Mike King: Next speaker is Andrew Prince. And Andrew is the senior SEO manager at Rocket. He has 13 pairs of glasses and I’ve seen all of them on Zoom calls because SEO isn’t the only thing he needs clear vision for. Andrew has lived in four states, which means he’s optimized everything except his forwarding address. He approaches SEO like a survivor challenge, outwit the algorithm, outplay the competition, outlast the updates. So today, he’s going to be talking about Navigating a $14 million Domain Migration. Everybody make some noise for Andrew Prince.

Andrew Prince: Hello. Hello. I’m gonna start off with an apology that Elton John is not here, but you do have a Rocketman. And the crowd groans it. That’s the worst joke, promise, but I had to. So this is a really, really cool opportunity to talk about a case study of a gigantic migration that we did at Rocket. So we had a mission, and it was acquire the domain name rocket[.]com, migrate a site to it by the Super Bowl, and kind of pull an Apple. Right? Rank for the term rocket and start getting the entity to change for that term. How big of a challenge would that be? Only the fourth largest domain purchase on record behind voice[.]com, chat[.]com, and nfts[.]com. So the next thing is that it was Q3. The Super Bowl is February 9th. So a very, very short timeline. And then I’m not someone who tends to look at keyword difficulty but Rocket’s a 95. 

So not easy. Also, the old owner of Rocket[.]com was an aerospace engineering company, so completely different entity. Bianca Anderson too referenced Katy Perry going to space. So also have to compete with that and Elon Musk blowing up SpaceX rockets all the time on that SERP. 

So why in the world would we want to do this? Well, Rocket has many different brands some of which you may recognise. So as the largest retail mortgage lender, we have Rocket Mortgage. All those ads of do you have subscriptions that you don’t know if you’re paying for? That’s Rocket Money. Our founder Dan Gilbert owns the Cleveland Cavaliers. There’s Rocket Arena. All these separate brands, sites, teams that we have in our ecosystem. But the main goal of Rocket is to sell homes, get mortgages, right? Many of these brands play a role in that home ownership ecosystem. Rocket Money. Let’s clean up your financial situation. Rocket Loans. Get a debt consolidation loan and get out of debt. You have RocketCard. So hey, you can get some money towards a down payment by using this credit card. There’s HomeSearch with Rocket Homes. There’s the mortgage side. So we’re maintaining all these different brands that don’t have that much cohesion but play into this end to end homeownership experience. And our issue is that Rocket Mortgage has this gigantic brand and all these other ones we spend time working on but they’re not quite at the level of mortgage. They share one thing in common and it’s this Rocket branding. So a couple of years ago, it was actually a few months before I started, the SEO team had a meeting. They’re like, Do you want to be really cool? If you just bought rocket[.]com and built this all in one homeownership platform, wouldn’t that be great? And it was kind of that SEO pipe dream of like, what if? But when you have legal challenges, you have different tech teams, different setups. Even in Detroit we have dozens of buildings downtown and some of these different teams sit in completely separate buildings. So it was a bit of a pipe dream and something we kind of shelved. Plus, in the summer of 2023, rocket[.]com got sold from an aerospace engineering company to a defense company. So it already changed hands that summer.  

So what does it actually take to be able to make a decision like this? It takes a visionary. And this man right here, Jonathan Mildenhall, is the most visionary marketer I’ve ever had the privilege to work with. He used to be the CMO at Airbnb. He led the integrated marketing teams at Coca Cola. He’s a board member of very prominent brands. And he had a vision to make Rocket one of the most culturally significant, inclusive, and optimistic brands in all of America. It’s a pretty audacious goal. So how do you make that happen? The first thing is let’s have more cohesion across all of the different Rocket brands. So have a similar font, logo, colors. Have matches across all of them. You see it, you understand that it’s Rocket. Some of our brands kind of came off like funny. We had the Rocket Homes, like Barbie and a Kendrick commercial. People are trying to buy a home. That’s really stressful. It takes a lot of time. It’s difficult. Humour in buying a home, although it’s remembered, it’s not really what we’re trying to accomplish. So we have a new voice and tone that’s a compassionate trailblazer. And then we needed a core concept, right? You look at a Nike, an Apple, things of that nature, they have this concept of like a just do it, right? So our core creative idea of owning a home is part of the American dream. So when three simple words of own the dream, when people think of Rocket, we want them to associate it with those three words. So how are we going to get into the market? We’re marketers, so a Super Bowl ad. So at the Super Bowl, we were going to announce this own the dream core creative idea. And hey, SEO team, we want to have a rocket[.]com website up. 

