartment in Dallas with a fitness center under dog park under $2,700 because the kind of learned consumer behavior is to go find indexes of information and then go deeper from there. Um, now when it comes to SEO, like a lot of things that do make for good SEO, making this information available does help u because SEO is also about machine readability. Um, but in terms of even just the scope of information or the types of information that make for good AI visibility versus SEO, there is a big difference in in the two. Um, and that's actually kind of where where you see also the differences in the tool. uh like SEO tools generally focus on keywords because they're only focusing on terms that people are going to go put into Google because that's what you want to measure. That's where you're getting your traffic.

Those are the keywords you're buying. Uh Nicole, I don't I don't think Maris is buying any keywords on Jersey City apartments with fitness centers under $2,700 as a long string, right? But when it comes to large language models and when it comes to AI visibility, like the way that these systems work is fundamentally different. Like you're going from search, which is like it's called lexical searches. You're matching the keywords. You're lining up the keywords and they'll say, "Okay, these keywords match.

How relevant is this domain source?" And it's also focused on the domain because you're optimizing for traffic on onto the domain. Um, however, with these AI engines, 70% of kind of the the information is actually sourced from these ILS's in our listing services. Only about 30 35% is coming from the property website. So not only are you not measuring just on the domain, you need to measure kind of you need to measure the entire visibility of the community as an idea, but because these large language models function using vector search um in which they're looking for associated ideas um having this additional information, having this content that renters wouldn't necessarily type into Google about the fitness center, about the dog park, about the finishes inside of the apartment, about the nearby points of interest or how close it is to a train station or a major highway. Like these formerly kind of longtail keywords become important topics for these large language models to actually essentially be able to find the community. Um, and if the marketing materials or the content on the website don't have any of this information, if it's not there or if it's not in a machine readable way, it's actually really really difficult to to be good at AI visibility because not only is this AI kind of engine looking for call it apartments in Jersey City, it'll actually break it up into a bunch of different questions and say, "Okay, if somebody's looking for an apartment in Jersey City, what are the things that are likely that or topics that are likely that they're going to care about?" And your visibility is not just dependent on one keyword or one small set of keywords.

It's based on a very broad array of topics. I guess you know this is the difference between search engine optimization, search engine marketing and generative engine optimization. Um Nicole, what do you see in terms of these differences? as as Austin was talking about the you know we know that the ILS has continued to be really important at this point um where the the LLMs are pulling information from um you know to get the those basic um queries but we're just finding that the content on our website the the content that we put out into the world um is uh really helping in a way that you know it wouldn't have previously helped with SEO. So, previously we had to um kind of do SEO the way that Google wanted us to. And you know, oftentimes you would sort of figure out with we we use a digital agency.

We'd figure out with our agency, okay, what what does it Google want to populate us and we would adapt our content for Google um to make sure that it was, you know, formatted the the right way or we had the right words. And sometimes that leads to websites that feel a bit off-brand, if you will. Um, that content is sort of written and and redundant and rewritten within a website, which could be really frustrating for consumers that you come there and you it it feels like there's not actually anything meaningful here. Um, so for us, what we're seeing is that the the content that feels more naturally written, that feels like, you know, written by a human for a human, um, is really helping the LLMs perform. Um, the the storytelling, the the storytelling of your building, the brand, um, what it's like to live at the building. If you can incorporate things like social um you know real authentic um pictures or reviews, the more content you have there that feels um iterative and nuanced is really helping the LLM see you.

And and for us it's something that we were already doing um before the LLMs became a thing. um we kind of fell into good success um because we were really trying to tell a brand story and a narrative about Veris residential and what it's like to live at our buildings that was really important to the brand. So while we were still serving SEO to Google um we were just trying to ensure that when somebody came to our website it felt like okay now now that I've landed here I can understand more about this place and that set us up for success with the LLMs. And are there some particular challenges in the multifamily market versus uh other people who are uh offering properties? Multif family is the most nuance. It it is it is complicated because it really is a a onetoone relationship.

