What does it take to lead SEO for one of the most trafficked real estate sites in the U.S.? Apartments.com VP of SEO, Zach Chahalis, reveals how he drives growth across 20 million+ pages using agile sprints, SEO testing, and hyper-local content that beats AI Overviews.
How do you manage SEO for tens of millions of pages that drive renters, landlords, AND national press? Zach Chahalis—VP of SEO at Apartments.com—joins host Jon Clark to reveal the playbook behind enterprise-level search success.Â
From agile sprint planning and split-testing metadata to turning MLS feeds and Matterport 3D tours into unique, ranking content, Zach breaks down the systems that keep America’s top rental marketplace ahead of Google’s ever-shifting SERPs and AI Overviews.
🏢 In This Episode
• How Apartments.com approaches SEO across 20M+ pages
• Agile workflows: integrating daily stand-ups, PI planning & sprint demos
• Scaling hyper-local content with proprietary rental and cost-of-living data
• Structured data, title formulas & other “boring” tactics that still move the needle
• Using AI agents for efficiency (content QA, Fair Housing compliance & more)
• Mitigating traffic loss from Google’s AI Overviews and LLM crawlers
• Smart redirect rules & platform pitfalls during massive site migrations
• Negotiating dev tickets: writing stories developers want to ship
Master the fundamentals and the future of enterprise SEO with real-world tactics you can steal today.
đź‘Ť Subscribe for more actionable SEO deep dives every week
đź’¬ drop a comment telling us your biggest challenge scaling search.
đź“Ž Mentioned Links & Resources:Â
[00:00:00] Jon: What happens when you manage SEO for tens of millions of pages, and the stakes include everything from rent prices to national media coverage. Zach Chahalis has the answer. Zach is the vice president of SEO for apartments.com. If you've ever done rental research online, there's a good chance you've landed on one of the millions of pages he's responsible for, with more than a decade of experience spanning Fortune 500 brands, agency roles.
Over 75 site migrations. Zach now leads SEO strategy for one of the most visited rental marketplaces in the country. This episode unpacks what enterprise SEO actually looks like at scale. We talk about how Zach builds and executes search strategy from SEO testing and sprint planning to managing duplicate content.
He shares the playbook behind turning syndicated MLS data into unique. Location rich content and how strategic acquisitions like Matterport are giving them an edge. With immersive 3D tours and proprietary rental data, we also dig into how apartments.com is navigating AI overviews and Google search structuring content for [00:01:00] LLMs and why even now, idle tags and structured data still matter.
Zach's challenge serving renters, landlords, and Google's shifting search landscape all at once. Let's get into it.
Zach, welcome to the show. As always, I'm joined by Joe devita, my partner in crime at Moving Traffic Media as well. Zach, you're an OG on the page two podcast, the original guest way back in 2017.
[00:01:33] Zach: Well, I guess original guest, yes. I think it was the first one in the official format, but I followed Rand back in the day, so that was a little bit of a intimidating follow through.
But yeah, I think I was the first official person,
[00:01:46] Jon: so I went back and listened to the episode before we had a chance to regroup with you. Did you ever get the Supra.
[00:01:53] Zach: No. So my, my, I'm still obsessed with cars. That's my passion. I went in a different direction. The [00:02:00] older supers hit that Paul Walker Tax for anyone that's familiar with that, where Fast and Furious made them a little bit too expensive, and the new ones started looking into them.
And then I went a different route, ended up grabbing A-B-M-W-M two at one point, which was basically just a faster version of it. So
[00:02:15] Jon: recently sold that, so that was way back in 2017. You've definitely made some moves since working at Search Discovery. Spent some time at IPO rank and even co-founded the Atlanta SEO organization, which sounds interesting.
And now@apartments.com, can you give us a, I don't know, a 10,000 foot view of what your career has been like since 2017 and get us caught up to where you are today.
[00:02:38] Zach: Yeah, so I, I continue to search Discovery for a while since the last podcast was recorded, left there in mid 2020 around the pandemic time, a lot of stuff was in flux in the agency world.
I had an opportunity to go in-house for global payments and just decided to take that opportunity just to do something different. I had spent my entire career agency side, so I just wanted to see what it was like being [00:03:00] in-house. Did that for about two years, and I missed the agency side. It was. Nothing negative with the experience of global payment has.
I actually loved working with that team and that company, but you kinda get in the realm of working with one website and you just go, yeah, I'm missing a little bit of the variety. So yeah, I went to IPO rank and worked with Mike Garrett and a couple other folks there for about two years. Started doing a little bit of my own freelance consulting and whatnot in late 2023, beginning of 24, before I had the opportunity to join apartments.com and.
They were actually a client of mine about 10 years ago. I work with a lot of the folks that that brought me back in. They were familiar with me and just were looking for some additional SEO expertise to help come in and drive some strategies hard
[00:03:42] Jon: that says something about the company. If you've someone, or I guess a group of people have been there for over 10 years, it's a good, a good group to, to be part of.
Um. So you've worked now agency of course, now you've worked in-house, went back to agency. Now back to in-house. Is there one that you prefer [00:04:00] over the other?
[00:04:01] Zach: I don't think we'd be talking SEO if I didn't say. It depends. At some point it really does depend. Like it, it depends on the variety of work, and I am a.
A type of person where I love chaos more so organized chaos, but agency world has that a little bit where you're jumping from different clients and projects and enjoyed that component. It's kinda what I missed when I was at Global Payments. I, like I said, I love working with that site and that team, and we built a lot of cool stuff while I was there.
But going back in-house, or sorry, agency side, it was an opportunity for me to do a wide variety of things, work with some different clients, some cool brands at the time, but at the same time, coming back in-house here is. It's slightly different because I have about 10 core websites underneath me as well as a wide variety of kind of extending off websites.