So we had to get to work. What in the world are we going to put on this site in only a few months? So we thought through all these various ideas. Is it just a landing page and we get people to the different Rocket brands? Is it just take all the existing sites and throw them on different subdomains and hey, they’re on the same domain but they still operate completely separately? Do we only bring a few core brands short term for this while we think about what we need to plan for longer term? We thought through all of these approaches and had to go to our leadership and say, Here’s the risk assessment tied to each of these. Here’s data for them. Here’s visuals of it because just throwing logic on a decision this big can be really hard for visual learners. So what would a site actually look like? And then SEOs can’t just make that big of a business decision as far as what goes. So talk to legal, talk to tech. What’s actually feasible? We didn’t want to recommend something, get buy in, and then it wouldn’t even be able to have legal approval and all of that in a short timeframe. And then, of course, case studies. Who’s done it well and who’s done it maybe not so well? The hard thing is when you have something of this type of size and this level of brand, it’s kind of hard to find things at the same scale, but we did the best that we could. So this was our core recommendation. Take Rocket Homes, put it onto Rocket[.]com. So things like listing pages and locations, market reports. But we needed to protect our cash cow of Rocket Mortgage, right? We couldn’t harm that. So we left that as is but took a lot of the things that are part of the homeownership ecosystem like rates calculators, like VA loans, FHA loans, those type of pages, and put them on rocket[.]com but no indexed them. So people could get from home search to mortgage and still have that ecosystem while we protect the Rocket Mortgage brand and plan for the long term. 

And so you think of the financial side of this slide that I showed earlier. We wanted to take our biggest brand, which is Rocket Mortgage, and just go one step before it to Rocket Homes and the home ownership side because it’s the next clearest step. How do we get people to mortgage? It’s not like Rocket Money, all of a sudden you get a mortgage, right? But searching for a home, getting a mortgage makes a lot more sense to connect those two. Home search, there are some big players that absolutely dominate, right? Like they get so much traffic from an SEO perspective. They have huge brands. But figuring out how mortgage works is not the easiest thing. Many of them, like Zillow, bought a mortgage company and they’re trying to figure out we have all of this brand and website traffic. How do we figure out the actual mortgage side that makes money? The cool thing is that Rocket’s figured that out. Rocket knows how to make money through mortgage. It has to figure out home search. And so that’s where the big part of Rocket[.]com’s idea was, is let’s put some investment into the home search side of things.

So we have to get started. And I simplified some of this to seven steps that I’ll go through in this talk. But when I say it wasn’t like week one, we do this. Week two. No, it was Tuesday we do this. Thursday we do this. Which as SEOs, it’s so odd because we’re used to like just so many different things. Or what if the SERPs change? Or what if this happens? Or there’s an algorithm update. This is like a business need, not an SEO need. So we need to say, by Tuesday, we’re having all of our redirects mapped out. By Thursday, we’re having our robots.txt file because we had other teams that relied on this in a very, very short time span. So the first thing we did, and I honestly think, looking back, it was the best thing that we did in this entire process, is launch a splash page. This was the old Aerojet Rocketdyne page that existed on Rocket[.]com, so quite literally a rocket ship. So we’re in homeownership. How in the world do we change the entity of Rocket[.]com in such a short time span? The first thing to do is just seed the domain and the content and make some connections that were homeownership before we actually do the gigantic migration.

