In short
Eye On A.I. Podcast Episode Summary
Episode Title
#158 Connor Leahy: The Unspoken Risks of Centralizing AI Power
Host
Craig S. Smith
Guest
Connor Leahy, CEO of Conjecture
Episode Overview
This episode focuses on the themes of AI safety, governance, and the implications of centralizing AI technology. Craig Smith interviews Connor Leahy, who shares insights into the risks associated with monopolizing AI advancements and the importance of open-source solutions.
---
Key Discussions
Introduction
- Sponsor: Netsuite by Oracle
- Overview of Connor Leahy's background in AI safety and open-source large language models.
Dangers of Centralizing AI Power
- Discussion on the risks of centralizing AI technology and the implications of monopolistic practices.
- Connor draws parallels to historical examples, emphasizing the necessity for governance in AI development.
Open Source AI
Safety and Risks
- Conversation around the importance of open-source AI in promoting safety and transparency.
- The potential hazards of proprietary AI models and the need for more responsible development practices.
Flaws of Self-Regulation in AI
- Addressing the inefficacies of self-regulation within the tech industry.
- The call for more robust policies and regulations to ensure ethical AI development.
Policy Proposals for AI
- Connor outlines his proposals for policy changes that could enhance AI governance.
- The need for interdisciplinary collaboration among technologists, policymakers, and the public.
Implementing Safety Measures
- Discussion on practical measures like the implementation of a "kill switch" for AI systems to prevent unintended consequences.
Public Opinion and Policy
- The role of public perception in shaping AI policies and regulations.
- Emphasis on increasing public engagement in AI governance discussions.
Risks of AI Disinformation
- Examination of how AI systems can be manipulated to spread misinformation.
- The need for vigilance and proactive measures to combat disinformation campaigns.
Survivorship Bias in AI Risks
- Discussion on the concept of survivorship bias and its implications for understanding AI risks.
- Importance of acknowledging failures to prevent future occurrences.
A Hopeful Outlook
- Despite the challenges, Connor expresses a hopeful perspective on the potential for AI to benefit society.
- The importance of collective action in shaping a positive future for AI.
---
Key Takeaways
- Centralization Risks: The concentration of AI power poses significant risks; a diverse ecosystem is essential for innovation and safety.
- Open Source Advocacy: Open-source solutions can lead to more ethical and manageable AI systems.
- Regulatory Needs: The tech industry requires stronger regulations to govern AI development and ensure responsible usage.
- Public Engagement: Greater public involvement is necessary for effective AI governance and policy-making.
- Disinformation Concerns: Vigilance is critical in addressing the potential misuse of AI for disinformation.
---
Conclusion Connor Leahy concludes the episode by urging careful consideration of AI's capabilities and advocating for a cautious approach to its development. He emphasizes the need for collective responsibility in shaping the future of AI technology.
---
Additional Notes
- Call to Action: Listeners are encouraged to leave a review and rating on platforms like Spotify and Apple Podcasts.
- Engagement: Craig Smith invites listeners to stay updated through social media and links to the podcast's channels.
---
This summary encapsulates the discussions and insights shared in Episode #158 of the Eye On A.I. podcast, highlighting the critical themes surrounding AI safety, governance, and the implications of centralized power in technological advancement.
Written by AI. May contain mistakes. Listen to the episode to check what was said.
Transcript
Automatic transcript. May contain errors.0:00Nobody wants to fail in full view, especially property team members who don't get a chance to collide with the corporate staff because they're on this tribute they don't want that they don't want to be exposed they don't want to have an additional opportunity to look bad they don't want that so you paint the picture of what good can look like this is going to mean this for you we're going to hand off qualified leads you no longer be in a conversation for an hour only to find that somebody's not qualified to rent a class a and it and you start to tether to the things that they care about owning your footprint is really important finding a really strong third-party partner who can help reduce your dependence on ILSs is critical.
0:33Somebody who's mobile first that understands kind of the ever-shifting landscape of doing that well and really can make sure that you're always competing. Okay, we're here with Dustin Lacey from Mark Taylor. Dustin, why don't you introduce yourself? Hey guys, Dustin Lacey here is indicated Vice President of Marketing and Technology here at Mark Taylor Residential in sunny Scottsdale, Arizona. Welcome. welcome to yourself cool um so dustin you've been with um do you guys always say the whole mark taylor do you say mt what do you guys do mtr mtr okay all right now i'm an insider um you've been with mtr for how long um so this is gonna be year eight okay yeah and i'm uh I didn't have multifamily experience, so I came in from American manufacturing.
1:32I worked in oil and gas, semiconductor and solar. And then most recently, I oversaw a business group within an irrigation company called Rainburn. Cool. Well, we're really excited to talk to you. One, because Karina, who had worked with you at Mark Taylor, has said so many good things about Dustin. But also, a lot of what Karina has talked about to me specifically has been, I guess I'll call it a lot of the change management that you've helped lead up there and Mark Taylor over those eight years. So, would you give us a little bit maybe of your, if you were going to do your eight years and three chapters, how you would talk about those projects that you've spearheaded?
2:15Wow, that's good. they brought me in because we had a you know American manufacturing you have a data kind of centric approach because of just the constant encroachment of international manufacturing and if you manufacture stateside you know erring on a business decision can have broad implications and as margins kind of compress, you're putting yourself at risk if you're not relying heavily on data. So when Mark Taylor pulled me in to multifamily and I'd lived in a Mark Taylor, I'm an Arizona graduate. We've got three teams in the top 20 right now, which is unusual, which is why I know it and why I'm also happy about it.
3:07Would that be all Arizona teams? Arizona Wildcats. Dustin, okay. I just want to know if we're talking strictly Wildcats or if you're looping in some of the other colleges. Yeah, yeah. No, no. No. We're doing well. And so I joined at a time when we were looking to make a pivot. We were at that time managing AMB assets. We are an owner-operator. Mark Taylor is. Founded in 85 by Jeff and Scott. we manage about 30 ,000 Class A units today. At the time I was brought in, we were managing B as well and moved to an exclusively Class A posture. About a little under a third of our units are our own properties.