So I almost have that in-house agency perspective where like I get to work with a bunch of different sites and it's never adult day.
[00:04:53] Jon: Yeah, that makes sense. Are there, so if you look across agency in-house, are there like standard things that [00:05:00] sort of hold true across either scenario or are they pretty significantly different than your experience?
[00:05:06] Zach: There's some overlap and similarity and certainly some differences. I think the biggest challenge, I always had agency, so I was actually getting clients to do the implementation of the things that we were talking about. I had some clients that were fan phenomenal at it, and I had a brand where they implemented everything.
I gave them out of an audit in two weeks, and I had others where you're talking two plus years for them to implement something simple. So in-house, you still deal with a little bit of that, but I feel like I've got a bit more flexibility and ability to have conversations as to why things matter with different leadership and kind of be able to justify it.
I, I guess the cool thing here is the SEO team is a part of the product organization, so we work very closely, directly with the development groups, and we even have our own development teams as a part of it. That focus specifically on just doing SEO product led items, we're able to get things done. At the same time, you still deal [00:06:00] with, there's always red tape and there's always, Hey, we wanna do X, Y, and Z.
Okay, cool. Well, we still have to figure out how we get that data or implement it, and things still take time to do so. There's definitely that overlap component, but I feel like I've got a little bit more ability to actually get things done and execute. Got it.
[00:06:17] Jon: Got it. I think one thing that. We experience as well is that challenge of deliverable implementation, and one of the things that we've started offering as part of that, I don't know, way to speed that process, is helping to create the dev tickets where we hand over technical audit, here's the actual dev tickets to go with it.
I'm curious if you. I'm sure you've had experience doing that, but do you have any tips on how to write a great dev ticket to help get that implemented quicker?
[00:06:49] Zach: Yeah, it's funny, a lot of reason why I ended up working so well with the apartments.com team back in the day, and this is when they were apartment finder.com, before they were acquired by CoStar, is we as [00:07:00] an agency, always operated in their ticketing system, creating tickets, doing full stories, a.
When you're able to help the developers understand what you're trying to accomplish and be able to talk their language, Hey, as a search engine, I'm trying to do X, Y, and Z as a user, I wanna do this and be able to have clear acceptance criteria and failure criteria, or whatever the company's calling that.
That's key because developers, they're thinking in a certain way and they're not versed in the SEO components. So they might go and look at something going, Hey, we just did. This particular thing one way last year. Now you're asking us to do it the exact opposite, and sometimes you have to be able to write and understand, hey, this is how they're going to be looking at, and here's I.
The information you need to convey for them to understand how to do it and why we care.
[00:07:53] Jon: Yeah, I think that's the key component is helping the, helping them know that you can speak their [00:08:00] language and sometimes it's less about the SEO reasons and maybe leaning into the UX reasons to help justify some of that.
The other thing that we found very successful is data that typically developers defer to well. If I do this, what's the potential uplift? So if you can just tee that sort of thing up, tends to help really well as well.
[00:08:18] Zach: And that's one thing I love here is, so we do, we operate agile methodology and we do three week sprints, but every sprint we do a sprint demo where every team that that has worked on something will run through, Hey, here's the new features we did, here's what we worked on.
But we always start that with a lookback cycle as to something that we've released. Recently, last couple months, what have you, and talk about the performance of that. Did it work? Did it not work? What did we learn? But it's also a great way to give recognition and props to all of the teams that were involved in building something, because beyond our development team, we have five other teams that may be involved in creating a story, whether it's ingesting the data, getting that [00:09:00] data piped over the analytics component, the design components.
There's a lot of folks that touch these things. So it helps to be able to show them. Hey, look what you guys did. Look at the impact that it had on these core KPIs. We're looking at.
[00:09:13] Jon: That's really smart. Yeah. Tying it back to actual performance is always a smart thing to do. I've worked in a couple agile environments where we're actually joining the daily standups and things like that, so that SEO is always as part of the process.
Are you, is there. Any phases in the agile structure that you found to be really important to be a part of, to make sure that SEO isn't forgotten about?
[00:09:38] Zach: Yeah, definitely. I, so at least having our own specific dev team that, that focuses on SEO, we do daily standups with them and we're able to talk with them about, Hey, here's an issue we're encountering, or, Hey, maybe this ticket's a little bit more complicated than we were originally thinking, or we need to figure out how we get.
This. So we're able to have those conversations daily and work through those types [00:10:00] of problems. Our development teams also implemented what we call PI planning on a quarterly basis. So at the beginning of every quarter, in fact, it's coming up, not next act is next week, but we'll go through, Hey, here's what we wanna accomplish in Q3.
Here are the major stories that we're looking at, and we'll work with the development teams to t-shirt size the items that we wanna work through. And that way they can say, all right, we need to at least think about where this is coming from. Or, oh, this is a lot bigger than you guys thought. We might have to lose this other item, but be able have those conversations and then we can work with them to plan out the sprints.
Let's say we have three to five sprints per quarter, depending on the timeline, and be able to say, okay, these are the things we can actually truly get done. And then we have to figure out any type of compromise or anything like that goes along the way.
[00:10:49] Joe: So do you start with like an annual plan? You say, here are the five things I want to do this year, and then you map it out with the DevOps team to try to see what can we do Q1, [00:11:00] what can we do q2
[00:11:01] Zach: we do, we.
Try and map things out for the full year. I will say that's the interesting thing about being part of the product organization is I'm one pod within the product organization. There are three other pods. They are a lot more forward thinking in a sense of they could probably define 80 to 90% of their roadmap Right now, for the next year, we have things we wanna accomplish, and because we're product led, we can go.
