The Report — Funnel tested their AI against 3 major competitors, with 169 real renters. They won 6/6

In this episode of The Report, Alex Howe, SVP Marketing + Growth, Funnel and Tyler Christiansen, CEO, Funnel share the results of a study they commissioned to answer a simple question: how does Funnel's AI stack up against the biggest names in the industry?

In the spirit of the world's best teams competing at the highest level, they put Funnel's voice and chat AI to the test - independently, blindly, and against competitors' best examples.

Funnel won every single matchup. The more interesting story is why - and what it signals about where multifamily AI is headed.

The setup:

• An independent research firm (Userlytics) ran six blind, head-to-head studies in April and May
• 169 real American renters tested Funnel's chat and voice AI against three major competitors: two legacy property management platforms and a self-described "AI leader" in multifamily
• Renters had 30 minutes and a broad mandate: schedule a tour, ask about pet policies, and interrupt the voice AI mid-sentence to see how it responded

By the numbers:

• Funnel won all six head-to-head matchups
• Nearly 3 in 4 renters preferred Funnel's chat AI; 2 in 3 preferred Funnel's voice AI
• 81% said Funnel felt more knowledgeable in chat; 67% said the same in voice
• In chat AI, Funnel outperformed the competition by 6-27% in renter satisfaction and 7-52% in human-likeness.
• In voice AI, Funnel outperformed the competition by 5-10% in renter satisfaction and 4-27% in human-likeness.

Four themes renters mentioned often

• Warmth. Competitors were described as "abrupt," "formal," and - my personal favorite - "the audio version of a brochure."
• Transparency. Funnel answered questions about fees, breed restrictions, and community specifics directly in the conversation. Competitors sent renters to links. Renters noticed, and they didn't like it.
• The ability to actually listen. In voice, renters could interrupt Funnel, pause, change direction - and Funnel kept up. Competitors talked over them. That was the single biggest complaint in the voice study.
• Anticipation. Funnel volunteered relevant details before renters had to ask. It felt like the AI was thinking ahead, not just responding.

Read the press release here.

Follow:

• Tyler Christiansen on LinkedIn
• Alex Howe on LinkedIn

Subscribe:

• Multifamily Unpacked podcast
• Multifamily Unpacked newsletter
• Multifamily AI in Action newsletter

Episode transcript:

0:00 The Report - Funnel tested their AI against 3 major competitors, with 169 real renters. They won 6/6

Alex Howe: Welcome to Multifamily Unpacked. Quick setup before we get into today's episode of The Report. We've always believed that if you want to be the world's best, you've gotta be willing to compare yourself to the best and if needed, to find out where you fall short. So in the spirit of the world's teams competing at the highest level in this year's World Cup, we put our AI to the test against the biggest names in the industry. Spoiler alert. Funnel won. Every single matchup. Let's get into it.

0:45 Funnel wins 6/6 head-to-head matchups of leading multifamily AI tools

Alex Howe: In April and May of this year, we hired an independent research firm called Userlytics to run six blind head-to-head studies. It included 169 real American renters-none of our customers or anyone who knew what they were testing. We ran tests separately on voice AI and chat AI against three major competitors: two legacy property management platforms and a company that bills itself as the AI leader in multifamily.

None of the renters participating knew which product was whose, and the order was reversed randomly. Every renter had 30 minutes to do what renters actually do. They were given a broad mandate so the results would be authentic: schedule a tour, ask about pet policies, try to interrupt the voice AI mid-sentence.

We gave them general objectives and just let them go. The result was Funnel's AI won every single one of the six head-to-head matchups. Today, we're going to dive into what we found and what it means. Tyler, before we jump into the specific numbers, why did we decide to run this study in the first place?

Tyler Christiansen: The truth is that deep down, it was not designed to be marketing. First and foremost, our mission as a business is to create the world's best renter experience. That's the promise we make as a brand to our customers. And we recognized that any given technology-we've been building AI products since 2020 with Camden-just because you're the best at one point in time doesn't mean you are.

So we did a big overhaul in 2025. On this podcast, we talked a lot about it. We did the acquisition of LeaseHawk. We made a big splash about our partnership with Sierra. And so we really wanted to test ourselves and quantify: how do we stack up against the alternative solutions in the market?