3:54And we manage fee-based as well for the balance of communities here in Arizona and Phoenix. And in 17, when I joined, we were looking to really start to evaluate alternative marketing approaches to kind of what is the standard approach within multifamily even today, which is ILS heavy spend. And when you consider the fact that lead gen is kind of the lifeblood of your community from a occupancy, rent growth, etc. standpoint, the more control you can have there, the better. So chapter one was really about spending a year evaluating the true value of what we were currently doing from marketing perspective, auditing, learning the industry, learning Mark Taylor.
4:44And then the decision was taken shortly thereafter to move off of ILS's altogether. So we knew that we wanted to bring in our own marketing function internally, manage that lead gen piece specifically. So So the first bite we took was establishing a dual website footprint with marktaylor.com operating as a private ILS in essence. And at that time, we used to get all of the unique, we were a heavy ILS user, right? We were spending, I believe, 15, 80, this is 2017 numbers, obviously much higher now. But I think we were spending just under two grand a month per property with parvus.com. We were on RentPath.
5:30We were on ApartmentGuide. We were on Zillow. We were spending quite a bit of money. And so the specter, this notion that perhaps we could own our lead gen and drive out cost was very attractive. It's not very common, though. So it was a bit of a risk. Our departure conversations with the ILSs, who I think had enjoyed a very long and prosperous relationship with Mark Taylor up to that point, I think that it was quite a surprise. And we caveated the conversation with, hey, and, you know, the powder's still dry. You guys will be available to us. We hope. We think we've departed on amicable terms.
6:12So you guys, if we need you, we'll call you. So we did that in 18. And we timed it with the creation of our new MarkTaylor.com and standalone websites for our properties. So today, each property enjoys a property listing on markdail.com and then has its own domain and own property website that is mobile-centric, right? Google started indexing mobile first in 18, I believe, in March or May of 18. And we knew we needed not the best-looking sites, perhaps, although we're proud of what our sites look like, But they're certainly not festooned with a bunch of bullshit. They're built specifically to meet the needs of, honestly, Google, right?
7:03So we spend a lot of time evaluating accessibility, SEO, performance, page, all the different things that they care about. And, you know, you guys certainly know this, but Google's refractoring their search algorithm every day, a couple of times a day on average, based on the limited information they allow us. And so we knew we had to get really good at producing sites that were fast on mobile. At that time, I believe 40 % of traffic, mobile traffic still matriculated on 3G, which is, you know, just a step up from dial-up. So you can't, you didn't have the luxury of requiring a prospect to load a pre-rolled video above the foldery and that stuff.
7:44And so we, and we were doing all of that. And it looked good if you had a great connection. But truly, that wasn't the reality. Most folks, 76 % of our traffic arrives via mobile device. Of that, 72 % are iPhones specifically. So our thought was, hey, if we step off of ILSs, this is chapter one still, we've got to be able to generate the requisite leads to satisfy the leasing appetite for each of our properties individually. So bringing in-house ad strategists and dedicated account managers was also a requirement to this. We found that when you were sourcing third party for ads, it could be challenging.
8:33This is our own experience. I know you guys offer this service and do it quite well, as evidenced by your client set. But for us, we had local folks who may or may not have had competing interests. And so in essence, they could play kingmaker on any given day as to who was going to be generating leads. Because we know, based on CoStar data, that on any given day in Phoenix, only 8 % to 10 % of the available, they who are searching for apartments can even afford a class A community. We don't have the luxury of not doing these things well. So we deployed two website, dual website footprint, both were mobile centric, started to own our lead gen and quickly realized after we were fortunate enough to succeed there and successfully move off LSS, we realized that lead management had to be next.
9:31We, at that time, utilized a tracking number platform called LeaseHawk. I think they're probably still the dominant provider in the space nationally here in the U.S. And in 2017, they used to come in and they'd bring us nice gifts. You know, Mueller's team is a great team. And, you know, they'd always extend all these plaudits as to how we're doing. And I finally was just like, you know, after all the back padding and they'd be like, you guys are top three in the country for a portfolio this size. And we felt great. And finally, I was just like, what is our call entry rate? Truly during operating hours and then full, right?
10:14Including after hours. And it was around 51%. And so in that first year, as I alluded to, we took a deep dive into our cost per lead, cost per lease. We really want to understand customer journey. This notion of reinforcing buying decision and reducing barriers to intent was really important for us. That's really what fed the UI UX piece of our websites. We really wanted folks to get where they wanted to as quickly as possible. So enabling to do that instead of creating obstacles with websites that probably weren't thoughtfully conceived was a non-starter for us. And in the process of doing that, we recognize that our cost per lead was 70 to$700 per lead, not even per lease, but per lead.
11:01and so here we'd stripped out all of this cost which in today's terms 2023 terms is about four and a half million dollars in ILS spend reallocated a portion of that to um paid campaigns paid search you know PPC
11:21remarketing geo etc we recognize that we were still falling down from a lead gen standpoint so So that brings me up to chapter two. Chapter two was really kind of the Genesis story for Centralize for us. Here we were now creating a healthy, consistent, peace of mind inducing lead gen engine, but we weren't answering our phones at the rate we needed to. And we figured that out when we took a deep dive in the probability of converting a lead to a lease based on whether we had answered the call or not. So when we answer a call, today's terms, if we answer a lead call, we have a probability of converting to a lease that's near 40%.
12:08If we don't answer said lead call, our probability of converting to a lease is less than 2%. It's almost not even worth calling folks back or working those leads. It became very apparent that we needed to solve for lead management. So we embarked on this centralized approach wherein we have established pods of three properties, identified really good leasing team members, created additional career pathways for individuals. It's no longer just about becoming, you know, starting as a leasing consultant, moving to a senior lease consultant, moving to an assistant manager, moving to a manager, and then, you know, the Holy Grail perhaps is corporate.
12:53Now we've got this additional tree of opportunity for folks, which has proven to be very, very helpful. And I'll speak more to that in chapter three. But we identified these successful leasing individuals, allocated three properties to each individual, and then started answering phones at a more regular basis. And again, it turns out when you answer your phones, you drive out costs, and you create a more favorable impression for prospects. You improve the probability of converting to a lease. These are all super valuable for us, especially today, given all that we're kind of facing as we step outside of the last decade of abundance in terms of lead gen for multifamily.