Hey, when we launched recently is adding cost of living pages and data to our website. We knew that was gonna take time, where we get the data from, how do we build it, how do we design it, and that was one, that was a couple months process what we knew we wanted to do. Of course, with SEO, there's an element of reactivity that.
Plays into it as well. So we're not as, I would say we're like 50 fifties where we are, where I like to be in terms of being proactive and always having a backlog that we can pull from. But then there's always things that we're working on, like we're actively testing [00:12:00] content, page placements, metadata, different things like that.
And we might go, oh cool. This test actually resulted in. A decent performance improvement to this KPI, let's figure out how to roll that out. That was not in our roadmap at the beginning of the year, but we might need to pivot and have it become part of the upcoming quarter or even possibly the upcoming sprint.
[00:12:19] Jon: Got it. I'd love to dig into the testing a little bit. I don't know how much you're able to share. Maybe it's a, maybe it's a test that's already been done or a little bit older, but maybe to start, are there tests in this. Hyper local SEO component that are very different than maybe what you were doing at Search discovery or I pull rank, or is it a lot of the same sort of tactics or approaches?
[00:12:44] Zach: There's certainly some overlap. If a marketplace site like apartments.com is very similar in nature to an e-commerce site. There's some differences, but a lot of what I've worked on in my career have been e-commerce sites, GameStop, michaels.com, under Armour, things like that. [00:13:00] So you can think about some similarities.
It's, and it comes down to enterprise, SEO. You're building on a certain template. You have an idea what works, and that's what I always lean on. My more inexperienced folks are the ones that are coming into it at the beginning there. A lot of it's, Hey, I've learned this, I've done this, I've tried this, and that maybe guides where the strategy goes.
So we've been able to look at it from that perspective and just we think if we do. This change on the page, it will likely it changed this KPI metric maybe, Hey, we, we played with the title tag formula. I'll use a very simple one. Play with the title tag formula. Oh, we saw a 10% bump in click through rate from doing that didn't necessarily change the ranking per se, but, oh, we have more folks that are coming into our site now because either they see the urgency or they see.
A better signal that's supporting that. It's not really any different from when I was on the agency side. I worked very closely with my clients to develop their [00:14:00] testing roadmaps and ideas and things from that realm. But we were kinda looking at things from the same perspective. Hey, you know what, if we did this, it's similar how we're looking at it here.
We just think about it more specifically to our brand where some things might apply to a certain way for a certain company.
[00:14:16] Joe: What like 20 million pages you, this is the a massive site. You probably have dozens of templates that you have to manage. How do. Build a test plan throughout the year. Do you have to work in those sprints?
Do you have to try to launch and wrap things up within the sprint cycles? Yes and
[00:14:34] Zach: no. We've been playing with our testing procedures lately and doing more of like the SEO split testing via SM tooling that can change things more on the fly. So we're able to. Get out of the sprint cycle requirement. We try and time things accordingly, but you know, we, we have an in-house analytics team.
We work with them to develop the cohorts and everything that we want to do and look at. We will build the test in the testing [00:15:00] platform that we're using. Test goes live, we'll run it for X period of time until it hits reasonable significance, and then we'll make the determination on how it performs.
Historically, that implementation function was done by the development team, so then. You wanna do some testing, you've gotta either maybe sacrifice a story or a ticket that you wanted to do that, Hey, we need to do this, but we also need to be able to implement this test. We don't do the test. We lose several weeks in terms of being able to know the result of that data.
So it's been cool to be able to work with our development teams and find a platform that. Hey, we're comfortable with running this. It doesn't have any type of impact on site load time. We're still able to implement the changes that we want to do, and that way we can use that to be more data-driven and efficient with them when we're giving them the stories.
I'm curious on
[00:15:49] Jon: the testing and changing on the fly, so how, maybe to get a little tactical, like how do you determine what you actually serve to search engines when things are changing that dynamically? For example, [00:16:00] you would want some stability, right? So they can. Determine what to, or how to re-rank or change based on what's being shown.
So how do you manage for that?
[00:16:09] Zach: Yeah. So that this is where, and I think a lot of even SEOs still struggle with the idea of SEO testing, but you're kind of choosing a cohort of pages rather than a cohort of users and let's say, from our perspective. So yeah, I think, Joe, to your point, like we, we have templates.
We have. It really only comes down to actually a, maybe a handful of templates, but let's say our search result page, it's the same template in theory that gets served across every type of geography, every type of filter, every type of property type. It's just serving a different set of data on those particular pages.
So we might look at and say, Hey, let's test something on our town homes page. We'll choose a list of town homes across the nation. Let's say it's a thousand URLs. We'll do a 50 50 split cohort. Identify what [00:17:00] URLs go into it based on traffic so that there's even traffic distribution. And then we'll implement the change, but that will be served to both the user and the search engine because we don't wanna serve them something different.
We wanna be able to understand, Hey, does this have a user impact? Does the search engine understand and respect this change? And then we just, we'll run it for until it hits significance and then determine, Hey, did this. Make a difference. Did it help our rankings? Did it make a difference on the search engine side or, and or did it make a difference on the user side?
Do we see more folks interacting with this element, this component, this metadata change? And we've done everything from I. Change elements on the page, rewrite copy, move things around, implement different approaches to the linking strategy down to something more simple. Hey, let's play with a meta description, see if it improves click through rate.
[00:17:49] Jon: Yep. Yeah, that's, I imagine that's the benefit of being in the product group because they're so accustomed to testing and being [00:18:00] okay with pushing something in. With the goal of getting a proper significance out of that test. So I imagine that has to be a benefit of sort of being positioned in that structure.
Love to dig into how you guys manage content across so many different locations and places. And I guess a, how do you. Make it unique enough to stand on its own, especially when you're getting basically probably seeds from the MLS, which are roughly the same for everyone, right? So like how do you manage that scenario where you're trying to take something that's identical and make it unique enough to stand on its own own legs as it were.