One of the other questions we always get is, "Surely you picked really bad examples." And we genuinely didn't. I don't want to throw anyone under the bus, but we wanted to go play against the best. Alex, you and I are excited about the summer of the World Cup starting today when we're recording this.

You don't want to play against the worst teams-you want to play against the best teams. And so we literally put the very best examples we could find of our competitors' tools. These are marquee brand-name companies using our competitors' tools, and we said, "Go against that and see what we find out."

So to begin, it was a benchmarking test for ourselves designed to give us product feedback. When we went through it, we recognized, "Hey, there's a lot of valuable information here we need to share with the market." So that's why we're starting to launch this case study and releasing the information, and we're really proud that our team and the investments we've made, the partnerships we've made, and the work we've done are showing results-so our customers, at the end of the day, can deliver the best possible AI experience.

And to back it up-to put the receipts out there-for those folks that are interested, we're happy to send out the numbers, in particular the phone numbers for Funnel and the chatbot for Funnel, so consumers can test it themselves and see if they agree with us.

Experience the AI renters love for yourself, chat with Funnel's AI right now at go.funnelleasing.com/canopy-701, or try it by voice or text at 202-807-1514.

3:41 By the numbers: Funnel's AI is the best in multifamily

Alex Howe: Let's jump into the numbers. The data showed Funnel-we won all six head-to-head comparisons. Those were two each per major competitor on both voice and chat. Overall, nearly three in four renters preferred our chat AI, and exactly two in three preferred our voice AI-66%. 81% said Funnel felt more knowledgeable in chat, and 67% said the same in voice.

In chat, we outperformed competitors by six to 27 points on satisfaction and seven to 52 points on how human the AI felt. Finally, in voice, the gaps were five to 10 points in our favor on satisfaction and four to 27 on human likeness. Tyler, what struck you most when you saw some of those numbers?

Tyler Christiansen: Yeah. Let's unpack both of them. Because I do think it's one thing to say, "Hey, we're better," and it's like a steakhouse saying, "Hey, we have the best steak." It's why-what is it that actually makes it the best?

For those who have ever been to Tampa, Florida, we've got this amazing steakhouse called Bern's. They'll take you on a kitchen tour and you see why. It's all the preparation: the way they dry age, rub the steaks, all this stuff. So what are the secret ingredients of the world's best AI? The questions in there tell the story.

On knowledge: all of us as consumers, the most aggravating thing in the world-whether I'm talking to a human or an AI agent-is that I don't get the answers. I just want to know the pet policy, the school district, the office hours-whatever my question is that I didn't find on your website or that I prefer to find via AI.

Knowledge is really important because there's a complete misnomer that exists in the world that a lot of our competitors like to say: that AI learns, it's just going to know. And when you really ask that question-literally ChatGPT is not learning when I ask, "What kind of grass is outside of my office today?" It learns through these very expensive training runs that Anthropic and OpenAI are doing. It's not learning from me when I say that grass is Bahia grass. I don't know what it is.

So it is learning through substantive retraining. In the case of our customers, the knowledge-how do we ensure that the AI agent has the right knowledge? It's by empowering the world's best operators with a knowledge base that is used across their platform.

So the same answers that I'm putting in to feed the AI agent are also there for my contact center associate; they're there for the sister property associate. So I know that Arbors Green is in the Hillsborough County School District.

And that knowledge base needs to be reinforced through self-improvement. And by self-improvement, I mean: identify that you don't have the answer, flag it for the associate or for the appropriate person to add.

But there is a real-we've talked in the past, Alex, about cognitive offloading. And when you think your AI is just going to magically learn the right answers for your communities, you're going to end up with this gap where it doesn't know.

So knowledge is really important and is something that we need to continue to invest in, train, and add to the knowledge base-and in particular, put it in the same place that the humans need that knowledge so that it's going to be there.

The second: on this human concept, it is an interesting thing that you determine if AI is good by how human it is. In voice in particular, this has been a radical transformation over the last couple of years. We all know that the voice agents of 2023-they were so bad you just wanted to hit zero.

