 2017, that is the day that Dick's Sporting Goods migrated its website. If you heard us talk, you would think that we were war veterans, our team. We talk about pages that we lost, but they're not forgotten. We would talk about those formidable experiences where we had to band together, unite together, even when spreadsheets got a little out of sorts hours before we were ready to launch. So I'd like to tell you some of our stories, not because I want to recount them so that you can hear of the chaos and courage that we had that day, but I want to tell you it because hopefully you're either in the middle of a migration, you're planning a migration, or you have gone through one yourself and you've seen the success that you can have. So let me show you what our plan looked like. Three months out, we had read all the technical audits, all the strategies, and we said, we know exactly what we're going to do. Here are the five things that will help make our migration a success. First, we want to start early and have a process. Second, we want to preserve those top traffic driving pages because that's what sends the majority of visits to our site. Third, as any good SEO would, we wanted to use one-to-one redirects. The fourth thing we planned to do was we wanted to set the expectations with our leadership that rankings would stabilize in eight to 12 weeks. We didn't really have a clear picture, but we gave ourselves enough of a buffer to make that our target goal. And then finally, we wanted to monitor performance so that we could analyze how we were actually doing at a high level. This looks like five great tips, right? Everyone's probably tweeting this out, good plan, good solid migration strategy. Wrong. Do not do these things. I will tell you what our plan actually should have looked like. We should have waited because, as any SEO will know, our work is the very last step in the process. We should have been a little more patient, not gotten in there too soon because we didn't have a complete website. Secondly, long-tail pages are what actually saved or sabotaged our traffic. We had our eye on the high-level pages, the workout clothes, the golf clubs, the running shoes of the world, but in reality, we had to go a little bit deeper because that's where we really saw the impact to our visits. Third, redirects work. We all love 301s. They're great when you're going through a site migration. No argument there. But the problem is those redirect tools, they have limits. In those limits, they'll sometimes fail you. It'll take three months for rankings to stabilize. If someone asks you how long, three months. If they're walking past you in the hall and they say, hey, how are you? Three months. Three months is what it's going to take. You write this down and you tell your leadership. That is how much time you need. And then finally, we really had to get specific with our analytics. Everything was looking great at the top level, but it's only when we started looking at individual pages that we were able to identify and fix our problems. So lesson one, have a process. Processes are good. They're good for your team. They're good for your business partners. Everyone wins when people know what to do, but wait to do the work. And why do I say that? Well, because we had a little chaos in the creation of our website. New pages, they're going to be created fast and often. We have about a 30 to 40 person team who's creating pages and it gets a little messy when things are just churned out. URLs were changing daily. Yes, daily. So not only from Monday to Tuesday, but from 9 a.m. to 3 p.m., you suddenly have new pages, new URLs, things can change and that affects your world. And then finally, we had an incomplete product catalog. So what this means is something like our yoga shoes page. We didn't know if we would have it upon migration because we didn't have any yoga shoes that were currently live in our non-prod environment. So maybe that assortment would go somewhere else and we would have to change our strategy. We weren't quite sure. So redirect mapping looks really, really messy whenever you have URLs that change every day, every week, every month. So on our old site, we had a page called Category ID 441-8838. God bless natural language URLs that came on our new platform. In one month out, we were told that that page would actually be called Running Shoes. Great, we love this. Very simple for us to understand. Then one week out, our partners came to us and said, oh, actually, we're going to change the URL. It's going to be Running Shoes-1. Being the flexible SEO team that we were, not a problem. We'll update our thousands of thousands of redirects that we have to this one page. Great, we'll update at Running Shoes-1. But then, plot twist. Four days out, we were told that that URL was now going to be called Shop Running Shoes. Now you can imagine, this is just one small example, but if this happens multiple times and you're mapping back redirects that you're getting ready to submit, it can get a little crazy. This slide, this calendar, this still gives me night terrors. Let me show you what I mean here. So January 29th, this is when we migrated our website. Everything was great. We had our early process, early plan. We wanted to get ahead of things. And so we told the team, December is when we want you to have the final site that we can go in and look at. UAT, map our redirects, do all of our SEO work. Didn't happen. Holiday got a little busy, not a problem. Again, flexible SEO team. We got this. January 5th, we said, maybe at the beginning of the year, we'll start fresh. We'll have a great site to work with. Didn't happen. All right, not a problem. Maybe January 10th, give ourselves a couple of weeks to really get things finalized. January 10th was going to be the day that we would have a final site. Didn't happen. January 18th, we thought. Didn't happen. January 19th, 20th, it did not happen. Now as all the SEOs in the room, this should panic you when you find out that we did 75% of our redirect mapping those last four days. My team will tell you, they would text their husbands, hey, I might be a little late. You know, at 6.30, I'm not coming home. Rolls around, I will see you tomorrow. I am not coming home tonight. So shout out to my great team. We're all working late hours, 90-hour weeks. But again, like the war heroes that they are, they