[00:18:40] Zach: Yeah, it's funny, I'm actually speaking at Pubcon later this week, and my topic is on this exact thing. It's how we approach hyperlocal content from a national perspective. So I, I think the cool thing is apartments.com is part of CoStar Group, and for those that aren't familiar with CoStar, they're a data analytics platform.
And basically the mantra is to try and [00:19:00] digitize the world's real estate. So as a part of that, we collect a lot of data. We have research teams that are going to these areas. They're taking pictures, videos, information about the particular area, and we're able to effectively tap into that to write our content.
My team oversees both SEO in terms of the traditional technical sense and all that, but also the content team. So we're able to work with them to say, Hey, let's develop content. For this layer, for this kind of level of property, type for these particular geographies, and we will go through and take the data that we have and the knowledge about a particular area and either write new content, refresh content.
We've been developing things like our local area guides for over 10 years at this point, so we will go through and regularly update those pieces of content, whether it's, hey, a stadium has changed names to more fundamental changes to the geography. And be able to do that. But then we'll also have our content [00:20:00] team create what I'll call long form supporting pieces that help our renters or landlords learn about a particular topic.
So as an example, we were looking at San Diego recently. And we wanted to just cement our performance and visibility in that area. So we revisited our copy, we revisited the neighborhoods and the guides that lived there. We developed a bunch of longer form data-driven articles with our content team. I. To help renters find what they're looking for in a particular area.
San Diego's right near the water, Hey, what are the best neighborhoods? If you want to be near the water, what are the best neighborhoods? If you are an outdoor enthusiast, one thing we always see from the multi-family perspective is the biggest thing our renters are looking for is price. Okay, well, let's help them identify.
What's the right place to live or what can they afford with their budget? So we'll create supporting content that goes into that. We also do data-driven style pages. I mentioned earlier our cost of living pages. We also [00:21:00] have things around average rents. We're able to take our internal data and curate, Hey, here's how rent is changing in this geography over time.
Here's how this quarter compares to last quarter, to this quarter last year. And that way a renter can be more. Educated on how they want to go through the rental negotiation process. Hey, I already live here. I know what my rent is. It's going up. Do I want to go find a new place to live that's cheaper, or do I want to be more educated to negotiate?
Where I am and go through that rental process there. It's even cool when I have my own team going, Hey, I learned so much working here now that I was able to negotiate my lease more effectively because I was able to look at our rental data and go, Hey, you're trying to charge me X, but the average rent is this and this place down the street's offering a concession.
Why should I stay here when I could go there? So we're able to do a lot of that from that national scale, but then scale it down to the local perspective. And that's something that I find cool. [00:22:00] But it is. It's similar to an e-comm side, like we're all getting the same product, right? So it's how do you differentiate the rest of the experience so that way someone's getting something unique out of it.
We do try and optimize our, what we call LDP or listing description pages to be different. A big thing is we recently acquired Matterport, so Matterport is essentially a 3D touring digital twin development tool. So a lot of. Agents, brokers, property management companies, they're now able to model their entire building for a particular unit and.
Create unique visuals around that, unique data around that, and that's something that the users love to see, especially since the pandemic where the majority of folks were buying or leasing or renting something UN site unseen. So now they're relying on these virtual tours. I. 3D twins video tours to be able to make a decision on where they wanna live without ever actually seeing the place.
So we [00:23:00] have a lot of that data because we now own and are able to integrate a lot of this unique content that comes in from Matterport.
[00:23:07] Joe: John and I have pitched that idea of data journalism to a number of our clients. It's very hard for them to execute because. What data do they have? What data can they share?
But it sounds like you guys can actually make news, like you have so much information coming in, you can make news about it.
[00:23:25] Zach: It, yeah, and that's actually the cool thing. One of my team members was really excited over the weekend because they wrote an article about rental statistics and it got picked up by a bunch of local TV stations, news sites, things like that, and they're able to go, Hey, cool.
I created this thing and helped a bunch of people, but I'm also. Getting this national visibility on some of these articles. So it's really cool to be able to see that. But these groups, from a digital PR perspective, they love having the unique data side of things, and that's where some companies are able to support that and some aren't.
And there's certainly things that, that we can't. [00:24:00] Share or fall into our internal or paid tooling. But when we're able to pre present this data in a way that's helpful for users, it naturally gets picked up on. And as much as the debate around links still exists in SEO, it's valuable. But I think just having the brand mentions is also key.
A lot of folks need that top of mind. Re way to remember something or kind of go, oh, these guys are definitely the authority here. Let me think about going and looking at their website, or things along that realm.
[00:24:30] Jon: Yeah, I've, I was taking down notes. There's so many questions I wanna dig into off of that. I guess maybe to start it, it would be impossible for us to not mention AI at some point.
So I'm curious, as in, in the guide structure, right? I imagine like those are a lot of the. Queries being typed into Google that are getting picked off with AI overviews. I'm curious if you guys are seeing traffic loss there and maybe a supplemental question to that is the trade off having that unique data that's getting picked up in the [00:25:00] LLMs that maybe allow you to be a citation?
Is there, I'll stop asking.
[00:25:04] Zach: No, it's a good question. From a traffic perspective, obviously can't go too deep into that one, but like we, we certainly do see an impact from AI overviews, as does everyone else. The interesting thing is like we still see solid performance from pages that existed before AI overviews even to now.
So folks are still clicking into these citations and they wanna learn more and read more, and I think it's maybe a side effect of the industry I'm in. Versus something where they can get the immediate quick answer. If I'm just looking for, Hey, what's the average rent in New York City? We appear in the AI overview.