What we have seen time and time again with the new voice AI that Funnel delivers to our customers-in particular customers like Camden and BH have talked about these things on earnings calls-is that renters are choosing to stay with the AI because it is so conversational. It is so fluid.

And really, one of the biggest anecdotes I could point to there is: a lot of voice agents-the V1s that we compete against-they just talk over you. And if you try and interrupt them, if somebody's coughing and there's talking in the background, it's really bad. It's what we all understand Siri to be today.

There was news today-I don't know if you saw-that Google's going to be replacing Siri for Apple. They gave up and said, "We need help."

So really good AI agents: you can interrupt them, they can discern between your voice and your roommate's voice. And that's where it starts to feel human: it's got intellect and the ability to understand what you are asking versus what is ambient noise.

So I'm super happy that we're seeing certain case studies where up to 80% of calls are fine talking to the AI agent. Always my personal opinion: you should have an off-ramp for the demographic that doesn't want to talk to AI agents. But that's what it means when it says it's more human: it's not speaking over you. It's not giving you a five-minute answer when you ask for the pet policy. It is far more conversational and demonstrates a degree of EQ that normally we wouldn't associate with AI.

Alex Howe: Shorter pauses, gaps-things like that. I think what you're getting at is: this is an art. There's a little bit of a science behind it, of course, in terms of how the AI is developed, but it is somewhat subjective.

So when renters explain their preferences in their own words-we ask that of each of them-four consistent things kept coming up.

The first was warmth. Competitors were often described as abrupt, formal, and my personal favorite: "the audio version of a brochure," somebody said.

Second was transparency. Funnel answered questions about fees, breed restrictions, community specifics directly in the conversation, while often in chat, competitors sent renters to links and asked them to go find it elsewhere. Renters noticed that. They didn't like that.

Third, the ability to actually listen-what you just mentioned. In voice, renters could interrupt the Funnel AI. They could pause, talk to someone in the background, change direction, and Funnel's AI kept up. Where competitors' AI voice products really talked over them, paused too long, got confused as to who they were speaking to. That was the single biggest consistent complaint in the voice study.

And the fourth was what we call anticipation. Funnel volunteered relevant details before renters had to ask. It felt like the AI was thinking a little bit ahead, not just responding to them.

So we're going to play two audio clips now, so you can actually hear the difference firsthand. These were actual clips pulled from the survey itself. First, you'll hear from a competitor, and then Funnel.

9:53 Real renter conversations with Funnel's Voice AI agent versus competitors

Renter: Yes please tell me about the move-in fees.

Competitor A: Here is an overview of some of the move-in fees-

Renter: What is the…excuse me. What is the monthly…

Alex Howe: Here's how Funnel handles that same moment.

Funnel: Restrictions apply for dogs or cats. There is a $300 fee-

Renter: What is the fees for the cat to be paid monthly?

Funnel: Monthly fee for a cat is $25

10:20 Funnel's approach to AI conversationality

Alex Howe: Tyler, three of the four reasons I mentioned are about how Funnel talks to renters, not just what it knows. Was that a deliberate design choice on our end, or did it come from building specifically for multifamily and the journeys that renters go through?

Tyler Christiansen: It's a very specific design choice. Funnel never optimized first and foremost for handle rates. That's something that you hear people talking a lot more about now-often called containment rates-which again, I referenced earlier: how often is the human trying to escape?

And you can get 100% containment rate if you don't give them an option to speak to a human. So it's not really the right arbiter of how good it is at first.

However, now as we're seeing that by building the workflows-answering questions, scheduling tours, making a payment-these sort of actual workflows that live underneath the knowledge is what renters are trying to communicate with you about.

And then you layer onto that this extreme level of conversationality-whether that's in text or in voice-and you start to have genuinely pleasant experiences.

Where we do think this is going to go, though-you talk about choice. One of the other reasons we did this case study-we paid for this test, which was genuinely a pretty large investment-was to prove a thesis that we have.

We believe that over time, although today we're the best in class, everybody is going to get really good at conversationality. And the bar of what a best-in-class experience is going to go from: how well can it talk and the pauses.