13:41You know, it's pretty stark what 23 and 24 look like relative to, you know, 20 and 21. even during the COVID era. So we started answering our phones, right? We created this centralized focus. We secured a second building, placed folks in it, and outfitted them with software. We're on the union platform, which is purpose-built for centralized specifically, a consolidated dashboard. And we started to create additional channels as well when we recognize that self-servicing as a behavior became more prominent with our target demographic. So we're talking about AI chat tooling specifically, deployed that to our websites, allowed folks to schedule self-guided tours and recognize, really recognize what Jay Parsons from, I believe he's from RealPage, recently wrote about on LinkedIn, where he said that upwards of 70 30 plus percent of U.S.
14:43residents, irrespective, or multifamily residents, irrespective of class type, prefer zero touches with their property team. And so at that time, we didn't have that data, but we were recognizing that, hey, if you can enable folks and create and allow them immediacy in terms of self-servicing, they're going to opt in for that. They're actually going like that. And in parallel, what we started to notice was that you go to your Home Depots and your Walmarts, your Targets, and they too had self-servicing. And so we kind of were timely there and got lucky. And so we started with lead gen, right? So chapter two is really about lead gen, perfecting that, understanding what that looks like, making sure we had software that met the need of the moment that allowed an individual to field calls and inbounds because we have live agent chat from our chat tooling.
15:37So if you don't want to talk to an AI chat bot, you can actually talk to one of our concierge live. What we recognize was folks actually quite like this. And from a convergence standpoint, we're converting at the same rate or better relative to our on-prem management of the process. And so with these distributed workforces where you've got 90 plus properties like we do, It's very difficult when you consider the turnover rates at the entry level positions, both on leasing and on maintenance, are very high in our industry. It can range from 30 to 60 plus percent for leasing consultant service technician.
16:16These are high turnover roles. And oftentimes, the lead management piece can fall to your greenest individual or your newest person. So not only are they not seasoned, they're brand new, they're probably just learning the industry, learning their community and learning Mark Taylor. But you are requiring of them to manage a very precious asset, which is your leads. And so when we started to take over lead management, it proved valuable from a lead, from a call answer standpoint. point, we moved from a portfolio average of 52 % upwards of 90 plus percent from a call answer rate for in hours. And when you factor in full day, when you don't have somebody answering, obviously that brings that down.
17:09But that was a material, material change that impacted our costs and our conversion rates sizably. So chapter two was really about that. And then chapter three really started during COVID when we closed all of our communities down as required by our governor. And we were deemed an essential service as property managers here in Arizona. And Governor Ducey allowed us to continue to go to work, but we couldn't keep our offices open. And when we did that, we thought that there would be a significant outcry. And there was around the amenities like pool and gyms specifically. And I understand that because that would have sucked.
17:51I'm paying full freight and now I don't have access to key aspects of the value proposition. But that was out of our hands. But what we learned was that, hey, even though our office was closed, if we allow our residency ability to schedule an appointment with us versus the walk-in scenario, which is most common, we have time to prepare for that conversation. and give dedicated time to that individual. In fact, we're able to improve their experience, which certainly reinforces buying decision, improves our retention rates and makes them happier in the end. And we know this because we surveyed them.
18:23And so that was heartening for us. So simultaneous to these kinds of learnings that started to percolate as a consequence of this variable, this COVID variable, for which none of us had that in our kind of our business model, we recognized we needed a full stop CRM that allowed us to at any given point understand the key performance metrics that really mattered for us. Not just call answer rates, but conversion rates of lead to lease, lead to tour, engagement rates. We have a sub two hour engagement rate requirement, meaning if somebody emails you, you need to be in touch with them via phone first.
19:05If they don't answer, text message, email within two hours. It's critical. Again, back to the point I alluded to earlier, only 8 % to 10 % of folks search for apartments in Phoenix can even afford a Mark Taylor. We don't have the luxury of taking a passive approach here. And so we deployed a CRM and you mentioned Karina, pronouncing her name perfectly, by the way, you guys are good. Thank you. She led that rollout actually for us and did an incredible job. I believe lockdowns hit March 11th of 2020, and we rolled out the CRM the following week in five business days. Karina did that. And that proved to be incredibly valuable for us.
19:50So now you have a consolidated cockpit from a reporting perspective that allows us to really ensure that we're doing all these things really well. Because even though we had our offices locked, we still had people on prem, right? And so as we're kind of cultivating the centralized approach, which we hadn't yet rolled out, we still needed folks managing leads thoughtfully. So the CRM proved to be very valuable for us. You know, we do all the piece, all of the call scoring and all the different data capture that's important to understand what's going on so that at the end of the day, and this is important, I think anybody from multifamily that's listening would agree with this.
20:31We just seek peace of mind at the end of the day. And it's very difficult with distributed workforces to really have confidence and certainty around delivery. And once we launched our CRM, we really started to understand that we could do this. Maybe this centralized approach, even though nobody else really appears to be doing it this way outside of Jim Ristel and Portland and a few other folks, Suzanne Hobson out in North Carolina and Tricap. Those folks have been very thoughtful. Irvine Companies had some self-guided touring. I mean, there's some folks that had towed the water and taken pieces of it.
21:11So at that time, we felt very comfortable with our lead gen and lead management piece, and we branched into resident management. Resident management, if you look at the total engagement for any given property, it makes up about 60 to 70 % of engagement for a concierge or a property team. So if you're thinking about how they're allocating or the utilization of your property team members, if it's anywhere close to what we're seeing, it's roughly 60 % resident, 60, 70 % resident. And the balance is on lead management when you're talking about the management of prospect and residents. So we knew this was gonna be a bigger bite.
21:58We knew this was gonna entail some additional requirements for our concierge. They needed to be fully fluent in the properties that they oversaw. They needed software that would allow them to answer questions in real time, just absent full fluency around that property. There's given days where our concierge may have never been to any of these three properties they're managed. So we deployed our generative AI chat tooling to the backend and listen to all of our calls and do all these really fun things that surface answers in real time, whether they're on the phone, penning an email, we generate all of that for them.