We appear typically number one in traditional organic search with the page, but we still see plenty of folks click through to that page and then they begin their journey deeper down into the conversion funnel. Okay, cool. Now I've learned about this. I wanna see a place for rent under this threshold. Oh, well I've got a pet, so let me make sure it's pet friendly.
And they dig down into [00:26:00] this more refined experience. So we see still plenty of quality folks and traffic coming into the site. But yeah, Google, I. With AI overviews, with AI mode with, I'll say outside of Google, like the generative AI search and chat GPTs. There's certainly folks that are going there and it's impacting click-through rates on every site, not just ours, but we're able to look at, Hey, what are the AI overviews talking about?
Oh cool. We have an article on this. Let's go back and update it. It's due for some refreshing and be able to go through and say, let's structure the content slightly differently to support. What AI overuse are talking about.
[00:26:39] Jon: Yeah. Are you taking a different approach to how you're structuring any of that content based on.
What we know about LLM so far? Or is it still letting the writer write and
[00:26:49] Zach: A little bit of both. Yeah, like that particular article I was just talking about, our rental statistics, that was a case of, hey, let's think about how we organize the outline in a way that's more digestible. [00:27:00] Let's make sure structured data's in place, but we still wanna give our writers free reign to, to express themselves and.
Provide a voice and a tone and an experience that our renters wanna read. They don't wanna read something that's written for a search engine. So we wanna be able to write for that renter, but we wanna make sure it's digestible as well. So even in, we found in that case, by structuring articles in a certain way, we see higher dwell time on a particular page or engagement.
They stick around longer. They then interact with something else and go read another article. Or they go into the the SRP results experience and be able to drill down there.
[00:27:34] Jon: Got it. Okay. I was looking through the site earlier, and one of the things that I always notice is like the URL structures and for most real estate sites, right, the challenge, well, similar to e-commerce, right?
Where once you get down into the product level, you're 3, 4, 5 folders deep depending on how that taxonomy is set up. And the real estate sites that do really well tend to. [00:28:00] Pull that listing up closer to the root. So I'm curious if you were involved in any of those decisions while you were there, or was that structure already in place, or y
[00:28:10] Zach: yes and no.
Like I said, we're a network of sites and I used to work with Apartment Finder back in the day before it was acquired. So when it was acquired by CoStar Group, we were trying to figure out, hey, how do we apply this brand to the platform that already exists for apartments.com, but have it be its own unique thing.
That was a case where the URL structures on a lot of the sites that we own and work with have been that way for. 15 plus years. A lot of our sites and brands go back even to the magazine era and things that used to have booklets and grocery stores. Those URL structures maintain where they are unless we see a dramatic need for change.
I. We try and stick with what we've got, but we wanna make sure it's user friendly. We certainly had a fair number of brands that we've acquired over the years that didn't have user friendly URLs, and we're trying to figure out going, all right, what's the impact to making this [00:29:00] change? Is it something we wanna bite the bullet and do?
But in a lot of ways, we've stuck with what has been there for the longest time to retain that authority that we have.
[00:29:08] Jon: Got it. Okay. Um. I guess last year at Pubcon, you talked a little bit about migrations and your title was Don't get left out in the rain, which I thought was great. Is there, I know this is one of your specialties, so I'd love to talk a little bit about migrations in general.
There's so many different reasons for going through that process. Some of them valid, some of them, you know, maybe just not so much. Is there, I guess, any key tips that you would recommend to mitigate that impending dip once that migration occurs?
[00:29:41] Zach: That's a good question. A as always, it depends if you're doing a replatforming and you're able to retain the URL structures.
Then generally I find that you're able to retain a lot of the visibility, performance and everything like that without seeing a dramatic dip. But there's, I guess there's a big, but here is one, if you're changing URL structures, making sure that you're doing proper [00:30:00] redirect mapping, not chaining URLs from one to another.
An example, I work with a major E major brick and mortar brand that had an e-comm site that. Two or three years ago, and it seemed like every two years they decided to redesign their local pages. And every time they did that, they would redo the URL structure and just keep chaining them together. So if I went back to the one that they had six or seven years ago, I.
It went through five hops before it got the new one. Every time you're doing this, I'm still a believer that there's authority fall off every hop that you take in a redirect. And it's been proven in a lot of case studies. So if you're doing that, some of the search engines will bail on the redirect chain.
The authority's gonna fall off. But that's a big one where I just see it's a, it can be a big miss, but also a simple solve. A lot of these are. Redirect rules where you're doing one rule that covers everything. If you're typically an e-comm site, you're redirecting off of, let's say a lookup table [00:31:00] for product pages or local pages, and you have a store ID or a product id, and you can typically do one rule that hits that lookup table and just have everything operate against the that particular table.
You don't have to do tens of thousands of redirects for every single particular URL. Try and be as efficient as you can with those. More regex based rules that exist there. I've also seen a lot of folks underestimate the impact of the platform that they choose. I still see a lot of developers that love spas, so they're, they're implementing JavaScript based websites and not really thinking through the implications of that.
To a pre rendering perspective what search engines are looking at. There have been multiple studies within the last six months that say a lot of the LLMF crawlers don't process any form of JavaScript. So if you're not thinking about the impact of that, you could have a perfectly well performing site.
You could keep your UL structures and do nothing but change it to, let's say, [00:32:00] angular. Without handling the pre rendering or the JavaScript component, and you can see your rankings fall off because you're not thinking about that. It also has an impact on site load time. It has an impact on the downstream experience of everything that, that the user's touching.
So there's ways of, of just. Not necessarily seeing what's coming, if you're not familiar with the technical implications of the direction you're going,
[00:32:21] Jon: is there, I guess looking back across the many migrations that you've managed, is there ever like a, like the right time to do a migration if you're, I. If you get pulled in and say, Hey, we're thinking about doing this for X reasons, can you help us figure out the best time to do that?
