Redirects are part of everyday life for every webmaster. With the popularity of on-page measures and the right tools, the basic knowledge of redirects and their benefits has been partially forgotten.
Some of what Google tells about redirects or various experts present is not (any longer) correct or at least incomplete. This article will explain not only the basics but also the specific applications and effects to show what indexing, On-page optimization (that is the basic idea of SEO) is concerned but also the deliberate manipulation of search results. In particular, the use of redirects for the targeted modification of results in the search engines is – rankings presented – and encouraged to try out.
What are redirects – and what are not?
One speaks of a “redirect” if one is redirected to another URL when visiting a certain page (more precisely a precise URL). “One” can be anyone here, a person in the web browser, the crawling bot of a search engine like Google, or just anyone who requests the specific URL via the HTTP protocol. There are often distinctions that are made for SEO purposes alone.
Redirects are implemented through technical instructions that the webmaster can incorporate into the webserver or the source code of a website. Then visitors to a URL are redirected to the specified destination, which can be on the same, but also a different domain.
Regardless of redirects, there have been “rel-canonicals” and other alternative instructions for a few years, which are only to be dealt with here in that they are explicitly not redirects, neither in terms of the impact on the user nor on the search engine.
The effect of redirects, but also of canonical tags, can vary greatly from an SEO perspective, depending on whether they are set on the same domain or between different domains (“cross-domain redirects”). Of course, the forwarding jump via different domains also offers some other options regarding the cookie setting, which is often used by advertising networks – but that’s a topic for a separate article.
Server-side redirects and client-side redirects
A distinction is made between server-side forwarding and client-side forwarding. This fine distinction relates to the technical way in which the browser decides to go to another URL. In the case of server-side redirects, the visitor or bot receives a response from the webserver directly in the HTTP protocol, namely, e.g., the status codes 301 or 302 or 307, because this is defined in the HTTP standard, and then loads another page – and maybe a few more cookies.
The technical solution of the redirect is usually not visible to the user, but online marketers and SEOs do have tools and advantages to analyze the implementation in detail.
Redirect types Permanent (301) and Temporary (302)
For almost a decade, the definition of redirects has been divided into two main aspects – permanent and temporary redirects. The permanent redirects are supposed to declare permanent redirection, i.e., if an original page is no longer available. In practice, this was implemented using the server-side status code “301”.
The temporary redirects are intended to indicate a “temporary” redirection and thus signal the search engine that the old content should appear again later. To date, however, it is often unclear how short or long a temporary redirect must be. In practice, the forwarding was implemented using the server-side status code “302”. Recently there is also the “307″ Redirect, which is also called temporary but can have different backgrounds and effects.
Also, client-side redirects, such as “Meta-Refresh,” were often understood as temporary redirects, although these were meant by the webmaster permanently.
If you look at the application goals of redirects in the next section, it becomes clear why misunderstandings keep occurring. It also becomes clear that Google’s representations have not explained – or want to explain – a very special area of activity for the redirects.
In the past, it was often said that 301 redirect is the best solution for permanent migration. Based on a few tests, this is very much in doubt today. Primarily this has to do with the fact that the 301 redirects previously passed on the rankings and link power as unfiltered as possible, but not other redirect types. However, this has changed, even if Google continues to recommend the “301” as the best redirect (see https://support.google.com/webmasters/answer/139066). But more on that later.
Interestingly, there is no statement anywhere on the web about the recommended duration of temporary redirects. From a technical point of view, days, weeks, months, or even years could apply here.
Redirects application objectives
There are three main application objectives for redirects:
Goal 1: Forwarding the visitor
The obvious goal is to direct the visitor to the most relevant page for them. So the visitor of a shop who has accessed the URL of the previous model of a product via an old link, the current model is displayed. To achieve this goal, it doesn’t matter which technical implementation is chosen; the main thing is that the user is redirected and gets to see the current version.
The redirection of the visitor from old URLs is important because links can exist via a variety of sources that a visitor wants to click on in the future.
In April 2016, Gary Illyes from Google Zurich briefly stated that redirects could be removed as soon as Google has crawled them.