It's us and that renter versus you know retail or office or or different types of real estate where you are sort of it's more of a B2B um marketing play. here we've got to appeal to every single person and their specific set of needs. So it it is certainly nuanced and that's why I think that the LLM actually present a really exciting opportunity because it's giving us this chance to tell our story in a more onetoone way and I and I think that will continue to evolve of you know like I can imagine a world one day where the websites themselves could be responsive to that particular person and what they need and sort of adjust the website to show up if the person comes through the LLM. We're obviously not there yet. I just it's the the world that I envision that okay, we know that it was the fitness stuff was really important to you. We're going to serve that up first to you versus you having to sift through the website and find it.

Um but it is like I said, it is nuanced. It is um it's a highly regulated industry on top of it. So that's another place that we have to be really careful um and really pay attention to what um is being served up um you know both what the LLMs are saying um as well as you know how we present information. Um so it is yeah it's nuanced but I think it's actually a really exciting opportunity in the multifamily space. I I think this brings up actually one interesting data point that we recently got um from looking at the visibility results um across all of our customers is that generic copy uh the generic kind of marketing copy paste that may work very very well for Google keywords because you're just optimizing for appearance of what everybody else has um actually performs relatively poorly with AI uh with AI search engines. Um for all the talk about kind of AI slop, it's almost ironic that these large language models can detect essentially human marketing slop and they can detect when something is just generic luxury apartments, high-end finishes, and great neighborhood.

like they'll pick up on on these generic descriptions and preference more unique or or more kind of contextualized information ahead of essentially this copy paste which seems to be relatively endemic in in the multifamily industry. Okay. So let me take a look again. Let's let's take a look from the websit's perspective. I mean, are we understanding how much of that traffic is coming from AI versus from traditional sources? Austin, um, so it's actually really difficult to tell.

Uh, we recently did another survey of just the results that are coming in through Claude and Chatg less than 25% of the links that are clickable. So now when you when you go type in, hey, I'm looking for XYZ, and they give you a list of results, they have those clickable links, and fewer than 25% of those actually have the correct UTM codes. There in many cases, they're actually incorrect UTM codes. Uh we found that in over 20% of the links for apartment communities, ChatGPT was misattributing it to Google My Business. Huh. So now like when when 70% of tra traffic from chat GBT is even going through the ILS's like it's incredibly hard to measure where really like the originating point of the traffic is.

And even if you're just looking at your website and say okay all right the ILS is that's its own kind of game. Even if you're just looking at your own website, because these LLMs aren't passing on the UTM codes, like your Google Analytics data is massively understating the amount of traffic that's actually coming from chat. Yeah. And and to that point, we've always had an attribute, sorry, we we've always had an attribution problem um of figuring out, you know, where people come from. We we know that they're looking at multiple things, right? Like so maybe they originally um started on social media.

They could have seen a post on Instagram and then they're like, "Oh, that looks really nice, but they know as a consumer that that's pretty polished. So, let me go look this place up on Google and they're looking at Google Maps to understand where it is." And then they're reading reviews and then maybe the reviews take them into a Reddit thread. Um and maybe the Reddit thread maybe they come out of that and they finally come to your website. It's so it's again the the person's journey is just as unique of how they actually validate and verify. But to Austin's point, we we can tell by the conversations that renters are having with our teams. We can tell by the conversations that um the renter is having on our AI tools um on our website, on our chat um that they have had conversations with an LLM.

They may not say that. um they may not say I found you through chat GBT or when we're trying to attribute their lease. It it's not at this moment going to necessarily say I came to you through Claude, but we can tell because their search is they're much further along. They've had some questions answered. They know specifics that they didn't previously know despite a lot of other tools being available to them. Um so I think, you know, it's sort of the the tale.

it will, you know, continue to grow and it will just, it's not going to ever be the only source of truth, but we think it will become a dominant source of truth for them and then verified by other sources. So, Austin, it sounds like the mainstream reporting methods have room for improvement when it comes to measuring LLM search for multifamily communities. How can owners assess their properties AI visibility? So assessing visibility is call it a bit of a a easier conundrum than the attribution problem which I I think multi-touch attribution if you look across marketing in every vertical is one of the most difficult challenges but in terms of visibility in terms of measuring visibility AI visibility actually presents a new problem because in the kind of old SEO world in the old search world if you're number three. You're always or you're going to be number three for every single render who types in that keyboard. Now, I think all of us have used ChatGPT or claude and know that you can put in the same thing twice and you'll get two different answers.