Like how would you approach that sort of scenario?
[00:32:40] Zach: Yeah, I think there's three components here is one, when do you get pulled into the conversation? And I always like to work with the teams that I'm connected to and say, Hey, if you ever hear these words, you ever hear you start talking about these things, call me.
Talk to me, Hey, I'm down the hall. Let's talk about this. So even when I was agency side, I would listen for certain phrases or, [00:33:00] oh, let me ask a little bit more about that. You guys are talking about redesigning something. Let's talk about that redesign. If you get brought in too late, that's its own problem.
I've certainly been asked about that particular I. Case and I think I'm long passed by NDA, so like I used to work with with GameStop back in the day and they did a brand new site launch in 2017 and in that particular use case they were going, Hey, we are gonna do this. We haven't touched our site in a decade.
We need to do it, but when should we think about doing it? And it turned into, alright, we think this replatforming is gonna take a year to rebuild the new site, make all the changes to get everything we want. That site ended up taking about 18 months. I. They were thinking like, Hey, let's launch this at this certain time period, because we don't want to impact our top revenue performing seasons.
We now add an extra six months. We're right in the lens of, Hey, you're gonna be pushing this right around the holidays. Do we want to do that? I, you might take a hit on revenue as much as we preserve things, [00:34:00] you might experience a bump. You could offset that with additional paid. Or other resources, but do you wanna do that?
And we had the conversation just going, Hey, let's push this back. Let's wait until maybe Q2 of the following year when flesh out our bugs, fresh address any technical debt that maybe we didn't resolve. So it was a cohesive decision where we all said, Hey, I. Let's think about how we want to do this and came to the decision to push it back.
[00:34:26] Joe: John and I don't do a lot of project work, but the, a migration project is one that we do pretty often and we the, when we write the scope, we usually start the timing with wire frames, like as soon as there are wire frames ready. That's, that can be our day one. And then the last day is a month or so after the launch.
[00:34:46] Zach: Yeah, I always looked at something pretty similar to that. I would give them initial specs, requirements, things to think about if they're doing platform selections and want folks involved in that, then I would join that. But yeah, I mean it [00:35:00] certainly, I would always go through the process of, of wireframe review, design, review.
I would use that to craft the requirements and say, Hey, look, we're looking at this page template. Here's how your heading structure should be organized on that page. Or, oh, you're missing an H one. Why and be able to have those types of conversations. But yeah, I would typically agree. That's where I like to start things off, but I at least like to ga give folks, Hey, here's a checklist of things to think about or talk about, or just some high level requirements that you need to think about as you're going into this process.
You wanna involve me sooner? Hey, great, happy to talk about that, but otherwise you should involve me here. And being able to be very transparent like that with clients is key. And you wanna make sure you get pulled in at that right time.
[00:35:41] Jon: I love the idea of getting involved in the actual platform selection for all the reasons that you mentioned, right?
There's so many platforms that are heavy JavaScript and just present unnecessary challenges. So I'm curious if there is a platforms, probably a tough question, but is there a platform you [00:36:00] prefer looking across the landscape? In, in your experience?
[00:36:04] Zach: I think it comes down to the site's purpose. A lot of e-comm sites I've worked with in the last five to seven years have migrated to Salesforce Commerce Cloud at the higher end or something like a Shopify in the more kind of mid-tier, lower tier type of size site.
The each of those platforms has its pros and cons. There was a good time where Salesforce Commerce Cloud was creating. Duplicate LDP URLs and you would have multiple iterations of your product pages. Or at one point they had a certain page template type where you couldn't set a canonical tag for it.
Like a lot of it would go, Hey, you guys are thinking about these platforms. Being able to read through the documentation or being familiar with the platform and go, here are the limitations that you need to think about, or Here are the things that you might need to solve for out of the box. Because a lot of these folks are typically hiring.
A third party development group, whether it's the agency that's also doing the SEO or someone [00:37:00] else, and they set specific scope and they're going, Hey, we're selling you what's in the box? This is what you get. Here's how we're setting it up. And then you start to go through that process and realize, oh well we are a multi-language site and this platform doesn't handle atrial lang tags.
Gotta figure out how to do that. That's outta scope and you're gonna get pushback from the development folks to do that. So if you can identify that earlier on, you can at least make sure the scope includes that. Or you can say, Hey, maybe this isn't the platform for us.
[00:37:28] Jon: Yeah, that's smart. I wanted to circle back to apartments.com or I guess, or.
Again, thinking about any large e-commerce site for that matter, once you get into filtered search results, you filter down so far, you end up with a bunch of software, four errors and things like that. Is that something you guys encounter on your site as well and how do you manage for that?
[00:37:47] Zach: Yeah. Software fours will appear in different cases.
So from us, from the marketplace perspective, you'll have either one a property that's gone offline. So we, for example, we have a lot of private landlords, Hey, [00:38:00] you own a. Condo, you wanna rent it out. That property can go offline. The URL typically continues to work, but might not be as prevalent. It, I'll say, it won't be prevalently linked to, it won't be included in XML site maps, but if that property does come back online, then we wanna make sure it's there.
You might, in theory end up with an orphan page, but that might be good because it might go in and out. I have the same thing with e-comm sites, where let's say they had. I work with a, a couple clothing companies, they would have winter gear. They wouldn't sell it three quarters of the year because they didn't wanna stock the inventory or deal with that, but then they would bring back the same product page later on.
But the out of stock still exists on the site, and it's necessary evil to have that type of software. Oh four, on the results experience side, you kinda have the same thing. Let's say you have that winter gear page, now you have no products to support it, but you don't wanna kill that page off. You wanna be able to use the again, so you're not starting over.