Goal 2: Forwarding search engine bots for indexing
The search engine bots should also be forwarded to the current content as well as possible. The indexing of the most up-to-date description of a product as from the previous example should be ensured alone so that it can be found as accurately as possible via the search engine.
The bots should be informed as quickly and consistently as possible on which landing page the correct content can be found. But that doesn’t mean that they will remove the old content or the old SERP snippet just as quickly.
It was always – and should continue to be – the case with
- Permanent redirect (301) the target URL is indexed and visible in the search results, and the redirect is cached;
- Temporary redirect (302), the source URL is indexed and visible in the search results; the redirect is not cached.
So much for the theory. In July 2016, e.g., For example, you can still find old subdomain snippets from try.linkresearchtools.com in the index. However, for more than seven months, everything has been “301” forwarded to other destinations on other (sub) domains (see Figure 2). The result corresponds more to the behavior of temporary redirects (302), the SERPs shown should no longer be found after the “old school”.
Figure 1: Even seven months after a 301 redirect, pages remain in the index.
At the latest at goal number 3, it becomes interesting technically and for search engine optimizers.
Of course, previous rankings and incoming links should be forwarded. The links are not only about links from external sites that have strengthened an old URL but also about internal links that may still exist on the page.
In the rankings, we expect that the new URL will replace the position of the old URL for a specific search word in the search engine. This wish of the so-called ranking forwarding is not always fulfilled.
Which redirect does what for SEOs?
If we assume that both the simple forwarding of the visitor and that of the bot will somehow work, we as SEO are concerned with goal 3 – the link strength and the rankings.
The impact of forwarding on the transmission of link power as well as rankings has been discussed in recent years – sometimes more openly, sometimes hidden – but has recently been neglected in “SEO research“. It has been widely believed for years that “permanent 301 redirection” passes on the link power (albeit weakened) and rankings, whereas “temporary 302 redirection” stops the link power.
This is no longer the case in 2016; even Google says – at least in part. Google’s statement by Google Müller that redirects 301 and 302 are treated equally (see https://www.seroundtable.com/google-302-301-permanent-redirect-21045.html) is more confusing than helpful. Google probably only refers to the interpretation of a temporary redirect for indexing.
In countless SEO experiments by the author and LinkResearchTools (LRT), some of which had been running for months, it was not once possible to confirm that redirects 301 and 302 behave equally when used for SEO purposes (influencing the rankings).
Specifically, Google still refers to statements that (internal) page rank would be passed on with both forwarding variants and that Google would smartly consider the temporary redirects to be permanent if they only existed long enough. Unfortunately, there were no statements about the “duration” that Google would understand as “temporary”. Also, there are only confusing statements from Google regarding the newer 307 redirects, and here only for the client-side 307.
The “temporary 307 forwardings” was described by John Müller as “client-side redirects”, although the status code 307 was clearly defined in the HTTP protocol and can also be returned from web servers. The client-side representation of a 301 redirect as 307 mentioned by Müller is rather a confusing interpretation of the Chrome browser, where a permanent redirect is converted into a “new temporary redirect”.
The “new” 307 redirect – two cases on the client-side and the server-side
Although the new code 307 was defined as an improved temporary redirect by the IETF the implementation in the search engines is a seemingly different reality that can be interesting for many SEOs.
According to Google, a 307 code is generated on the client-side in Chrome when redirecting from permanent 301 forwardings via HSTS(HTTP Strict-Transport-Security) to HTTPs. The example from SEroundtable deals with this case and explains why the Chrome browser displays a 307 although Google displays 301 sees.
Figure 2 shows the view of the Google Search Console for www.seroundtable.com and a very clear 301 redirect from the HTTP to the HTTPS variant.
Figure 2: The view published by Barry Schwartz of the “Fetch as Google” of the domain seroundtable.com.
The Link Redirect Trace for Chrome shows that Chrome shows this 301 as a 307 by changing the protocol, as can be seen in Figure 3.
Figure 3: The client-side view of the “Fake-307” in the free “Link Redirect Trace” extension for Chrome.