So, when it comes to measuring AI visibility, kind of the the hard simple and kind of inadequate way is just to type it in yourself. Type it in yourself. Try and start a new browser window, fire up another incognito session to try and measure how often a community has come up. However, that's really hard to measure the breadth of keywords and the breadth of interest that that renters may be touching on or routing to on their way to discovering your community. Um, and you can theoretically do this and run through an entire battery of of questions and ask each question 10, 15, 100 times, but that's incredibly timeconuming, especially when we're talking about large multifamily portfolios where every single community may be positioned slightly differently. It's in a slightly different neighborhood.

they may have a slightly different call um demographic or price point that typically rents at that community and testing for that would would take you all day. Um now there are AI visibility tools that are built for more kind of general purposes or or even AI visibility tools that are offshoots of existing SEO tools. Now existing SEO tools, there are some multifamily specific ones, but many people in the industry do use kind of general purpose SEO tools because you're just measuring on a handful of keywords. It's not hyper domain specific. Now, if you want to use their kind of AI visibility functions, you have to set up all these different topics. You have to figure out all these different price points.

You had to figure out all these different points of interest that may come up in the conversation. You may you would have to look up community by community. Okay, like here's the likely places that people are probably going to be working around this community so that the so we can measure the commute queries. And while these tools exist and while they do work, it's a lot of setup. It's a lot of figuring out community by community what a renter would care about and then also when it comes to the information parsing through them. Um because a lot of these general purpose tools like whether it's a call it general purpose tool specifically designed for AI or an offshoot of an existing SEO tool, they don't know the difference between an ILS and a community website.

they don't know the difference or they don't know that Zillow and Trullia are actually the same parent company and that all those results should really be looked at together. Um, now that's really where we kind of we came in and when we were doing our own efforts to try and measure visibility in these large language models. Um, so to take kind of a step back, we for our 3D touring product, we always looked at how to get the most renters through. You can create all these, you could create as many 3D tours as you want. If you're not getting in front of the renters, it's useless. So, we looked at this and said, "Okay, we need to be able to measure this." But we didn't want to go through and use one of these tools and have to set it up for every single different community.

So our product peak discover starts really with with first principles and says okay all right how how would a render engage with this and automatically populate all these topics automatically figure out okay here are the likely commutes here are the likely points of interest and then automatically run all these tests like essentially have our servers spin up thousands of conversations and compile all the results so that you're not sitting there just recording or copy and pasting results into a spreadsheet and saying, "Okay, all right, River Trace at Ford Imperial was mentioned 16 times out of 50 on when I talk about the pet park." Um, so when you can actually do this at scale and when you can automatically essentially individualize the battery of questions that you're testing, you can actually get a true picture of AI visibility without having to go and essentially do all the manual setup and do all the leg work yourself. Okay, that sounds like a really efficient and effective approach. Austin, um, let me ask you, Nicole, have you used Peek Discover? We have. Um, you know, it's interesting because at, um, I don't know, probably six to eight months ago, my team and I were talking about, you know, things that we wanted to focus on for 2026 and things that were really important. And you know, like I said earlier, we we have seen the LLMs become um an important part of the search for customers.

And we felt good about how we were showing up. Though to Austin's point, like we never had the um the the capability and sort of the bandwidth as a team to go and do all those searches ourselves. It's just, you know, our traffic was really good. We were getting really high qualified traffic into um our leasing office. their conversion was really good and high quality. So, we've always been really focused on ensuring that that the level of renter continues to be what we're providing actually matches their needs so that we we have that, you know, shortened lease window is the ideal goal.