Six months from now. So I tried to encourage folks to, to keep some type of permanence there. Game Sub was a good example. Every year they would make a new Black Friday [00:39:00] page. Okay, why don't we just keep the same Black Friday page and have it be three quarters of the year. Hey, check back later for Black Friday deals, come back in November and see them e.
Exactly. Yeah. So something like that on, on the apartment side, it comes into, Hey, we wanna make sure folks are finding the right thing. They might be looking at something rural going, oh, well we don't have any listings there, so we might wanna stretch out how far the results go to support some properties going, Hey, when you're looking in this rural area, there's nothing really here, but we have these other properties nearby.
Maybe look at these. We do certainly have instances where there's no results pages. We, from a logic perspective, try to avoid linking to those where we're able to say, Hey, there's no inventory here. Let's not link to this page. If it becomes something that has inventory, then the logic will say, okay, let's reinject that back into the experience.
[00:39:55] Jon: Got it. Yeah. It's a smart way to to approach it for sure. Having those rules on the back end to [00:40:00] manage how those things are surfaced. Which I guess leads me to a question that a lot of other real estate companies deal with, which is properties, well, I guess two scenarios. One is the new properties, right?
So how do you. Ensure that those are getting indexed quickly and getting the visibility that they deserve against maybe a property that's been a while live of a while and maybe has some links pointing to it, so search engines know about it. Is there any tips on. Discovery for new listings or new rentals?
[00:40:31] Zach: Yeah, given I'm sure competitors are listening to this conversation, I'm gonna keep it a little bit high level, but there's plenty of ways out there to make sure search engines are seeing new pages, having certain linking that goes to, hey, new builds in an area or XML site. Maps structures a big thing. I see folks not using as much as they really should or understanding are.
Some of the submission APIs that exist or index. Now on the Bing side, that also serves multiple other sides. If [00:41:00] you're able to say, Hey, we have this new content, come check it out. It encourages the search engines to come see it. Like a, A big example we deal with on our side is apartment buildings are popping up all over the place.
Construction has fallen off, but over the last couple years there's been a big influx of new buildings, so. The property management companies wanna start advertising them. They wanna be able to start getting leases so when it's ready that they have, folks don't wanna move in, but sometimes they want, they don't wanna do that too soon.
They don't wanna have the page exist. So we need to be able to have certain logic that says, Hey, this is under construction. Let's not open it up yet until this certain criteria is met. And then, okay, this criteria is met. Now let's go do these things, this checklist to ensure that search engines are seeing this new page.
[00:41:45] Jon: Yeah, I've noticed a lot of real estate sites moving a lot of those links to the homepage, for example, and trying to get visibility on new construction and new listings that way. Do you have to deal with properties going away? I imagine that's [00:42:00] through some of the logic you just talked about with maybe pulling it off or out of the site maps and things like that, but.
I guess you're a little bit unique compared to homes where they sell and go off the MLS. Does the, do apartment listings stay in perpetuity or
[00:42:15] Zach: generally yes. I guess it's two angles. And even when you look at the homes or some of the sites that do more of the for sale buy type of experience, they tend to just put a page into an off market type of thing.
So, hey, this property sold well. It just stays an off market until it goes back on the market. And that way they don't have to make a new page somewhere. We were talking about before with e-comm products works same way. I find a lot of our multifamily properties, so like apartment buildings and complexes from a property management company, they generally are always there.
They're generally usually looking for an advertiser and they usually have at least a unit or two available because of the rent cycles that exist there. So those pages very infrequently have no listings available to them, but they, if we do hit that type of experience, then we'll go. [00:43:00] Hey, this property's currently not advertising, or currently doesn't have any available units.
Check out this other related property or go see what else is available from this property management company. Same thing with individual owners. Hey, you have a house that you're renting out in theory that LDP stays around. It just goes into the background. And that way when you wanna bring it back later, we can.
If it's a case where that page needs to go away, then we'll account for that. But we have the logic in the backend that will say. Cool. This property is no longer on the market. Let's not actively link to it or surface it
[00:43:30] Joe: to entice a, a big property manager to advertise with you guys. Does your content team develop the content?
Do you build a page specifically for the property? Does your content team build the pa? Do you offer that as a, is it a service or.
[00:43:46] Zach: Generally not as a service. No. We do have aspects of our brand and company that will actually create property management websites for them and handle all that type of components.
We typically work with the PMC or the [00:44:00] feed that's coming in to inherit their content, and then we try and figure out how we either supplement it or make it more unique or provide a just at least a unique experience on that, that LDP. So you're not seeing the same thing as every other brand is seeing.
But typically for the most part, that comes in through the feed, whether that be an MLS or working directly with a vendor or a feed vendor. If it's a property, if it's just like an independent IO landlord type of thing, IO being an independent operator, we have an experience in our rental manager tool where you, that's actually a good example of AI where we've said, Hey, upload some photography of your property.
Upload some additional thoughts on your property, and we'll write the description for you. You like it, you can use it. You don't like it, you can regenerate it, but we're able to say, Hey, based on what we see about this property, we're able to put together a property description for you that you can then modify.
We have fields that allow us to add additional content to pages. It's just on a case by case basis. Yeah. Um,
[00:44:58] Jon: I was gonna [00:45:00] ask if there are ways that you can share around how you guys are using ai, either for efficiency or for SEO, specifically, content creation, things like that. I'd love to hear a little bit more about how you're making things more.
AI process oriented.
[00:45:14] Zach: Yeah. Uh, be transparent. Say we're not using a ton of AI for the actual content generation. We use AI from an agent's perspective to help make our team more efficient and prosperous. Like I said, we've got all of our in-house internal data. We are able to create ways to pipe into that data and say, Hey, I wanna write an article about.
