SEO false myths, what to know on site migrations
No particularly new information, but a useful recap on some key topics that SEO should fully understand to avoid mistakes and trouble: the new episode of SEO Mythbusting on Youtube is dedicated to debunking the myths on the migration of sites, offering an overview of domain name changes, site mergers, partial migrations and more.
informazione particolarmente nuova, ma un utile recap su alcuni topic fondamentali che i SEO dovrebbero comprendere a pieno per evitare errori e guai: il nuovo episodio di SEO Mythbusting su YouTube è dedicato a sfatare i miti sulla migrazione dei siti, offrendo una panoramica anche sui cambi di nome del dominio, sulla fusione dei siti, sulle migrazioni parziali e altro ancora.
False myths on site migrations
Given the vastness of the topic, it is not surprising that this episode is also the longest ever in the series, with a duration of over 20 minutes; the guest is Glenn Gabe (Digital Marketing Consultant, G-Squared Interactive), which opens the episode telling an old experience of consulting, useful to introduce the topic.
Once, he says, “I helped a large-scale e-commerce site that had not made a redirected 301 to images, which therefore had not been transferred (we also cited it as a frequent error in our insight); so, The advice is to never forget to redirect even to images and visual content and then check if the process was successful”. For example, checking the log servers of the old domain to see if traffic has dropped, and when we notice that the crawling activity has dropped say goodbye to the site.
A topic that can be scary
It is still Gabe to point out that many site owners have a real fear to start the migration process, because they do not know sure what can happen; others instead perform the operations too quickly and without having prepared everything properly, as evidence of the complexity of the topic and the strength of the main “false myth”.
According to the SEO metropolitan legends, in fact, there will always be a drop in traffic after a change of domain name or the migration of the site, but in reality it is not so.
What really is the site migration
Martin Splitt then takes the floor and debunks this myth, explaining that first of all what is a site migration, meaning – literally – the complete transfer from one domain to another, copying practically the entire structure of the URL and the entire content, so that at the end of the process you will have a perfect copy of the old site on another domain.
This process does not always imply a decrease in traffic: to be precise, traffic begins to decrease on the old domain to resume on the new one. Overall, this does not mean that we are losing traffic, and generally doing this operation cleanly allows us to complete the work smoothly without losing anything. More critical is only a partial transfer of the site, which could lead to traffic anomalies.
The analogy with the restaurant or the food truck
To better explain the concept, Splitt launches into a colorful analogy: in his opinion, the migration of the site can be compared to the change of location of a restaurant or a food truck.
When we go to a restaurant, we look for answers to a series of questions, such as “do I feel welcome? Is the staff friendly? Is the quality of the food good? Is the price right?” , and we store the answers in our mental folder for that place – just like Google does with its signals and ranking factors for each Web page.
If a friend asks us for advice on a restaurant to try, we will probably use the signals we collected to give him the right information and recommendations – “it’s perfect if you like Asian cuisine, it’s a really very nice place, but rather expensive” and so on.
If the restaurant moves to another area and another place, we will probably have to re-evaluate some of the answers, to find out if the characteristics have remained the same or if there was some variation – to determine if it is the same restaurant or food truck that has only changed area, for example keeping the same quality cuisine and the same prices, or if something has changed.
This also applies to search engines, which have to re-evaluate what they see and find on the new domain.
New domain name and traffic anomalies
Gabe then asks to focus attention on a specific topic, the change of domain name: sometimes the process went smoothly and the site takes strength over time, but in other cases there may be anomalies, such as “a site that, three days later, fills up completely by 70%”. The expert asks if this different behavior can be based on the history of the domain, especially in cases of acquired domains and subsequent migrations.
According to Splitt, these cases are unrelated to the history of the domain, which plays a role mainly in the situations – defined as “complicated” – of sites basically used for spam purposes and then purchased and switched immediately. The advice is to take all the precautions to avoid getting into weird issues, using all the monitoring tools also in Google Search Console to check that everything has been set up properly before making the switch.
Anomalies may occur even if we make other changes during the migration, a “risky thing” because it puts Googlebot in front of difficulties of understanding and differences between two versions of the site. And this, in turn, can translate into the need for the bot to run new scans to better understand, also negatively affecting the crawl budget, especially for large sites.
Situations with acquired domains
So these particular situations may depend on various factors, but according to the Googler in principle switch to a domain that we are sure does not have pending past loads should be fine. And even if we move to a domain with a negative history, Google is aware “that domain content changes” and the Manual Action Report has a tool dedicated to the request for reconsideration of newly acquired domains. However, we need to know that depending on how the domain was previously, Google may not immediately consider the action as a migration, and this will therefore require you to do a re-crawl and a reworking of the content, and therefore more time.
In this regard, Gabe cites another example from his consulting work: there was a customer, an e-commerce site, which had a very long domain name and wanted to reduce the name to the four letters that represented the company; after finally acquiring the new domain and making the transfer, they realize that there is something wrong. They had not done the right checks and they bought the old domain of a “kind of rock band of the past” that was full of “crazy spammy links and all kinds of things”: so, in the immediate had a crash of traffic, which then settled over time.
This confirms that content can change and Google’s consideration can also change – even in cases where the domain has a spam history or has been hacked – so it only takes a little time to fix it.
How to correctly start with an acquired domain
Splitt’s advice is to be sure to clean up everything that could be problematic in advance, so as to give Google time to understand that “things have changed” and there has been a clean slate of the past.