Um, and we had been talking about that we feel good about how we're showing up because of our approach to marketing, but we didn't have anything to validate it. Um and so ironically we were asking some people about hey how are you guys approaching in terms of like on the vendor side how are you guys approaching um LLM visibility and then ironically Austin came to us and said hey will you take a look at this thing that we're doing you know they were doing it for one reason but interestingly we we think this this could actually be really useful and it was just perfect timing because our team was already thinking about that and we were looking for a way to validate um and also to give us something on a you know right now we do it on a monthly basis my marketing managers meet with people on Austin's team to understand uh how are we doing how can we tweak things and you know I've in the past compared LLM visibility to a bit of a rebuk's cube because you do get different results every single time from claude or Gemini you want to be careful about how much stuff you are tweaking as a marketing person it can be very tempting to get a report and go and change everything on your website and hope that it is going to result in um you know increased visibility and that maybe not the case. Um we're still we're all tinkering around with it and playing with it. Thankfully Veris um we happen to have an umbrella website so all of our properties sit under one website. Um and I think that it's been really helpful and powerful for us as as marketing tool but not every company has that. every, you know, some companies may have their own um a different website for every single um property.

Um but the same strategy I think still applies is that you need to personalize that content. You've got to um provide that nuance. You've got to provide storytelling. You more content is probably better um less generic sounding things. So, it's actually really exciting opportunity for marketing people and brand people to finally be able to tell the the narrative and like how are you different from your competitors? What's the one thing that every tour comes in and says that they're impressed by?

Maybe it's your artwork package um throughout the building that the interiors team could have spent hundreds of thousands of dollars on, but there's no checkbox for that um on Zillow, but you can finally tell that story. to the person who is you know loves artwork and that would really appreciate that that story comes through to them. Um so it's been for us it's been you know still a journey and I think it will continue to be the the interesting thing about these things is they're going to these LLMs are going to continue to evolve and new models are going to come out and we're going to um continue to play around with the content that we're you know serving up to them. Um but we you know we know that we have a benchmark now a place to start. We have some metrics that we can actually measure um and understand how we're doing. Um so that gives us the room to improve.

So Austin, what do you think about Nicole's strategy? So I mean I I would say measure measurement is is really key in that um as Nicole said like V now has a benchmark there's now a baseline of okay are we doing better are we doing worse um and that's really like that that's really number one is just to understand kind of where you Now, when it comes to the individual strategy of of the various properties and one of the things um that actually um that actually stood out to our team was the percentage of these LLM citations that were pointing towards the various property websites or the umbrella website. actually consistently across top performing communities, we find that because they have that breadth of content, because they have unique content um on their website, they get a much larger percentage of their citations or their their mentions by these LLM pointing directly towards their website. like to go from kind of good to great really like that that's the biggest differentiator and from there it's not just having this content um machine readability is also key um making sure that these websites are able to be picked up uh by these LLMs is a big piece of this. Now many of the kind of stock solutions or templated websites um that are prevalent through the multifamily industry aren't very machine readable. So when it comes to augmenting this content, when it comes to telling this brand story, it's also really really important to make sure that it's not just visible to a human being going on to rubber trace.com, but it's also visible to an AI engine pointing towards that because now it's not just call it your own domain survey information you're going to have people getting the same information looking at claude looking at chat so assessing that readability and and and that's actually one of the big pieces of having this reporting in place is that it allows you to verify whether or not the content that you've invested all this time into creating and the story that you've invested time into crafting is visible by the AI engines and and hence by the millions of renters who are starting to go through these in for their rental rental search process.

And that's a great uh lead into the final portion of our webinar. Let's talk about some best practices for owners to improve their communities visibility and AI search and LLMs. Nicole, what do you suggest? I mean, at the end of the day, um, as much as we've talked about has changed in the the last 20 years in terms of how renters find the information, it still starts with location. Like people, that's that's the first thing that people, it's still location, location, location. It's still really important.

Um, that you know, if your building is not in a place that the person wants to live, that it is what it is, right? Like they they have a specific place that they are looking to live and then it goes sort of from there. I do think that the LLMs, as we talked about, are opening up the aperture for the renter of places they could live um that they are not necess that they maybe wouldn't have thought of themselves or looking on Google Maps. Um they wouldn't have understood that, oh, it actually would be very easy for me to live in Jersey City and get to downtown um and they can have that conversation. So, we're finding that, you know, neighborhood pages extremely extremely important. um listing out the amenities with specifics, not um generalization, like as specific as you can get.