put in those late hours to make sure that even though our business partners got us a little crazy at the end, we were still able to do those redirects. Which brings us to lesson two. Preserve the top pages but don't neglect the long tail. Long tail keywords are really, really important. Our plan, even though we wanted to keep everything good at the high level, we should have focused on those long tail keywords that actually added up to really, really big traffic. Something like sand spike rod holders, I'm not even sure what those are, but we sell them on Dick's Sporting Goods. We had the number one rank. And this little guy gets about five visits a day. There are five humans out there somewhere searching for sand spike rod holders. This did not make our cut off during migration. We saw the numbers flat line and we said, oh, we actually need little pages like this because when you have about 3,000 pages that get two, three visits a day, those pages can actually add up to a lot of volume. So focus on the long tail when you're going through a migration. Because when we looked at it, 76% of Dick's traffic, it comes from pages with less than 200 visits a day. So we were really focused on that maybe top four graphs here on the right, but we should have been focused on that last and final bar. Fun fact for you is that on Dick's site, we have 50,000 pages that get less than 10 visits a day. And so the question then becomes, how do you guide 50,000 pages through a migration? I don't know. It gets really, really hard. It's a lot of volume. We have a very large e-commerce website and it's hard to devote all the attention, especially when you're talking about a lot of pages. So lesson three, redirects. This is any good SEO we'll tell you. This is what you should be focused on when you're actually migrating from an old platform to a new platform. So they're very, very critical, but what's even more critical we found was the actual process that we use to execute these redirects. Our team submitted 120,000 redirects. This is a large e-com site. So not all of this was manual. We did use some tools to map back certain products to where they now lived on the new site, but our team did a large portion of manual mapping back, finding those running Shoes pages and mapping them back to their new equivalents. The redirects worked just fine. 301s were in place when we launched. Everything was great, but we did not test the tool limits. We worked with our business partners. They told us, not a problem. Submit all the redirects you want. They underestimated how aggressive our team would actually be in wanting to submit one-to-one redirects, and the tool itself could actually not handle the volume. What this meant is we actually had to reduce our scope 12 hours before we launched. We are Dick's Sporting Goods. We should not be reducing anything 12 hours before we actually launch. I am on the phone with my director. We are looking at a shared spreadsheet determining where we were comfortable not launching certain redirects when we flick that switch on January 29th. All of the redirects eventually went in, but upon launch, we did not have those full map backs that we wanted. We should have talked through with the team how to handle certain things as well. We assumed they would be the same from the old platform to the new platform. Do not assume anything. We should have talked through how to handle dead pages, how to handle 404 pages, how to handle parameters that were appended to our old site and how they get passed back on the new site. Specifically for something like dead pages, we had the team that was actually putting the redirects in place, and they decided that it's not really good for customers to see a dead page when they come to the new site and the new platform. This was a decision made not by a lot of people, probably maybe by one or two people, just haphazardly, hey, not a good customer experience to show dead pages that air out. Why don't we put a blanket redirect on the site and take everyone who hits an error page to the home page? Sounds reasonable, right, from a business perspective, but as an SEO, you should hate this idea. What that did is it actually hid some errors and it didn't allow those to surface for our team to then go and use error reporting and quickly fix those after we launched. So it was something that we had to have a couple meetings on and had them lift that redirect in order for us to find the errors. Talk through these things with the teams that are actually executing some of your redirects before you go live and you'll save yourself a lot of trouble. This brings us to lesson four. Rankings will fluctuate, but expect stability in three months. I'll show you exactly what this looked like. Google understood our news site structure. Very happy that they were able to crawl, index our new pages, understand our full migration when we transitioned everything over, but it took them about three months for ranks to stabilize. So if you're going through migration, set this expectation for three months not only for search engines to understand your news site, but also for you to understand your new site. Everything was completely different for our team. Processes were different. How pages got created were different. How the site handled out of stock products was different. Everything completely changed. And so we had to evaluate and redo how we did some of our processes. And it took our team about three months to get on board with everything that was changing. Of course, technical SEO, large portion of site migrations. We had everything in place. Everything was really nice and buttoned up for this portion. Canonical tags were set. We had our site maps in place. We updated our robots file. So from a technical perspective, we really felt strong here. One of the interesting things we learned about migration was actually how Google interacted with our old pages. So they kept our old pages from our old site indexed for 100 days, pretty much to the exact day after we migrated. They weren't sending any traffic to these pages. They weren't still getting ranked for appropriate results, but we still saw in Google Search Console that they kept our old URL structures ranking for 100 days until that suddenly dropped off a cliff just like that. So currently we have about 3% of our 250,000 old URLs that are still in there in that index, but