However, there is a permanent redirect – this 307 is the “lie” in Chrome described by John Müller, which is already questionable from a technical point of view. However, the (often shared) statement that the 307 is only a client-side redirect is much more surprising, which is wrong.
There is also the option, as described in the IETF, to have the 307 status code returned by the webserver. In an SEO experiment, this then looks as shown in Figure 4.
Figure 4: The view of a “real server-side 307” in the Google Search Console.
Please note the number (1) HTTP / 1.1 status code 307 and the HTML message generated by (2) the web server that this would be a temporary redirect.
In the browser view, this server-side 307 redirect is displayed correctly, and at first glance, it is indistinguishable from the “Fake-307”. However, the analysis confirmed after only a few days that this server-side 307 also behaves like a temporary redirect because the content of the destination URL was indexed for the conscious URL – but with a snippet, and the title for the actual one Source page ranked.
When evaluating a 307 redirect, maximum attention is required, and it is important to make these case distinctions in daily work.
Moving rankings – which redirects?
The question that is now obvious is: Which redirect is the best and most sustainable way to “move” rankings? Google recommends the “301 redirects” for permanent postponements, but that doesn’t mean that this recommended option is the best from an SEO perspective.
You want to “move” rankings when you
1. Introduces new product pages to replace old ones,
2. Changes domain,
3. “Forward” links on old pages to new pages.
Although it would be intuitive to use permanent forwarding (301) based on Google statements and based on history, from today’s perspective, this can no longer be recommended.
In months of tests, the impression was confirmed that the “classic” redirects 301, but also the 302, are filled with various filters, delays, and “devaluations.” A comprehensive publication of the ongoing SEO experiments has not yet taken place. But the following observations make you think:
1. The rankings that are passed on by 301 disappear after a few weeks without ever appearing again for months. The effect of this is that without “post-control” everything looks good at first, but rankings “get lost” at a later point in time.
2. The rankings, which can be passed on very quickly (within hours) with a 301, then take days via a “temporary” 302. Even the “new” temporary 307 redirect is faster.
3. The rankings are transferred fastest via a 307 and remain stable for months. Some of these rankings seem to compete with other rankings that are transferred at 302, but apart from a few fluctuations, they are the most sustainable. Interestingly, only on July 1, 2016, did the “307 rankings” completely disappear, which allowed the adoption of a new internal filter.
The “old” 301/302 redirects have been discussed, used – and sometimes misused for years. The new “307” has not yet been defaced by SEOs or is so new that many employees who have intensively dealt with SEO and its tricks at Google are no longer employed there.
Corresponding SEO filters, which have long existed at 301 and 302, were accidentally not implemented on the 307. Of course, this motivates further tests and experiments and can be a real opportunity for some.
That may have changed with the “new Google version” since the beginning of July.
The following questions make you think about Google’s statements:
1. Who benefits if websites lose rankings a few weeks after a migration?
2. Which redirect has been used intensively by the SEO community to manipulate the rankings for over a decade?
3. Which redirect has been said since the beginning of the Google era that it does not transmit rankings, and when was the last time it was checked by whom?
4. How many independent research approaches and experiments on the topics of SEO and link building have existed since “Self-Publishing” has become so trivial with Facebook and Twitter that often only comment on a retweet or share is the only contribution sufficient for “Visibility “Is?
For ambitious search engine optimizers, a detailed and ongoing review of the knowledge and assumptions can only be strongly recommended.
Redirects and their effects (official Google statements)
- Permanent redirect (301): The target URL is indexed by Google and is visible in the SERPs, the redirect is cached.
- Temporary redirect (302): The source URL is indexed and is visible in the SERPs, the redirect is not cached.
- Temporary redirect (307): According to Google, only a client-side redirect.
The statements from Google are definitely to be enjoyed with a pinch of salt, and please test yourself.
The long cultivated opinion and recommendation about the usual “best” Redirect 301 should be known, but also critically questioned. It looks as if switching from 301 redirects to other redirects would be more advantageous for many applications. A long-term view and more experience reports from other SEO experiments would be very desirable.