22:38So to create efficiency for them because of the diversity of things they may confront as a concierge from a centralized standpoint. So for us, chapter three has really been about perfecting a couple of, a handful of key areas. The lead gen piece, which I've discussed, lead management to discuss, resident management, which I just hit on, IT, accounting, collections, and then the big, big bite for which we are still pushing a lot of effort towards, I mean, all of it still, but service and maintenance have been kind of the biggest challenge because there are just so many different factors that impact performance there.
23:22And so originally when we embarked on this chapter three efforting where we were going to take the whole bite from a centralized service standpoint, and we knew maintenance would be difficult. But I think that we've been, you know, So we were right. It's been very difficult. Where we've seen a real impact of payroll and performance on the leasing services side, the maintenance services side is more about creating efficiencies. And we're still mid-stride and kind of fully forming our approach there. Fortunately, we have some good partners. We're on the happy product from the service side. But that brings us up to current three pods of three or four.
24:09I think we're on our fourth pod of three properties. We have fully closed properties. I think our chapter four is going to be really interesting because unbeknownst to us at the time, the feasibility of closing a Class A community in full from a leasing standpoint, leasing office standpoint, seemed out of reach. right? Certainly because we refuse to offer less than a white glove experience. And yes, we fall down all over the place, just like anybody else for all of the reasons that I indicated to you guys. But today we're operating a handful of closed class A communities. Did we have a few property residents cry foul when we announced that we were doing this?
24:56you know, almost a pseudo return to the COVID era, because we have since reopened our leasing offices just for background. And the answer is, yeah, we had five or six on each of these properties that were like, hey, I really enjoy coming in and chatting with your property teams. And I said, you know, I get that because I would be one of those people that I because I like to chat and I would that would have been a bit of a loss. but for the vast majority of our residents, a fully closed leasing office, as long as they are enabled and have immediacy around what they need to do, meaning your resident portal offers them all of the abilities to submit a service request, pay a bill, do those different things.
25:40It really is a viable approach. And so chapter four for us is really about extracting as much value from the centralized model as possible um and seeing how far we can take this without reducing our performance not not only from the resident prospect standpoint but just overall operationally and so far we've been successful um certainly had many many hiccups but um we're excited about the future so that's chapters one through three boys yeah well i like that chapter one and two felt like meaty it's like um or like very precise it's like harry potter and the philosopher's stone and then when you got to chapter three you're like you know what i'll just put the next trilogy all into this one book and not not not write six books um it's just giving you a hard time well no i get it well there's a lot there we can unpack it but there is you're right there is a lot there i just don't want to bore the listener too much i'm working really hard on my us and ums just so i can not um belabor somebody's eardrums too much but yeah chapter three has has been a big bite i'll say this here's how we know it's working obviously i've hit some of the key success metrics like engagement rates call answer rates we survey so we're constantly checking we utilize jay turner for surveying both prospect and residents.
27:05So we get to the benefit of benchmark data relative to our marketplace and nationally, as well as what we've done ourselves portfolio-wise. And the numbers don't lie. We're seeing improved customer satisfaction scores, if you will. We've moved into top 20 Aura score for a portfolio of our size. While Karina was here, she launched our review and reputation efforts she's that good for folks that are wondering. So congratulations to Digital. And we now operate full portfolio 4.0 and above on average for review and rep, which you guys know has really strong implications from an SEO standpoint in particular, but also voice of customers, very important.
27:55Social proof is very important for prospects. But there's still a lot left to do and I can unpack whatever you guys want to relative to chapter four and we can get into a little further if you'd like. Well, let's first back up to chapter one. Reed, I figure I toss it to you because this has been something that's been near and dear to you over the years as far as like removing the ILS or what would that experiment look like, possibly internalizing the marketing, yada yada. So it's just that you might want to dive in there. Yeah, well, a lot of companies, I think at this point have tried and at the very least have talked about it.
28:33And so I'm curious how you kind of set the whole thing up. Did you rip off the bandaid? Did you stagger it, you know, so that you could prove, Hey, let's, let's do this with, you know, 10 or 20 % of our portfolio before, you know, we make the bigger move. And then is there a point where you mentioned that the powder is still dry and that you might call them back, you know, are you like, is that, has that ship sailed or you know as modern as you guys are and the change agent you know would you ever look at the ILS's again or is it almost a cultural you know element now you know source of pride to a degree but yeah the beginning of that was just how did you approach the first three to six months um as you were preparing for for this big move yeah it's a good question so So I'll hit the second 1B first.
29:27It's certainly a brand now for us, this ILS independence, because it is unusual. You know, we talk quite a bit about it with prospective clients. It is seen as a value. Our marketing costs per unit annualized are materially less than, you know, competitors. We underwrite a lot of the deals here in the finance marketplace. So we have true apples to apple data. but it doesn't mean that we're look no there's no pride and kind of authorship here we will do what we have to we're not bigger than the market and we'll do what we need to so have there been instances where we've deployed you know a rent.com spend or apartments.com spend certainly it's often when a new client who is perhaps unfamiliar with our approach and not yet trusting of it, particularly in the lease up period says, hey, it won't hurt.
30:24We've got, you know, we've we've we budgeted for this anyway. So, yeah, I mean, we'll we work for somebody, certainly. So we will we will we're obligated, but it's very rare, very rare and very unusual. But we've we've done it a handful of times. We have nothing negative to say about them. We recognize they play a very important role in our marketplace. But for us, you know, it's about doing the same with less money. And we've been able to successfully do that and haven't had to backslide or return, you know, knock on wood. Now, the caveat being that we know there is change coming relative to Google and how they handle their availability listing.
Read the full transcript
31:16So today, the multifamily approach that Google employs is much like what they did for travel about six or seven years ago. And for anybody that's searched for flights on Google now knows, then it's, I mean, they'll still surface in organic and in ads, you know, your kayaks, your travel offices. But what they also service front and center, which has reduced the available organic page real estate is their own module, travel module. And so we know that that's coming for multifamily, wherein Google is going to cut out ILSs or partner with them. We'll see how this works. And we've talked to the various ILSs about it, and they've confirmed our suspicion.
32:04But we see that being pretty disruptive from an ILS standpoint. in what way I wouldn't dare to kind of presume, but it's coming in much the same way they've upended travel. So - Well, did you - Go ahead. Directly, did you do it all at once or did you stagger it all at once? Wow, that's pretty brave. Or had you already proven on some level, because I'm imagining you pitching this up, that had to have been a pretty significant presentation, especially if it was going to be ripping the band-aid versus we'll prove this out and then we'll roll it across the rest of the properties. Yeah, it's a good question.