Google did have a brief period of separation anxiety where they didn't want to get rid of some of the URLs that they had known and come to love from the past decade on our site. The other interesting thing we really learned is that Google had a grace period for deleted pages. They were kind to us in terms of preserving our traffic for a bit until we were able to recoup that and rebuild the page. So in this particular example, we had a page that was ranking. We replatformed. It was not carried over. Why? No one knows, not going to point any fingers, but that page did not exist on our new platform. Google still sent visits to that page for about a week after we migrated until it completely bottomed out. They're like, all right, we tried to give you a little grace period. This page doesn't exist. We're going to stop ranking it and sending visits to it. So our team actually saw this by getting really deep in detail with our reporting. We were able to rebuild that URL and Google was kind to us in that we regained our rankings. And you can see that our visits came back maybe even a little higher than we were previously. So it's a grace period. They understand when you're moving a giant site like this that things happen, especially on a very large scale like our site, and they had a very kind grace period to us where we were able to regain our visits. One of the particular ways we were able to do this with this page is because our URLs for some of our pages didn't change before and after we migrated it. If you want a number here, from the top thousand URLs that we were monitoring, 91% of them maintained the same URL before and after they migrated. Now remember I was showing you some of those natural language URLs. So we actually work with a fantastic partner called SearchDex. They're our SEO agency and they really help us contain our same authority on one URL. And so before migration we had this category ID that was one, five, two, four, nine, six, three. Very keyword friendly, you might notice. And then after migration that page was called Workout Clothes. But to the search engines, they only saw the canonical tag that was set to this page slash product slash workout clothes. And we really think that partnering with SearchDex was a key way that we were able to successfully migrate our website. All of the authority for our category pages, our family pages, everything is really consolidated and focused on this one keyword friendly URL before and after we migrated. Which brings us to our final lesson. Analyze high level reporting because everyone wants to know how you're doing. You're monitoring that very closely the first couple of days after you migrated. But don't be afraid to get really deep and really specific with your reporting. Because when we reported on things at the very high level, traffic looked great. Year-of-a-year visits were growing at the same rate. Everyone was happy. Our team celebrated first days after. You know, we didn't see this big dive. We didn't have to spend any of our reserve funds in paid search. Everything was looking good. High level's great. Everyone take a vacation. SEO team, we are done. We did it. We migrated Dex. But when we actually looked a little bit lower we found that the true impact was seen at the divisional level and at the page level. And so it takes a lot of time when you start reporting at this level. You're looking at running shoes. Then you're looking at tree stands. Then you're looking at golf balls. Then you're looking at yoga mats. And so you have to get really specific and really deep to find things like that graph where I showed Google had the grace period of our deleted traffic. You're not gonna find that on any of your reports. Only by looking at lot of page numbers year-of-a-year, month-over-month, day-over-day are you able to identify these kind of opportunities. My father has a saying that he would tell me when I was in college finishing my finals, he would say, run through the wire. Much like a marathon, our finish is a race very strong. He would encourage me and say, give all you have at the very end, don't slow down, don't stop. And so that's the same advice that I gave to the team. Guys, we can do this migration. I know it's January 28th and we're flipping the switch tomorrow morning at 5 a.m. But we can do it. Run through the wire, finish strong. Let's make sure everything is nice and buttoned up. What we didn't know is we would have to run some more. We weren't done with our race even after we hit that migration day. It did take us about three months to understand new processes. But one of the good things about this, even when we were still working those late hours, is we had the chance to educate our business partners. After we migrated, people really started to care about SEO because they saw some of those declines in their divisional reporting. And they said, hey, what do you guys do again? Keywords, pages, Google, traffic? You're the ones who send all of our visits to the website? Sure do. So now we now have a seat at the table and we're involved in a lot more campaigns and a lot more business initiatives. So one of the great things that came out of this migration was we were able to really influence people and educate them a little bit more. So if you're going through a migration or one's up coming for you, set the expectation with your teams that you're going to have to work really, really hard up until that point of migration, but then you're going to have to work really, really hard after you migrate. So we learned some very valuable lessons from our migration. We should have waited, do less early on, and wait until the very end. Then you don't have to go back and redo your work. We learned that long tail pages were really the key. They saved or sabotaged our traffic. So we love long tail keywords now. Some of our top level pages, they're always going to rank running shoes as running shoes. But what's really going to make the difference is something like pink kayaks. Not only pink kayaks, but 100 of its friends exactly like it. Redirects work, but don't forget to test the redirect tools because they do have their limits. It'll take three months, three months, for rankings to stabilize. And then you should always remember to get really specific and detailed with your reporting so you can find errors, identify them, and then quickly correct them. Thank you.