Storytelling, again, I think it's a really exciting time for brand people and marketing people because you can actually now sort of tell that story in a in a a descriptive way that you wouldn't have been able to um previously. Also, you know, we were talking a lot in sort of the marketing world and multifamily about transparency. Um there's a lot of push for transparency across fees. Um I think it's really exciting because again my goal is always to get the most qualified person to our team who is really ready to sign. Um, so I think you know just as much detail as you can provide um on the apartment sizes, on the features, on the finishes, on um fees, um you know the the price points um just as much content really, you know, as you can um provide. And I will say um one thing that I didn't get to to share earlier is that like right now it is not making our team's jobs easier right now.

It's it's making our jobs um more nuanced. there's more things that we have to pay attention to. Um, but again, I think it's a really exciting opportunity to actually share information with renters that previously would have just been discarded um or sort of deemed unimportant or actually could maybe impact you in a negative way through Google. Austin, I mean, I I think like a lot of the things Nicole mentioned are absolutely spot on and even to go into kind of the the location piece. Um, I would almost liken it to the types of location information that wouldn't necessarily be relevant or or or help you on Google, but that like one of your human leasing professionals would be engaging in a conversation with the renter. When we analyzed some of the top performing communities that we've seen on on uh Peek Discover, like the best communities, they're mentioning nearby highways four times more often.

They're mentioning transit hubs three times more often. It's all about weaving in the entire narrative into the content as opposed to, hey, we're in this neighborhood. This neighborhood is copy paste from Wikipedia. I'm sorry. I'm just I'm having trouble. I'm having trouble seeing anyone who's who's gonna say that we're close to the Long Island Expressway.

That's going to be a positive. There a little bit of editorial discretion does go. Sorry. Yeah. Just had to jump. If you're going to have to like if you're going to have to take some highway for a commute because the major employment center or or the headquarters for a major employer is along the highway saying, "Hey, you have easy access to these highways." Not necessarily say, "Hey, the the property is right next to it and we hear it." that probably isn't the right kind of contextual information, but letting people know that okay, this fits into their lifestyle and and um that type of information um plays extremely well.

Now, there's also a difference or at least currently there's a difference in how these LLM are sourcing information. That means that this can't just be on your property website, right? like Claude is much more heavy in sourcing information for property website, but catch GPT is very ILS heavy. So if you have all this on your property website, but it's not in your Zillow descriptions. It's not in your apartments.com descriptions, again, you're you're going to be missing a piece of the puzzle. Um, now this is continuing to evolve in terms of the information that matters, in terms of how these elements are sourcing information, um, in terms of where they're sourcing their information.

And I I mean, if you think back about a year ago, every single SEO GEO agency was saying, "Hey, we'll get you to show up on chat GBT. We'll do a ton of Reddit posts for you. pay us and and we'll we'll put us on a Reddit post because chat GPD is sourcing everything from Reddit. Guess what? Today, Reddit isn't even in the top 10 for sources in terms of citations. So, there's an evolving landscape that also just needs to be monitored and what worked three months ago, six months ago may not work with the next model.

So it it's very hard to or or I would I would say impossible to fully predict the future but kind of the table stakes to this is really to number one have a way to measure have a benchmark so that if things are if things you're doing are are inflecting positively that you know if things that you're doing or or changes that you make um for example a lot of the fee transparency uh changes that multi family properties made to their website inadvertently removed pricing and availability information from uh from being machine readable. So any renter looking for availability of uh of apartments at a specific uh at a specific property or even in a neighborhood ChatGPT and Claude were preferencing properties that would actually tell you, hey, we have these three units available. it is priced between X and Y versus the hey I know that this property exists but I don't know the inventory. So having a way to assess, having a way to detect those blind spots and missing information and gaps in call it machine visibility um are I would say like kind of common threads that will remain important u no matter how kind of the landscape of L and visibility changes. Those are great insights. Austin Nicole, what do you think of all these suggestions for your fellow marketing leaders?

Yeah, I I agree with Austin. Um first step is to measure um to understand where you are and and it's okay if um it means that you're not particularly visible right now. Um so it's okay to get the response back um that you're not performing particularly well. At least you have somewhere to go from there. So I I am always a fan of understanding where are we today? Um you know what can we do about it and then being able to make the changes.