32:53The nuance there is that we rolled out the individual property sites because we were finding clients were wanting something that was bright and shiny, they've got a new asset or they've got a new brand, they're rolling out. We rolled out the early BTR product with Christopher Todd and they had really broad brand expectations and wanted very explicit websites and all these different things. And it became apparent to us that this was not the exception, but eventually going to become the rule. And what would happen is, what's really important for us is owning the Google Business Profile orientation relative to a property and where you point prospects from a website standpoint.
33:40And what we couldn't have happen was have all of these third-party managed property sites co-opting. It used to be Google My Business. I believe it's Google Business Profiles listing, the website listing. What was really important for us was kind of maintaining the primacy of MarkTaylor.com as a private ILS. And by primacy, I'll give you hard data. You know, CoStar used to share Apartments.com's entire network, unique traffic data with us on a monthly basis. They've since stopped. But, you know, there were months where MarkTaylor.com itself generated more unique traffic than the entire apartments.com network.
34:28Now, this was 2018 era, so I won't pretend to say that's the case anymore, but it's quite an engine. And so we needed to preserve that. And if we had 80 property Google business profiles pointing to to 80 different property websites, that was gonna hurt us. And we saw the appetite for these custom sites costing anywhere from 50 to$200 ,000 that ultimately didn't perform for all the reasons I mentioned earlier. They're not built for mobile, right? They're replete with all this rich imagery and all these design kind of attributes and et cetera, but they don't load on mobile and that's problematic.
35:12We knew we needed to have a alternative that was viable. So to answer your question, we rolled out the standalone websites, those individual property websites, established them, polished them. That was about a six month process. And then in tandem with rolling out the new MarkTaylor.com, we turned off ILSS. Got it. So staged. We didn't just, I mean, it was already scary. Yeah, I bet. I mean, it was already, yeah. In fact, I remember my CEO at the time saying, you're really exposing yourself here. And this is how much I believed in it. Because I'd spent a year studying. I said, listen, if I'm wrong, we can still go back, number one.
35:58Number two, I will submit my resignation. This is how much, how strongly I believe in it. And he says, at that time, I think John, who is now our president, he was our president then, but he is, John Carlson leads Mark Taylor. He was literally the only person that believed this was possible as well. So fortunately, it's proven out. It was a little bit of a scary period there. But again, we haven't had to return to ILSs out of need, right? Out of necessity. In fact, I'll give you a data point that's helpful. We average about 200 active leads today per property in our CRM. So while we are, yes, out of the decade of abundance from a multifamily perspective that masked a lot of the poor performance from operators um you know that's no longer the case today you know which is why you're going to see operators exposed because that abundance period is over cost per posture continue to rise right um and uh you know payroll ad etc etc and all the associated performance induced costs because you don't do things well that those folks are going to start to get exposed.
37:11But for us, we've had to just be really cognizant of and really simplify it. And as long as we were seeing the key metrics that we needed to, we knew we were pointing in the right direction. And so here we are, I guess, six years later, almost, where, yeah, we're coming up on our sixth year. No, we're past it. It was in October. Yeah. So we've really been cranking. Yeah. And been successfully. Go ahead. What were you going to ask, Rick? Well, I had two questions, but maybe I should narrow it to just one because I want us to be able to hit some other topics. But if there was one lesson that you could pass on to a marketer, an owner, operator listening to this about making that switch, what would it be?
37:55Yeah, that's a good question. Well, number one, avoid indulging on really expensive sites that look good, but that don't perform to mobile requirements. That's a heavy no-no. Strip out all of the additional kind of cutesy bullshit. Allow somebody to convert after a single click. Like don't make your form feels too heavy. Most people are thumbing that, right? So just ask for their name, email address, phone number, maybe the unit type they're interested in. But really, really understand what Google cares about from an SEO accessibility performance standpoint and really build to that. Trust that you can bring in these marketing capabilities in-house.
38:50It's something that you can do. And it can be really valuable because, you know, we've experienced with it. And, you know, when you utilize third-party managers, some will turn around sites very quickly. You never know what your situation may be. You may be taking over a property in three days. If you have a three-week requirement to get a website up and all the feeds connected, that's going to be a challenge. And I know you guys do this piece well, so this certainly isn't you. but there's a lot of folks that that's that was the standard um so think about the immediacy of having somebody in-house bringing a specialist in they will pay for themselves um if you bring in the right ad strategists and the right account managers um and then really take a hard look at providers who can give you websites that um you know are mobile for first oriented and are going to be additive.
39:47And in doing so, that is going to create a new problem for you, which is answering your phones. But that's a good problem to have. So know what your needs are, know what your goals are, come in prepared. And then when you're pursuing partners in this space, unless you do it yourself, know what you're going to ask them, ask them what their success metrics or how do you justify your existence? How do you perform? Because when I first arrived, we used to have these meetings with our third party ad strategists and they come in and say, you're averaging 4 ,000 uniques per property this month. And we'd be like, oh my God, this is great.
40:31Then you'd ask the next question, if you're thoughtful and we weren't particularly at that front end period it's well what's my conversion rate of unique to a lead right because i only need 100 uniques if they're converting at a 20 rate right and so i don't need 4 000 all you know for which most of them are paid since at that time our website wasn't indexing well for um search so we were nowhere near page one of search results for organic um so if you're somebody that's listening the podcast and you're saying all right where do i start with this you know owning your footprint is really important finding a really strong third-party partner who can help reduce your dependence on ilses is critical somebody who's mobile first that understands kind of the ever shifting landscape of doing that well um who monitors and audits how the algorithms are changing from a search perspective who understand the integration of now generative in that process what that now means from an SEO and structure data, what those implications are, and really can make sure that you're always competing.
41:42Because we know it. We track our performance via Google's development tool, web.dev, and we run our sites every week through that to see what's changed. And I can tell you that things change every week. We may have a high SEO score one week. actually SEO doesn't change that much, surprisingly, nor does accessibility, but performance certainly does. So we can be in the high 70s and drop down to the low 30s suddenly, and you're like, holy shit, what just happened here? And so you want a partner who's paying attention to that and making sure that you have a reliable lead gen motion. That's really critical.