In more detail, when you detect a domain (not only to transfer our old site) it is crucial to measure what happens through tools like Search Console and learn about the health of the domain, possibly considering the possibility of removing the content, waiting for Google to understand this intervention, eliminate negative signals and make things normalize. Only at this point should we start the domain transfer in a progressive way, so that – while Google discovers the moved pages – it begins to evaluate the contents as a new beginning for the domain.
The false myth on site fusions
The video then goes on to face another myth to debunk, related to the fusion between two sites: many, says Gabe, think simplistically that putting “one plus one gives as a result two”, but it’s not always so.
Also because – adds Splitt – combining two sites together is no longer a migration, but creating a new site from an amalgamated version of the previous: this means that Google must understand the new content, understand how they moved from the starting domains, but also whether it is in front of a completely different domain or not, because maybe only the structure of the Urls changed in the passage, or there was the transfer of some content.
Anyway, nothing in this process is as simple as a migration and Googlebot has to scan a lot of pages again. Depending on the size of the site, this could mean that it takes a long time for the search engine to have a clear view of the site, its structure and its current content, and of course much depends also on what concretely unites in this fusion and on the attention dedicated to the process.
Migrated sites, how Google recognize the new domain
Later Gabe moves the focus on another topic, that is what happens to Google when the change of domain name is properly completed and recognized, that is when a site moves from one domain to another and all redirects are active.
As a first step, Google first checks the similarities between the old and the new site, to be sure that the new site is exactly a perfect copy of the one on the old domain – which, Splitt reiterates, is the real migration of the site. When it ascertains that this is what has happened, Google will begin to forward all signals from the old domain to the new one, but the speed of this process is completed varies from site to site (going from a few days to a few weeks).
Initially you might notice an increase in the activity of crawling on the new domain, which gradually decreases when Google understands that it is a copy of the site on another space. At the same time, if everything is normal, the crawling and the signals will disappear from the old domain to move to the new one.
More clarity with the Change of address tool
Webmasters engaged in the migration process can use the Change of Address Tool to give Google additional and clearer signals about what is happening and simplify the understanding by the search engine.
It is, according to Splitt, a useful tool to give explicit indications to Google that the site has moved permanently and therefore is not a temporary change, that could then speed up the completion of the process because it gives way to Google to skip some steps, having the certainty that the transfer was voluntary and intentional.
Migration and content quality: new evaluations after the transition?
Another “myth” that Gabe had the opportunity to meet and listen to in his work is the reassessment of quality ratings by Google after a migration. In fact, Splitt explains that Google constantly re-evaluates the quality of content, regardless of whether the site has been moved, and reiterates that “if your content is now considered high quality, it does not mean that it will always be”.
This also holds true on the contrary: low-quality content or spammy could, in theory, be considered high-quality if improvements are made.
However, as far as migration is concerned, in principle if there were high-quality content on the old site that is moved identically to the new domain, then the signals will also follow.
Issue with the transfer, it is better not to go back
Another advice coming from Martin Splitt is how to deal with any problems that arise after completing the migration: a common feeling, especially when traffic drops occur and you do not recover the previous positions, leads to consider reversing the process to restore the initial status.
For Google, this is a step to never perform except as an extreme ratio, when there are no other options and all checks performed on the new site have not given outcome or explanation to the collapse.
In most cases, in fact, just check that there are no technical issues that can interfere and cause the negative effect: for example, Google may not have recognized the redirects, or the old site was crawled with little frequency and therefore takes more time to allow the bot to pick up the redirects. Other situations could be algorithmic changes occurred in the meantime, or even spammy content reporting or manual actions and so on.
If we are sure that we have done all the steps correctly and, after a month, the traffic has not improved and returned to the levels of the old site, it may be the case to ask for external help to find out the cause of the situation.
However, it is always important to compare the traffic between old and new domain: if the old site continues to have all the historical traffic, this is a sign that there has been some error in the migration process. Only in this sense, says Splitt, could it make sense to “reverse the process for a while, go back, understand what happened and then regroup”.
Migration and robots file, tips to make no mistakes
Another technical aspect discussed during the episode concerns the management of URLs blocked in the robots.txt file: Gabe asks if it makes sense to unlock resources in the transition to the new site, but Splitt is rather clear in stating that it is “not needed”.
According to the Googler, there is a reason that the site does not want those Urls scanned, so there is no reason why the crawler should analyze them by migrating to the new site.
The most common issue with site migration
The last theme analyzed concerns the most common problems that Google finds on sites just transferred: the list includes many technical variables, such as robots.txt that completely blocks the pages of the new domain, a noindex meta tag on all new content or failure to pass the Google Search Console settings on the new domain.
Yet, according to Splitt, the most frequent error is making too many changes and changes at the same time during the migration, adding “too many variables” which also make it difficult to understand the effects of work. In these cases, in fact, it becomes difficult to identify with certainty the cause of a problem, which could derive from the new structure of Urls, from the different technology used, from the new content, from the migration, from algorithmic changes of Google, from penalties and so on.
Google’s final advice is to take it one step at a time: first focus on domain transfer, then change text decks, then possibly intervene with other things. “Whatever you’re doing, do it step by step,” Splitt says, waiting for Google to do a new crawling and process the site before you put your hands on it again.