Um, you know, LLMs are not going anywhere. Um, so I think we have to accept and acknowledge that, um, AI conversations are going to be a huge part in a growing part of how renters continue to find us. Um, and again, I feel excited that it probably means people who are coming to you are going to be more qualified. It means that by the time they get to you, um, if you do this well and you really commit to doing it, it means that you are going to get just a higher quality of renter. Uh, I do expect at some point this will positively impact our marketing budgets. We're not seeing it right now because everybody's kind of just getting going with it.

Um, for the people who are hearing this who feel scared and feel behind. You're not. It's okay. Like everybody's, you know, we're all just starting in this, the good news is that there's not Austin and his team are ahead. Um, but there's not any real expert in this area. So people can, you know, should you choose to commit to this and accept that this is a really valuable part of your marketing plan, you can catch up um and you know, certainly take advantage um of this soon.

So I would also recommend hiring um someone to help even whether it's the benchmarking um whether it's just understanding where you are and then you know you're going to take it from there. Our team does a lot of the work internally. Um, if people have the budgets that support having an agency, try to help them improve their results. That's always, you know, fantastic. Um, I think my biggest recommendation as a leader is to get your team excited about AI. Um, they may be scared or nervous about AI.

Um, get them excited that it is an opportunity for them as marketers, um, to really tell the story. Um, tell a nuance story. tell a story of how your buildings are different, how your company is different, um, and really paint the the picture for the renters. Um, you certainly have to pay attention still to everything. You can't ignore your ILS's. You got to double check that, um, where they give you opportunities to tell the story, build it out, you know, a bit more.

Um, and then for the teams that, you know, I've worked on all kinds of multifamily. So, I've had properties that had very very small budgets and we wouldn't have been able to afford to do this and maybe not even be able to afford to um hire a team like Austin's to give us that benchmark. Yes, it is um timeconuming, but it's really important. You can check these things yourselves. You can have the conversations with the different LLMs. Um as Austin mentioned, open up different windows.

Um try chatting with them. Close out that conversation. Try again. you can get a good sense of where you stand, maybe you won't have the metrics to back it up, but if you never show up, if you're having a conversation where you know what your desired outcome is to show up, then that will flag to you that you you have problems. And then again, you can just play around with the content that you're providing. So, if if you're not showing up, you can try adding stuff to your website.

You can try certain social campaigns. You can try adapting your ILS content um and see if that improves it. Um, so you know, I think organic content is going to continue to be um, king and important and you just have to be willing to adapt. Um, that's the best thing about marketing and being in our field is that every day is different. So when these LLMs change, it's just going to provide a new challenge for us to respond to. Terrific.

Thanks, Nicole. And Austin, any final words before we get to our Q&A? I think the last thing maybe I would emphasize is just the magnitude of this opportunity. Um this is a call it once in a decade or or really generation shift like I would say akin to the shift from those paper apartment guides to the internet um where call it the front door to the renter is changing. it is starting to move away from from Google and traditional search. And again, it's uncertain kind of how this may end up playing out over the next couple years.

But in this time of rapid change, there's so much opportunity for the marketing leaders, teams, and organizations who get this right to really capitalize on this. We're seeing that the top 5% of communities are getting over 50% of the mentions in these large models. This is like nailing SEO in like 1995, 1996. This is an opportunity to get an early jump on something that's going to be huge. Um, I mean, I think I've already gone through a number of kind of suggestions for for how to really get started, but really it's this is the magnitude of opportunity and the potential for this to even just shake up where marketing budget can effectively be spent. Um, if you look at kind of a typical property marketing budget today and where it's allocated, a lot of that is allocated towards sources that their efficacy may change.

Their efficacy in in their ability to drive leads or the quality of leads that they're driving may go up, may go down. their share of essentially the renter eyeballs is going to be shifting because you do have this monumental shift particularly when you're talking about the renter demographic this demographic that is likely being pushed to use AI at work um to their entire kind of behavior and or interaction with online information is is moving just to a different platform. So um with all this opportunity there, it's it's an exciting time for for any organization team or person who wants to invest the time into capitalizing on this. And with that, thank you Nicole and Austin uh for your insights and actual actionable steps for our audience. Great conversation. So let me uh kind of summarize what I heard.

the the multifamily industry is is not immune to the impact of AI technology and increasing use of LLMs in everyday life. And by assessing your community's AI search visibility, measuring it, you have the opportunity to get ahead of the market and grow your reach today. And uh with that, let's hear from you in the audience. Uh uh keep your questions coming. We only have time for a couple. So um the ones we don't have time for, we'll be sure to share with our presenters.