So had this header of Rocket. Again, it’s not Rocket Companies. It’s not Rocket Homes. We wanted just that term Rocket, used our branding, talked about some of the different brands in the ecosystem. We had six different links to some of the various Rocket websites and brands. What I think is cool is the last line of this is a bit of an Easter egg that gave away that core creative idea we were sharing at the Super Bowl. So the American dream is owned, not rented. And so part of that own the dream kind of hidden within that. So we had some metrics we had to track. What does success look like? I had an executive a week in after the splash page launched, and he goes, Is it working? Are we starting to get traffic for it? And I said, we have a KPI actually about losing traffic and that being a positive thing instead of gaining. We want to stop showing up for Aerojet Rocketdyne terms. We want to stop showing up for aerospace engineering, and that’s a success because it means that we’re disconnecting from what the website used to be, and that was the whole point of the splash page. We also wanted to make sure our index pages were where we wanted them, meaning one, just the splash page.

We realized when L3Harris bought the site, they redirected the home page but not dozens of others. This is one of the use cases where the GSE removal tool was actually very helpful and necessary. Knowledge Panel had changes. SERP brandings, we launched the splash page, and it still said L3Harris above rocket[.]com. That was something we needed to monitor. Favicon, logos changing. So all things to monitor. Entity and linking, this is still ongoing, and this is a big part of this timeline.

I would not recommend this long term, but again, we were wondering for the term rocket, and we had a very short amount of time to be able to make it happen. So I reached out to all of our tech teams on our various sites and said, hey, in that About section in the footer, can we just throw a link that says rocket and link it to this Flash page? And some of them were like, we can put it in the main nav or put it in a prominent place on the home page. I’m like, I actually don’t want people to click this at all. It’s literally just seeding to Google that hey, this thing exists. We want to associate it with this term and it’s part of the Rocket brands. And so very quickly, in only a handful of days, I’m like, man, this is nice. Can we do these migrations more often? Because we got this implemented across all these sites in a short time span. And then one of the most robust corporation/organization schemas that I’ve ever seen. So we put ten people, Dan Gilbert, the founder, Varun Krishna, our CEO, a lot of people that we wanted associated with the Rocket brand. And then we nested underneath that eight of the Rocket organizations, so Rocket Mortgage, Rocket Money, Rocket Loans, etc., etc., and did same ads schema with their Wikipedia pages if they’re available, their social profiles, their contact information. We essentially were trying to tell a search engine as much information on this splash page about Rocket as we could to try to differentiate it from the aerospace engineering side of things.

This was really interesting. So when we launched the splash page, Aerojet Rocketdyne originally had L3 Harris in their knowledge panel. As soon as the splash page went up, it reverted back to rocket[.]com because it used to have that site and there was such a strong connection from that company to that website. So that was something where we needed to figure out how do we separate this splash page from this knowledge panel for Aerojet Rocketdyne. There’s the three vertical dots to the right of an organic listing that you can click. And it says about this source. Gives you some really great information as far as what sources Google is pulling in and what it thinks of you. So Aerojet Rocketdyne, when we launched the splash page, was originally the source. Over time, you can see on the right it switched to Rocket, but it was still showing Aerojet Rocketdyne sources. So it’s something we continue to work through, tried to figure out how do we remove those links on there, how do we update things on those sources as much as we can and track the changes over time.

Now SEMRush doesn’t impact directly what you do in search, but some of these third party tools can give you hints as to how these entities are put together. So we threw Rocket[.]com into SEMRush and realized it thought that we were Aerojet Rocketdyne. What I found out is that SEMRush just uses Crunchbase to be able to populate these. So I went to Crunchbase, went to Aerojet Rocketdyne’s, uploaded L3 Harris as their website, and then went to Rocket Companies and made it rocket[.]com and, oh, it fixed it. Does that necessarily help you directly in search? It’s incremental. And when it comes to entities, that’s a huge thing. 