42:22I mean, that's high, high level, but that's how you want to think about it. Think about what you care about, have your goal to be creating more qualified traffic and reducing your cost per lead by extension, your cost per lease, your cost per unique. And if you do that well, you deploy a CRM that allows you the visibility into these different things, you're going to be able to understand how much you now need to spend to satisfy, say, an unforeseen event at your community where you You had 15 residents suddenly submit notice of vacates. And now you need not on top of the existing lease turnover.
42:58Now you need 15 more leases in the next six weeks. And you're like, oh, holy shit, how much do I need to spend? And what you'll see is that folks will go right or wrong. They'll make a long-term commitment to solve a short-term issue. And they'll go and sign 12 months with an ILS or they'll go do. And it's, you know, it's not good. What you want to be able to do is go, okay, how many leases do I need? 15? Okay, let's back into this. Do I know what my tour to lease conversion rate is? Do I know what my lead to tour conversion rate is? Do I know what my unique to lead conversion rate is? Okay, what's my cost per unique?
43:35It's$2.25 a unique? All right, I'm going to need 1 ,000 of those that yield 50 leads. that are going to convert to a tour in a 60 % clip, which are going to convert to a lease at a 40 % clip. And you can literally be precise with your ad spend. And yes, there's other factors. If you've had turnover in parallel on your property team, you've got some competing factors that may exacerbate this, but you can really be precise with your spend. It's incredibly helpful, especially today where your third-party clients, your fee-based clients, they're not interested in seeing wide deviations to that agreed upon budget.
44:18And when they do see them, it invites further scrutiny as to what else are these guys missing? You don't want to have those conversations. And it's already a challenging conversation to have when you have to tell them that we suddenly need 15 more leases, so we need to spend X. So you want to make sure that you're reinforcing buying decision with them as well, that they've partner with a thoughtful organization that has a real understanding of what it's going to take to not only navigate the good times, but navigate the contours of the challenging times, which are inevitable. Well, Dustin, I think that's going to be a good segue into my next topic, which I think will be what we have time for today.
45:02But what I've heard that third parties struggle with, with, um, whether it's central is I'll just say centralization, but we can stick on, um, the ILS portion if you want. But when, you know, when you guys went to like, I read the ILS, as you mentioned, alluded to, you hired some in-house people to also help out here. And what I've seen third parties struggle with is like, well, shoot a lot of times that the, that budget is considered part of corporate, uh, headcount and I don't get to charge it back to the property. And then And therefore, they are actually cutting their margins while they're saving the owner the ad spend that you're talking about.
45:40So how did you guys manage through that where you're saving the owner all this ad spend, but now you had to bring on people, which dollar for dollar, it works out in your favor, as you said. But I've just seen other companies struggle with that. Yeah. So let me be clear. We have providers for our websites, right? Our partner union. But we have, and so, and this was probably should have been part of my answer to the previous question, but you want a pretty robust content management system, CMS on the back end, so that you can manage all of the various aspects of running a, you know, a potent web footprint.
46:20different. From a cost standpoint, you have these fixed costs around your monthly website maintenance fee, your SEO fee, your ad strategy fee, and there's VIGs that are associated with all of that. These are fixed costs. And by the way, they're worth it if you have a strong partner. It's absolutely a peace of mind play. At the end of the day, as an operator, owner, you want peace of mind. For us, by creating this marketing motion internally that offset our reliance on third party, again, our websites are still a product of a partner, but we own the management of those because we want immediacy.
47:06We can get things done. We can place a concession. We can do these different things, update pricing, whatever it may be, feature a whatever needs to be done. And so you what you do is you you paint the picture from a dollars and cents standpoint. These the folks on the other side are our asset managers, which is another way of saying detectives. They are truly trying to uncover wasted spend. And so if you can say, while maintaining current parity, from a performance standpoint, we're going to take these fixed costs, as I enumerated, and we're going to be able to strip out a portion of that cost while maintaining parity with today's performance and perhaps even improving it.
47:54That's a strong value story that helps move you forward. And for us, it's really important, you know, as managers, if you're not going to be paying a management fee to a third party ad stress or whatever, we aren't able to allow you to abandon that in full. We certainly have a management piece that represents our monetization of this so that we can employ these folks to do this and do this well. But the sum total picture is still very favorable for our clients. As I mentioned, they look at marketing costs per unit on an annualized basis, which can veer from$300 to$700 a unit for a community. And some of our bigger properties that are$100 or less.
48:48And so when you tell that story, that's the story they care about. How you've engineered it or the construct thereof is less important. It's that you're able to do two things, carve out costs and improve performance simultaneously. That's your winning ticket. And it can be done, I promise you. There's no, Phoenix is as competitive a landscape as anywhere right now. You know, most folks that are in this space know that in migration to Phoenix is top three in the country. We've got all kinds of, we've got all kinds of new product coming online, 34 ,000 units this year and next year, for which we've only got about 12 ,000 units of demand.
49:28And it's going to be a very challenging marketplace to be successful in, specifically with new product in Lisa, because it's going to be a race to the bottom from a price standpoint if you're not thoughtful about your approach. So finding savings, especially now, is incredibly important for your clients. And if you can do that while improving performance, you have the winning equation. Got it. So on this entire effort, through your four chapters, you talked a lot about, I'll just again put the overarching centralization, whether from marketing or lead handling or what have you. How much of that does Mark Taylor get to benefit from?
50:09Meaning, it improves your guys' margin versus how much is that more of a marketing strategy, as you said, when you're pitching new clients, that it's like, hey, this is how we're different so that they're not just hearing the same pitch from someone else? In both environments, they're very they resonate. The upside for us is that if we provide the framework for this instead of, because otherwise you're a client, you're trying to build this community, you're also trying to figure out your brand. And so we have a full consultancy now where we've extended our centralized services for competitors.