This particular geography, help me identify the top facts, how far it takes to get from here to here. What are the top things that that folks are looking for in that particular area? What are the top amenities? And we're able to use our data via AI agents to be able to do something like that. We also, we have physical editors that do exist as people and they do a killer job.
But we've also created some basic agents that will do something like, Hey, does this [00:46:00] content violate fair housing Acts? And that way we can reduce the amount of time that editor has to spend doing something like that. So we're trying to just improve the efficiency side of things there. So it's more educating and informing the process rather than pure generation.
And I'm, I've always been a believer, and this goes back to some of the conversations and podcasts I did with Mike when I was at IPO rank. There's the idea of information game, right? If everyone's using AI to generate the same content from the same sources. At that point, what's the unique value that you're providing to anyone?
So this is where I'd rather use AI to be efficient and say, all right, let's take this data, let's improve that process. But let's still provide something unique that other folks aren't automatically inheriting. Not to say they won't scrape it later and or try to and use that in their own LLM, but you know, at least in that case, we're using what's unique to us.
[00:46:54] Jon: Yeah, I think that's why we have to figure out the revenue model for publishers, right? 'cause they're, [00:47:00] they. We need them to be incentivized to continue to create that unique content. Otherwise, everything just ends up being flight variations of the same thing. Yeah. I'm curious if there are any, I don't know, boring tactics that you would see as being still foundational in this age of ai.
I think the more we learn about L lms, it's like, well, actually they still key off the title and the meta descrip, so I'm just curious if there are things that you're. Pounding your fist on the table internally. Let's not get too crazy. Let's still focus on the boring things.
[00:47:32] Zach: Yeah, I, I spoke at SEO week a couple of weeks ago with.
And Lori and Jo Ford and y and made a good point where it's just kinda, Hey, don't try and do the fancy stuff before you do the basics. I love that you would wanna try and do this really cool thing, but your site's not even crawlable today. So you know why you're, why you're focusing on that. I'm still a, a core believer in making sure your technical foundation that's solid is your content.
Crawlable our search engine, seeing everything [00:48:00] that's there. If you're using an extensive amount of JavaScript, then you know, not only have you still struggled with. Search engine crawling, but now you're struggling with the generative engine crawling as well and the LLMs because they can't see that anyway, the one that threw me the most, as I was seeing folks going, Hey, structured data is meaningless.
Now I'm a big structured data, I'll say nerd is. That's my second biggest passion after re-platform. These migration, I'm sitting there going, no, from the search engine side, they still rely on heavily. It's great to tie together and define entities. Provided search engines information. My cheesy analogy is always like structured days is giving Google a Shakespeare novel At a kindergarten reading level, you're simplifying that contents on the page.
Then you have both Bing and Google come out and say, Hey, copilot relies quite heavily on structured data. Google, hey, Gemini looks at this too. It's, and suddenly people are like, oh, I took structured data off my site. Which is a little bit questionable. Oh, now I need to put it back on it. It's [00:49:00] stuff like that where I go, Hey, this is still there.
And even going back to the whole rail equals next in prev sunset setting from Google. Other search engines still use it. So hey, just because someone's saying they don't use that field anymore might not mean that other folks aren't using it. Make sure you're doing the research and understanding. What you care about, what search engines care about and how you can do that effectively before you try and get too fancy.
And do you know something crazy like building a new an AI agent on the site that you want to use? Cool. Well, if no one's gonna find it. Right.
[00:49:34] Jon: Imagine being that dev picking up the ticket to re-enable J schema after he just removed it a month ago.
[00:49:39] Zach: We were talking about earlier, right? Sometimes you do have to go back to a dev and say, Hey, we recommended this.
Now we have to about face. That's a case where, hey, the SEO made a poor decision. Not that the search engine changed anything. You just, oh, we don't need this anymore. And that's a little bit more awkward and a little bit more embarrassing if you have to go back in that sense versus, Hey, Google [00:50:00] said they were doing this, now they're not doing it.
Now we gotta do it again. Which they've done a fair number of times over the years. It's a little bit more straightforward and less embarrassing for you in that sense.
[00:50:10] Jon: Right, right. Well, since Zach, this has been amazing. Before we let you go, we'd like to ask a prediction question, which is, if you go to google.com in the next call to 12 months, what does that experience look like or how has it changed?
[00:50:24] Zach: I'm still on the fence of. I think it's gonna still be somewhat similar to what today. There will still be some element of traditional, I'll call 'em 10 blue links now. It could be as little as three blue links. You'll still see all the featured snippets. Google will continue to take, try and take more visibility, share of their results pages.
AI overviews will continue to be more prevalent. AI mode will eventually become, in my opinion, the possibly the default experience, especially with the agent component where you know, they're bringing that in. There was $250 a month, now it's $125 a month. [00:51:00] Eventually it's gonna be free. Probably depends on computing costs and all that type of thing.
But I could see that eventually becoming the down the line experience. I just don't know that's gonna be there in a year from now.
[00:51:11] Jon: Got it. Got it. Alright, before we wrap up, feel free to let everyone know where they can find you online.
[00:51:16] Zach: Yeah. So yeah, obviously with apartments.com, so you know, always great if you wanna just check out random real estate and look around and just see what's around you.
You can find me on twitter@zachjoha.com or just as Zach Hels and then LinkedIn add Zachary Chaus and then also in Blue Sky as well. So come check me out and like I said, I'll be speaking at various conferences throughout the year or so. Come see me, come talk to me.
[00:51:38] Jon: Sounds amazing. Thanks again for joining us on the Page two podcast and if you enjoyed the show, please remember to subscribe, rate and review.
We'll see you next time. Bye-bye.