Oh, boy, this one. Main navigations on sites with a couple million pages are obviously huge to figure out internal linking. And so I asked someone, do we have a Figma file of the mock ups of what the main navigation will look like that I can leave comments on? When I got this Figma file, I went in and found all sorts of other page template designs. So I started making comments to the point where they said, I think we need to have a meeting to figure out what SEO thinks about some of this. Because again, we were changing colors, fonts, logos, Didn’t realize that meant an entire redesign and not just a rebrand for Rocket[.]com. So it got to the point where I annoyed the heck out of them and they said, we’re gonna have this war room thing where everyone that goes in there needs to have an NDA. It’s gonna be a small group of people, and an hour or two every day we’re gonna get in there and bring up a page template and talk through it, and then we need approval from five teams, product, tech, design, legal, and SEO, which is awesome to be able to have sign off privileges in that process. 

We created the URL structure. It’s been communicated upfront and documented what the standards were there. We need to make sure that mortgage got no index. 1H1, I know some of you are like, well technically you could have more of the biggest font on the page. I did a talk last year about accessibility and SEO. They go hand in hand. I’m going to get on my soapbox for a moment. I think that we have a moral obligation to leave the web better than how we found it, and especially the websites that we work on. That starts first and foremost by making it accessible for everyone, including those with disabilities, seen and unseen. So make your websites accessible first, then you can worry about SEO. And so I’m going to sit up here and talk about correct header structures, not for SEO but for accessibility. And if we all thought with that mindset and we went to teams and said, we could actually get sued because we’re not accessible, that actually gets stuff done in large organizations sometimes. But it didn’t happen, right? So we set these standards. We communicated it. Some of these templates took the H1s and turned them into H2s or the main navigation wasn’t able to be crawled and the links weren’t found. Because we set those expectations and standards and got buy in upfront, it wasn’t, hey, create an argument as to why we should fix this. No, like we got buy in and standards upfront. You didn’t meet them. So now we need to make sure that they’re fixed. On a very short time span, we were able to shift engineering resources from other teams in cases where we set standards that weren’t reached and need to be fixed.

So then with QA, a lot of times it’s like, oh, does this link work? And is this H1 and title tag what they need to be? But you need metrics tied to it, too, especially when you’re dealing with a couple million pages. So we took Rocket Homes and used those metrics as benchmarks and then also set goals. So we said, absolute worst case is we want to be the same as what Rocket Homes was. But if we have more resources, if we do things a certain way, this is what we could achieve. And so when we ran crawls and understood Core Web Vitals and various page feed metrics, things of that nature, we were able to track that against it and say, well, these pages look really pretty and they look better, but they’re slower, and here’s the potential repercussions. Do we want it to perform or do we want it to look good? And how do we balance that mix? Indexation rates, when you’re dealing with a couple million, the biggest thing you can do is have the most amount of pages in the indexes you can. And then do they convert? So we set benchmarks for all those, ran a bunch of crawls, found some things like image alt text was not updated properly for some of our home listings. We found that our robots.txt file, which we actually block everything and then have an allow list rather than allow everything and block things because we found that that helps remove a lot of just the crazy crawlers and things that come in and random parameterized URLs that end up there.

And so the PR team originally said, let’s launch rocket[.]com and do brand refreshes at the same time. What they realized is that the website would overshadow a lot of the brand refreshes, and from a business perspective, the new branding needed to take precedence, so we pushed those up a week. And I don’t know, I think like I worked four overnights in like a week and a half, so my sleep schedule was all messed up. We redid Rocket Mortgage, Rocket Loans, things of that nature with the new font, logo, colors, etc. And so if we’re building an end to end homeownership platform, we used to have this global navigation where it was like, Go to this site. Go to this site. Well, we’re trying to put all of them in one place, so we killed that. The chat was up there. We’re like, Kill that. Branding’s old. So you can see more white space, more aesthetically pleasing, rounded corners. We use the AI chatbot now, so just cleaner design. 