50:50We're answering their phones for them. We do brand creation. We do market studies. we do all of these different things now in the space. And what's important for us from a client perspective is being able to say, hey, brand new brand, Mark Taylor's established, let's borrow from the trust and the confidence and the awareness around this brand to quickly elevate your brand in partnership, that halo effect, that knock on effect. and we'll direct flights of ad capital, of ad spend to both digital footprint. So both marktaylor.com and your property website so that, you know, there's a lot of merchant builders in our space and there's value for them from a brand standpoint.
51:37Although oftentimes a transaction yields a brand change that's written into agreements, but very helpful for a client who's looking to lease up quickly and transact to be able to say, Hey, look at this digital footprint we have that yields x y and it has proven to be very valuable that's important it's also important for clients to see that they're pushing traffic to their specific product versus just the mark taylor brand now it's important to know that we aren't pushing we don't um direct ad traffic to the mark taylor home page it's very targeted to a specific property right and so we have a it's not sophisticated but we have a a an algorithm ourselves for determining based on occupancy and what's something called long-term netters 60-day forward look absent any new leases what would your occupancy be that's a bit of our canary in the cage in the cave scenario we're able to um uh direct ad spend to specific properties within the mark taylor.com domain so they're not encountering Mark Taylor, they're encountering the property and its brand.
52:49Right? And this is a question that comes up because if you're a thoughtful client or owner, you're like, Hey, I love you. I love Mark Taylor, but why am I spending to generate awareness and traffic to Mark Taylor when I'd rather allocate it all to my property website? And so we talk them through our blended ad strategy approach. We protect brand with our, and we're hyperlocal with our standalones, our property websites when we're national we have different blend a blended approach that dovetails well so we're not competing um in terms of driving up ad costs because we've got two property two property websites for the same property that are that are bidding for the same search queries so um we've we've tried to be very thoughtful there and we go a long way to allaying um their concerns because you know there are a lot of thoughtful clients out there who will go And like I said, they're detectives, which is great because if they're detecting good things, that's only serving you well, right?
53:51But who will ask those questions? And we certainly have an answer for that. Got it. Cool. Well, Reid, why don't you bring us home?
54:02Well, I have a ton more questions. Let me think what I should bring us home with. uh you got a hot takes yeah uh david was saying hot takes i feel like you got you're loaded with those which is just more the contrarian viewpoint how do you normalize the picture related to performance and this is probably not the best question to ask to take us home but what i mean by that is you have all these metrics you said it's really important which i was going to say maybe that will be the biggest lesson is is which metrics to actually pay attention to because that's That's one of the hardest things in digital marketing and advertising, website analytics.
54:42There's so much data. When you're testing new strategies, and we've talked a lot about the move away from the ILSs, that's obviously a big one. But I imagine you mentioned the conversion rate, the call data. Is there a paradigm for you, Dustin or Mark Taylor, in how you guys go about trying to to normalize that picture before you test? Or is it more kind of project by project and how you go about the data analysis? Again, not maybe a great one to end on, but I'm always fascinated when you kind of come across, I'll say change agents that are really looking at new ways of doing things and trying to prove it out through data, how they handle that part of the equation.
55:31Because a lot of times it's not easy with all the noise going on in the market, all the different technology, different strategies that are in motion. How do you really distill that down and feel like you're getting an honest look at what's driving results and what's not? Yeah. That's actually a really great question. Look, when I started in 17, SEO wasn't a table side conversation. You would have asked what SEO means to folks here, decision makers here and nobody, you would have gotten, you know, a hundred different answers. The important piece is to maintain focus on the strategic objectives for the organization.
56:12For us, you know, we've gone to great lengths to simplify it. We have three core objectives, right? People, excellence, and growth. And that's the basis for all of our decision making. So, you know, partner to all of this change, and there is a lot. You have to be able to append that to the key and known strategic objectives so that folks have context for it. Remember, you take it from a centralized standpoint, you're taking generalists who are asked to do a myriad of different things. And for us, disproportionately now we have more Gen Z than we even have millennial on print. These are folks who want to explicitly what is asked of them.
56:54what's next for me? What are your plans for me? Like they want to know that they're doing well. They actually dispositionally are tuned to being specialists. They can derive more happiness from the role instead of being asked to do a million different things on any given day on-prem. So for us, the change piece was absolutely tethered to the success metrics that mattered. But in some cases, you're going to be introducing success metrics that nobody's even heard of for folks. So it was really important for us to paint a broad kind of future cast picture. So here's what good can look like for us if we do the following in service of our three objectives, whether it's people, excellence, or growth.
57:37If we want to get here in the timeframe that we've stipulated, we're going to have to do the following series of things. You have to be thoughtful about it. You have to map it on high level at a high level while also being abundantly communicative about the fact that, hey, there's going to be some learnings and deviations. This is theoretical today, but we're starting to see that this can happen because we've done, we've run this preview or we've trialed this or we've done this and here's what we've seen back. We feel it is a statistically sizable enough data set or sample set to really paint a this is what good can look like for us picture.
58:18And then you start the education process. Heavy communication on the front end is really, really important. Emphasis around each individual contributor's role within that process so they feel a part of it. A lot of feedback. I'm not saying a lot is kind of nebulous. Just, you know, guys, this is right. So we've got this objective. we are moving in this direction, which, which means the following is going to happen impacting you in the following ways, which is great because it's going to allow you to do this, this and this, um, or it's going to expose you potentially if you're not doing those things well, but really, um, you know, painting the picture as you're kind of future casting and trying to peek around the corner and allowing folks to, um, uh, to understand and buy into what the vision is, is really important so that they know what their role is.
59:17They can see it. They feel that they understand their importance in it. And they also feel comfortable raising any kind of questions or feel like there's this feedback channel where they're not gonna be judged. Nobody wants to fail in full view, especially property team members who don't get a chance to collide with the corporate staff because they're on this tribute. They don't want that. They don't want to be exposed. They don't want to have an additional opportunity to look bad. They don't want that. So you paint the picture of what good can look like. This is going to mean this for you. We're going to hand off qualified leads.
59:45You know, you're going to be in a conversation for an hour only to find that somebody is not qualified to rent a class A. And you start to tether to the things that they care about. And that becomes super helpful. But there is an educational piece, David. You know, you are going to have to explain kind of the the more nuanced aspects of what we do. And again, I've alluded to SEO, but SEO is broad. That's massive impact to our business. Review and rep has massive impact to our business. Seemingly inconsequential, not, big deal. So being able to paint that picture thoughtfully, articulately, and then one last thing, is we then arm our asset managers, we call them managing directors, with talking points, and then objection handling, because not only is it new for them, but it sure as hell is new for their clients.
