The New Moz SEO Q&A: 100K URL Migration Case Study

Should you constantly expect a traffic drop during a site/URL migration, even a momentary one?

In case you didn’t see, Moz just recently released a shiny new SEO Q&A platform for all the world to see, check out, and utilize to find out about SEO.

Originally released as a personal function for Pro members many years ago, the Q&A was opened for public– and online search engine– viewing back in 2011.

In the years since, it grew to over 60,000 posts covering every SEO subject possible, and tens of millions of page views. For a very long time, a substantial part of Moz’s natural traffic came from the Q&A.

Sadly,&however, as often happens, gradually the Q&A saw serious neglect.

• As a result: The platform collected a ton of technical financial obligation, making it nearly difficult to upgrade

• Pages filled so slowly numerous users quit completely

• Spam ended up being increasingly more common

• Small amounts tools were outdated, and couldn’t maintain

For these factors, two foreseeable things took place:

1. The Q&A became less helpful and satisfying to users

2. Over time, traffic dropped significantly

So Moz had an option: enhance the Q&A right away, or kill it.

The good news is, we choose to improve it.

Working with the fantastic team at NodeBB (highly advised, by the method), we rapidly spun up a brand-new Q&An using our existing database, but with totally contemporary innovation on the front and backend.

Why this migration was challenging

We were under extreme time restrictions. What may normally take months, we needed to accomplish in a number of weeks. This presented unique obstacles from an SEO perspective.

The biggest difficulty? Our whole URL structure required to change. (If we had more time, we could have avoided this, but it was a high-end we didn’t have.)

That suggested we needed to migrate thousands of URLs that looked like this:

The migration also consisted of all of Moz’s user profiles, which number in the hundreds of thousands. To be reasonable, most of the user profiles aren’t really indexed.

Regardless, this was a huge migration!

The other possible red flag was that the majority of the Q&A would use client-side making— not considered a finest SEO practice! We could’ve executed a service for server-side making, but again, we just didn’t have time. We were worried Google would have problem rendering the content, and this may tank our rankings (more on this later.)

How we performed the migration

To pull off this huge migration while decreasing the risk of traffic loss, we followed fundamental SEO website migration best practices, in addition to a few “special” additionals for an included boost.

1. 301 redirect mapping

To put it simply, how you implement your 301 redirects is either going to make or break your migration execution.

For us, this was in fact the easiest, most simple part of the job, as we have a great deal of experience with site migrations!.?.!!(Does anybody remember seomoz.org?)

1. We made a list of every possible URL and URL path. It’s remarkable how many URLs and patterns you may miss out on. A good crawler is important to aid with this to make sure you do not forget anything. For Moz, we had the ability to accomplish this with information from Google Analytics, Search Console, and our own Moz Pro website crawl.

2. We mapped every URL to its matching URL on the brand-new NodeBB platform. While we found lots of edge cases, this was relatively straightforward.

3. We made sure to redirect whatever through 301. This is essential due to the fact that numerous platforms and designers might utilize 302s by default. While Google has actually informed us that they pass PageRank equally through 302s and 301s, Google has also suggested that 301s are a more powerful canonicalization signal.

Speaking of canonicalization, we likewise ran crawls of the new URL structures using the NodeBB platform. In instances where we discovered URL courses that didn’t match our old patterns or we thought were extraneous, the NodeBB group was able to easily establish canonicalization patterns to avoid Google over-indexing our URLs.

2. Optimum sitemap management

An essential part of our migration strategy was sitemap management. This involved 2 actions:

1. Old URLs: We currently had sitemaps of all the old URLs in place. Significantly, we kept these sitemaps live and registered in Browse Console. This way, Google would continue to crawl the old URLs and “see” the redirects.

Typically, web designers make the mistake of removing sitemaps too early, which may cause a reduction in crawl rate by Google. This suggests it could possibly take longer for Google to process the redirects.

Sitemaps aren’t a perfect guarantee that Google will go to all your old URLs, but they do offer a hint. In fact, we still had several thousand URLs after a number of months that Google still had not checked out, even with the sitemaps in location. Regardless, without the sitemaps of the old URLs, the concern might have taken much longer.

2. New URLs: Our old sitemaps were organized into lists of 50,000 each– the optimum enabled by Google. There’s some suggestion in the SEO community that organizes URLs into smaller sized sitemaps can actually improve crawling efficiency.

Thankfully, NodeBB allowed us to construct smaller sized sitemaps by default, so that’s exactly what we did. Rather of 2-3 sitemaps with tens of thousands of URLs, we now had 130 specific XML sitemaps, usually with no more than 500 URLs each.

3. Spam + cruft&clean-up As I mentioned earlier, the old Q&A had over 60,000 private posts built up over 10 years.

Undoubtedly, a number of these posts were really poor quality. We thought both the poor quality of the posts, together with bad user experience, might be causing Google to rank us lower.

Once again, time restraints suggested we could not do a complete material pruning audit. Luckily, NodeBB came to the rescue again (this is starting to sound like an advertorial– I swear it’s not!) and ran all 60,000 posts through their spam plugin to get rid of the most obvious, low-quality offenders.

In total, we eliminated over 10,000 posts.

We did not redirect these URLs, and merely let them 404 after the migration. No one seemed to miss them.

FYI: another excellent resource on content pruning is this excellent webinar with Bernard Huang, Suganthan Mohanadasan, and Andy Chadwick.

4. Better internal linking & & user experience

Even though we were porting over the very same material and standard design, the migration provided a terrific chance to improve user experience. To accomplish this, we made two tiny tweaks to the general UX:

1. Added breadcrumbs throughout the app

2. Included extremely appropriate “related questions” in the sidebar

The old Q&A had neither of these functions. Users who landed on a concern had no options to check out other questions. As an outcome, we suffered for many years with a frustratingly high bounce rate and poor website engagement metrics.

Results: Prior to and after the migration

To be truthful, I’ve never ever seen a migration rather like this. Having actually performed numerous migrations, I did my finest to prepare everyone for the most likely situation: be prepared for a 15-30% dip in traffic for 1-3 months while Google processes all the URLs.

In fact, absolutely nothing even near to that took place.

As you can see in the chart below, we in fact saw an increase in traffic, almost starting at day one.

In fact, in the two months after the migration&, natural Google traffic to Q&A pages was up almost 19% compared with traffic to all other pages.

What triggered this immediate lift in traffic? Was it the improved sitemap coverage, the much better internal connecting, or something else?

We just don’t understand for sure, however we do have a hint.

As quickly as we released the brand-new Q&A, engagement numbers shot through the roofing system:

• Higher time on site

• Lower bounce rate

• More pages per session

Simply put, users appeared to be much happier and more engaged with the new experience.

Could the improved user engagement have assisted rankings?

Once again, we do not know. Google is rather tight-lipped about how it may or might not use user click signals for ranking functions, however we do have our suspicions.

Transferring to the future

We’re still continuing to improve the Q&A experience. Most notably, we’re working to focus on speed enhancements, particularly in light of Google’s work around Core Web Vitals.

Regardless, this was certainly a wonderful migration where we didn’t experience a traffic drop– not even for a single day!

Maybe if you greatly improve your user experience, website architecture, and SEO finest practices, migrations may actually cause a fast net win.