And then January 16, this was my Super Bowl. So it was a Super Bowl project. This was my Super Bowl. We actually launched the site. And it looks pretty, honestly. We’re going to put a bunch of spends. We’re going to run a Super Bowl commercial. As an SEO sometimes, we’re like, oh, but this is going to be really slow. Like, sure. But the homepage on a two million page website, it’s going to be mostly branded. Make sure the navigation and the footer is optimized and linked well internally. We wanted people to do home search. I think we made it abundantly obvious what you’re supposed to do when you get to this page. The one thing I’ll say is that the H1 originally was Find Your Dream Home. And I was like, with Rocket, right? With Rocket? With Rocket? We’re going to put Rocket in there? And so we got that in. I got my extra two words in there. And then two days later, on that Saturday, is when we actually put the redirects live. So we had the site up for two days without any of the redirects in place because we didn’t want to launch it, have the redirects happen, have some sort of tech issue, have to redo that, and you have all sorts of cross domain issues. So we waited two days, implemented the redirects, and it worked unbelievably well. 

So how did it work? This is such a cool thing to send to executives. They love this. Put your name next to Amazon, Walmart, Google. When we looked in SEMrush, we saw out of 10 million sites for most of the first week, we were top 10 as far as most keywords gained with the rocket[.]com website. So next to nothing on that domain, right? It was a splash page, so it didn’t get a whole lot before. First week, all of a sudden, over a million keywords, 1.4 million traffic. What’s crazy is you take the Rocket Home site and then you look at this and it was doing several multiples better. So when you think of domain migrations, usually it’s like for a period of time, you’re going to lose X%, and this is our expectation. I almost had to reset expectations that we weren’t going to do this well. I was like, hold on. They’re like, oh my gosh, we solved the problem. And I was like, Google’s kind of testing us out, figuring out how we’re going to perform. But it was really cool to see the short term performance on it. And then we had the Super Bowl. So how many of you remember the Rocket commercial? Right? I won’t sing up here. That was Mike yesterday and Busta. That’s not my talent, but country roads. And I was able to be in the office for the Super Bowl, and we’re all swaying and singing, and everyone deserves their shot. The American dream. Own the dream. We put that core creative idea out there. And then, like, as an SEO, how cool is this? Like, it’s not just like Rocket, but like Rocket[.]com, like, Hey, hey, I did that thing. Then our team is just unbelievable. Was the first ever transition from a Super Bowl commercial into a live in stadium activation where they got the whole crowd timed up to sing Country Roads when it cut back from commercial. And the TV person said, a moment of unity inspired by Rocket. Which, I mean, I had chills up here. Just when I heard that that idea was happening, I’m like, I work for a pretty damn cool company that’s able to really push the boundaries on some things. And we used some of that creative from the Super Bowl commercial as our new background. I know Rand talked earlier, and he used to do that thing on a conference stage where he’s like, Pull out your phone, search this term, scroll down, click it, and then let’s see how much time it takes to jump up. Like it works. So if we wanted to rank for Rocket, that homepage only had a handful of words, big old picture, all of that. So we ranked like, I don’t know, 6th, 8th around there throughout the country. I was monitoring the night of the Super Bowl and within three minutes I looked at 10 different locations around the US and we ranked number one for the term rocket. So if you want to rank really well for a really difficult term, just spend $10+ million on a Super Bowl commercial and you’re good to go. 

Keyword research was not going to find this. No one’s writing about rocket[.]com because they didn’t know what rocket[.]com was. Right? So we anticipated the needs of our users. They knew RocketMortgage[.]com. They know RocketMoney[.]com. They don’t know what Rocket[.]com is. They just watched this commercial and learned about what it is. So Kevin, one of our content writers, wrote this What is Rocket[.]com Learning Center article. And the night of the Super Bowl, it showed up in the AI Overviews, the featured snippet. And if you put in the ChatGPT, WhatIsRocket[.]com? It all showed up. So it was a really, really great strategy by our content team to be proactive, to think about what are people going to ask and what questions are they going to have, not like, Oh, the MSV for this is X, Y, and Z. 