Let's start off with a question from Charles who says, ' Like the early days of SEO when websites were embedding or camouflaging keywords into their web pages, how susceptibles are the large language modules to being uh models to being manipulated in their responses. Manipulated I think is the probably the wrong word. Okay. Um they can be influenced um they can and what these large language models pick up and what they read isn't necessarily the same thing that a human renter is going to read. Um there can be uh things in the metadata in the descriptions or even just pages in the website that aren't indexed um that these large language models will find, they'll pick up and they'll integrate that information. Um so right now there because of the call it computational complexity for these models to be able to determine okay is this something that would show up on a human read version of the website versus does this just show up when I point the AI model to it.

Um they don't do that comparison. So you can add a lot of information. You can add a lot of pages. You can add a lot of context that is even in addition to the content on your website that may not be visible to um that may not be visible to a human renter but that can augment uh your visibility on on these large language models. Yeah. And if I can expand on that thought, I I think when please back when we were looking at SEO only, you were competing for very generic terms.

So you were very very top of the funnel. Um it wasn't nuanced. It literally was luxury apartments Jersey City. There are a lot of luxury apartments in Jersey City. So the goal was just to ensure that you got a good chunk of market share. And that then the person would come into your website and after a while they might discover that this is way out of my price point or they would hop out of your website for a variety of reasons or you know wrong location what whatever it may be.

It was always something that was um we could have sort of sifted out early on but because we were competing on generic terms you know you you really needed to get that person into the website. You couldn't um sift them out very very early on. Whereas with the LLMs to to Austin's point, um it gives us an opportunity to get somebody here who's actually qualified, you know, they've had that nuanced conversation and most marketers our goal is not to just send tons of traffic to our leasing team. It's to send high qualified traffic to the leasing team. Right. So I I agree with his Yeah, I agree with the assessment that it's it's a manipulative might be the wrong word, but it it's providing enough information to the LLMs so that they are able to serve up um the right person based on their needs.

Okay. Okay, Charles, I hope that answers your question. And this one uh Austin is for you from Joyce who asks, "With all the moving parts in LLMs and AI technology, how often are you checking the accuracy of your AI agents or servers that are testing and reporting on your customers visibility?" So I mean we we have continuous monitoring uh on uh on all of our servers, all of our systems uh to ensure that all they're functioning as we expect. Um but we also do frequently assess our approach for testing as well um in ensuring that we are getting or or or figuring out the right topics um that should be tested in making sure that what we are or how we are testing um isn't essentially influencing itself. Um, one one actually issue with manual testing is that if you don't clear out that browser window, uh, if you don't start the session as a new user and you start asking more and more questions about a certain type of property or certain features at a property, even if you don't mention it in the next chat, it'll start biasing the results um, in subsequent or in in subsequent chats. Um so that is something that we that we keep an eye on.

Um but we are constantly monitoring and and making sure that okay are the results that we're getting back valid are they are they credible? Are they call it obviously correctly being tabulated? Um, but as a as a whole, it's a lot more of kind of the accuracy is just ensuring that we're covering the right topics and the right concerns and that we're mirroring the behavior of the renter. Um, now that does tend to be a bit more stable. Um, renters do tend to care about the same call it on a broad scale 20 to 40 things. Uh, and we're not going to kind of suddenly see a resurgence of renters who are primarily focused on their ability to connect satellite TV or or something kind of like out of the blue.

So, um, really like the the bulk of the accuracy in being able to determine the overall visibility of the community is based on a relatively stable set of rental preferences. And with that, that's all the time we have today. I want to thank you both, our speakers, Nicole Jones and Austin Lo, for an eye-opening discussion. We appreciate hearing about your insights on the growing influence of AIdriven search and LLMs in the marketing multifamily communities. I'd also like to thank our audience for joining us today. We hope to see you again soon.

Thanks again. Have a great day.