We think that our AI is the best today. We think that three to four years from now, most folks will have voice agents that can speak in 37 languages and recognize a cough. That probably is not going to be the differentiator on Funnel's AI in three to five years.

What will be, however, is the configurability of where does the human get involved. And that is up to every operator to make that determination. We're not trying to force everyone into a world of autonomous buildings.

And then second: really, what are the workflows that can be done? And that's where choice and taste and depth of product ability is really important.

So yeah, it's fun to see that consumers are getting more comfortable with this, and it just opens up more and more doors for us to think about what else could we do with an AI agent as the front door, and what avenues does that open up for consumers and for operators?

12:34 Real renter feedback from the head-to-head study measuring Funnel AI versus competitors

Alex Howe: So we went through every one of these 169 renters that we spoke with. A few quotes that really stuck with me, I want to read quickly.

One said, "I prefer Funnel because she sounded more personal. I felt like I was talking to somebody on the phone-almost even like a friend."

Another said, "Funnel definitely understood the cues, understood the way I was talking, my pauses. That was much better than the first AI voice experience."

Another said, "I prefer Funnel for the transparency on fees. Gave me those fees up front, told me exactly what breeds were welcome."

And finally: "They could predict what questions I was going to ask next. Seemed very intuitive and thorough."

[Talking about] our closest competitor in voice they said: "It continues talking while you interrupt. Obviously not human-like."

So we're hearing words like friend, personal, and intuitive to describe our software. What does it mean to you that we're earning that kind of language, and why does it matter so much in multifamily specifically?

Tyler Christiansen: One of the words that our partners often focus on is trust. And again, the more that consumers choose to use the workflows available to them and highlight the value that they perceive that they're receiving from that-it's trust.

And that is our job as a technology provider: to deliver tools that our clients can really differentiate their service level. Because we all know at the end of the day, we're selling homes. It's spaces for people to live in-not widgets.

And so I think that trust and that sort of warmth that renters are communicating back to us is really important. So I'm excited, Alex, that this really validates the path we're on is the right one.

But it makes me really excited-and anxious-to start working on additional workflows. Today, for instance, we know there are tens of thousands of communications happening about mid-lease changes. And it's not something that we're solving for yet. It's something that we are deploying here at Funnel.

But at scale, we think customers have hundreds of thousands-millions-of conversations that are yet to be solved. And the vast majority of those today are not answered by a human. They're answered by a voicemail or an email two days later.

And so I think the ability to embed AI into every single workflow in multifamily and enhance the human experience-both for the renter and for the associate-has really been validated in this study.

14:47 What did Funnel learn about building a better product?

Alex Howe: Last one, Tyler. What did we learn about our own product from this that we didn't know, and what should listeners take away?

Tyler Christiansen: Yeah, I think the takeaway for me is that you have to be ahead of the curve on technology transitions. These results are not an indication of work we've done in the last two months. This is really work that we've been doing for the last five years.

We speak multifamily. The knowledge banks that we've created inside of our system, the workflows that we've created-they're years in the making. But we had to be willing midstream to say, "We can do better."

And so we really made a huge investment in 2025, as I discussed, to lean into different partnerships and bring this technology to a wider audience-and I'm really happy that's paying out.

Similarly, though, we have to be making bets and investments as operators and technologists today, looking ahead into the future. And that is going to be agentic workflows. It's going to be how you connect multiple systems to still complete a task-something that people call orchestration.

So AI agents are not done. We're going to get into an era of deployment across hundreds of thousands and millions of units over the next couple of years.

But as a technologist, seeing the fruits of investments you made in the past show up is encouraging, and it makes you want to do it again. So we're making big bets again here at Funnel about where AI goes.

It goes into workflows. It goes into orchestration. It goes into things that really aren't being done today.

But for those listening that are still thinking that their AI agents could be better, I think it's really encouraging to know that we've turned a corner-from okay to great-and I'm very proud that the Funnel team has helped deliver those great experiences.

Alex Howe: Thank you, Tyler. And as he mentioned, if anyone's interested in digging more into the numbers, hearing some of the clips specifically from the actual users that went through it, we'd be happy to provide those. More on the methodology-if you're interested-reach out to myself, Tyler, anybody on the Funnel sales team, and we will share those with you.