1:00:35And so you tell a client, we're pulling your property off ILSs, there's a lot of red flags and warning bells that will go off. So how do I talk about this confidently with my client and not put myself in a position of vulnerability? I'm not a marketing expert. So you've got to think about all these different scenarios and arm people with the requisite information so that they're empowered to go and coherently communicate and act as an ambassador for the effort. Not only with their, their, their, their internal teams, but also with the clients. Well, this has been incredibly enlightening, Dustin.
1:01:09I've not heard, I can't remember the last guest we had, maybe Jay from CoStar who had as many of his KPIs like offhand. Like I just want to run through some of them for the folks that, that didn't write them down, but I love how you talked about that through all these efforts that you improved your answer rate from 51 % to 90%. I'm directionally correct on a lot of these, Dustin, so you don't have to get out the red pen. Then I love how you said, like in today's dollars, you're able to migrate that$4.5 million in ILS spend to other means that allows you to be more flexible versus getting stuck in these long-term contracts.
1:01:48And I love that you said that you've really worked on, that you guys even track now the 60 % to 70 % of the inquiries to your on-site teams or current residents versus, let's say, lead handling, which shows where their attention goes, right? That means two-thirds of their attention is on the current residents, not on placing new ones. That's right. And then obviously, you made it super focused to get your guys' reputation above 4.0, which just helps you convert more of those leads. So I just love how tight you are with all these KPIs. I'm sure next time, I'm predicting we'll have you on again, that you'll have more KPIs that you drop on us.
1:02:25So he's been great. I knew you guys would like him. You guys, after all, no, we're fighting the same fight and you guys are doing it quite well. So kudos to the digital team. Appreciate you guys having me on. This was really nice. I hope that this is appetizing for your audience. And if anybody has any questions or want to engage further, I'm happy. I'm an open book. I'm happy to talk about our centralization efforts. You can find me on LinkedIn or elsewhere. Cool. So if they want to get in touch, as you're saying, they'll just go find you on LinkedIn, mention that they heard you here, that they'd love to learn more about centralization or what have you.
1:03:01Yeah. I mean, look, we've got, we've kind of plotted our path. I'll share again. I will share explicitly what we're doing because at the end of the day, you still have to hike that mountain. That's still you that's got to do it. But we've had a ton of learnings. We've made a lot of mistakes and sometimes it's helpful just to hear you know what those are so that you can maybe you know avoid some of those in the process so feel free to reach out like I said I love to chat and I do often now a lot of folks are starting to do that reach out and kind of for a read on kind of our approach and I'm happy to have that conversation so feel free oh and I forgot a plug for you you guys also started a podcast to do to share a lot of these insights as well.
1:03:42What's that one? Yeah, Multifamily Forward. Check out our first season. It just dropped. Would love for you to, I'm on there as well, so you'll hear some of my stuff, similar stuff, but a deeper dive into the centralization piece specifically. First episode is myself. I think I'm starting to get a little oversaturated here in this marketplace. Maybe I need to just disappear, but check that out. I think we've got six episodes and we'd love to have you. This is a big family, multifamily for us. We count a lot of our competitive peers as really close friends. And, you know, there's enough for all of us.
1:04:26So we want to see folks do well. We are so pleased to see the digitals of the world come in and do really critical pieces really well. So proud of what you guys are doing. I've been obviously rooting on Karina from afar. And I knew that her addition indicated you guys were thoughtful folks. So you guys are an important piece, you know, for what we're doing and what we need to do in the future. So excited to be here. You guys have me on again. I'd love to give you an update in a couple, you know, a year or so and let you know how we're progressing from a centralized standpoint. It'll be fun. Awesome.
1:05:04Well, thanks so much, Dustin. Get down.
From the publisher
This episode is sponsored by Netsuite by Oracle, the number one cloud financial system, streamlining accounting, financial management, inventory, HR, and more.
Download NetSuite's popular KPI Checklist, designed to give you consistently excellent performance - absolutely free at NetSuite.com/EYEONAI
On episode 158 of Eye on AI, host Craig Smith dives deep into the world of AI safety, governance, and open-source dilemmas with Connor Leahy, CEO of Conjecture, an AI company specializing in AI safety.
Connor, known for his pioneering work in open-source large language models, shares his views on the monopolization of AI technology and the risks of keeping such powerful technology in the hands of a few.
The episode starts with a discussion on the dangers of centralizing AI power, reflecting on OpenAI's situation and the broader implications for AI governance. Connor draws parallels with historical examples, emphasizing the need for widespread governance and responsible AI development. He highlights the importance of creating AI architectures that are understandable and controllable, discussing the challenges in ensuring AI safety in a rapidly evolving field.
We also explore the complexities of AI ethics, touching upon the necessity of policy and regulation in shaping AI's future. We discuss the potential of AI systems, the importance of public understanding and involvement in AI governance, and the role of governments in regulating AI development.
The episode concludes with a thought-provoking reflection on the future of AI and its impact on society, economy, and politics. Connor urges the need for careful consideration and action in the face of AI's unprecedented capabilities, advocating for a more cautious approach to AI development.
Remember to leave a 5-star rating on Spotify and a review on Apple Podcasts if you enjoyed this podcast.
Stay Updated:
Craig Smith Twitter: https://twitter.com/craigss
Eye on A.I. Twitter: https://twitter.com/EyeOn_AI
(00:00) Preview
(00:25) Netsuite by Oracle
(02:42) Introducing Connor Leahy
(06:35) The Mayak Facility: A Historical Parallel
(13:39) Open Source AI: Safety and Risks
(19:31) Flaws of Self-Regulation in AI
(24:30) Connor's Policy Proposals for AI
(31:02) Implementing a Kill Switch in AI Systems
(33:39) The Role of Public Opinion and Policy in AI
(41:00) AI Agents and the Risk of Disinformation
(49:26) Survivorship Bias and AI Risks
(52:43) A Hopeful Outlook on AI and Society
(57:08) Closing Remarks and A word From Our Sponsors