Now what? This came out. Gianna posted something of, hey, have your crawl rates gone down and your server response time’s gone up? Like, yeah, ours did. How’d we know that? Because we track it with dashboards now because we set those benchmarks and those goals so we’re able to get better not just because of this project but long term at the company. And Bing, even after the Super Bowl, it was kind of embarrassing, honestly. If you searched for Rocket, it would show Aerojet Rocketdyne. And our logo wasn’t there. We went into Bing Webmaster Tools and made some updates. We reached out to Fabrice Cannell. Fabrice is the best. And got us a bump up in the amount of pages that we could submit in IndexNow. So we really turned around our Bing performance. This is a oh slide, right, where, all right, we’re at 1.4 million, and then all of a sudden it’s down to 430k the following month. But look at those keywords. It went from 1.1 million to 1.4. So we realized we don’t have a visibility issue. We have a ranking issue now. So we show up for a bunch of things. Google saw that we’re not being engaged with as well as our competitors. That’s seen in a bunch of third party tools. It’s like we’ve work to do. We need to figure out how do we make it more sticky, how do we get people to stay on the rocket site.

And thank you to the conference organizers because they put myself and then Nick Eubanks. So Nick, I’m going to throw this to you. If you can’t beat them, buy them. So I had no idea that this was going to happen. We’re working on Rocket[.]com. We’re doing all this. And then they’re like, Oh, by the way, new announcement. We’re gonna buy Redfin. So where I said that HomeSearch is trying to figure out Mortgage, and Mortgage is trying to figure out HomeSearch, rather than figuring it out, like, what if you could just buy it? So it’s supposed to close in Q3. I’m so excited to think about building that end to end home ownership ecosystem and having Redfin be a part of our team. And then a week later, we just happened to spend $9+ billion on the largest loan servicer in the United States in Mr. Cooper. This is supposed to close in Q4. When it does, Rocket will service one in six mortgages in the US. Honestly, from a regulatory standpoint, for the next few years in the US, I think we’ll see a lot more of this. So I know in SEO it’s like AI inefficiencies and automation.

I think a lot more acquisitions from larger companies to scale quickly in those problems is a great way to go. I know that I personally, dealing with a company that thinks this way, every week I sit there and I think about competitive research and keyword research and things I’ve done, and I think of, If we could just buy that company, would it solve our problems? I really like the tech of how they built this thing on this site, which is what Rocket did to buy Dictionary[.]com and Thesaurus[.]com. A lot of people don’t know that Rocket owns those. They liked some of the technology and the setup of Dictionary[.]com. Did we want to be like a dictionary site? No. We wanted some of the tech. We bought it. So acquisitions can be a really great way to be able to scale things, whether it’s behind the scenes or Redfin Mr. Cooper.

So a couple of takeaways here. Build your brand. And by build your brand, build it the right way. So we were building Rocket Money, building Rocket Loans, building Rocket Mortgage, and now it’s like building Rocket. So that works for Rocket and what we’re going towards. Getting budget to claim knowledge panels and update things is not the prettiest, most fun thing in the world. But I think sometimes in SEO, there are table stakes and being able to control your entities as much as you can or putting organization scheme on your website. Are you going get a bunch of traffic and results from that? Probably not. But it’s table stakes. You need to do it. With executives, set very clear expectations, set standards up front, get their buy in so that if things don’t go the way that it’s expected, you have documentation that they said they would do it so then you don’t have to re argue it after the fact. And then have very clear roadmaps to mitigate risk. Being able to escalate things on short timelines with migrations is very important. 

Then, like, have a good time, right? I am a Pistons fan, so I won’t say too much about beating the Knicks at the garden the other night. But I was standing outside of the Garden looking at the two big video boards that were out there and seeing all the fans. I’m like, man, this is so cool. And I turn, and on the street corner is a three story high Own The Dream Rocket banner. Like, man, that’s cool. So I’m taking pictures, and I’m sending it to the marketing team because, like, I sit on the same floor as them. We’re SEOs, and sometimes dealing with other teams, it’s like, oh, we need to do this. We need to do that. We work for some really talented people. Like, I work at an awesome company. That was my favorite career project I got to work on. It was stressful as all get out, and the timeline was insane, but now I get to be on a stage talking about it because I have those opportunities at my job. So thank you. Appreciate it.