Preventing migrations from flying south

So far 2017 has been the year of the migration, with Google increasingly pressing on web masters to be secure and mobile responsive.

Migrations however aren’t cookie cutter projects, and vary wildly in terms of both scope, complexity and risk. Switching to Https from Http on a small/reasonably sized website of a few hundred URLs carries a lot less risk than switching domain names, or platforms.

Domains and websites are fragile eco-systems, and can vary immensely in terms of build quality and code base (so even moving from WordPress to WordPress could pose issues).

Types of migration

SEO migrations can take a number of forms, however the most common that I’ve come across are; from Domain A to Domain B (domain migration), Http to Https (protocol migration) or from one platform to another (platform migration).

Ahead of the impending mobile first index, we’ve also seen a number of migrations and consolidations of M subdomains and separate mobile URL structures.

Domain migrations

There comes a time when some brands need to change domain, this can be down to internal factors or through acquisitions and company mergers.

In my experience these types of migration often carry the most risk as the brand profile and gravitas that you’ve accumulated over the course of being active online needs to be transferred to the new domain (which most likely won’t have any history).

A good example of this is the migration of Demandware to Salesforce Ecommerce Cloud. While this still currently resides on the historic Demandware domain, the transition in brand on the domain happened over a number of months in 2016 as to not compromise organic search visibility for Demandware.

This process is often essential as users still resonate with the previous brand name, and search for it.

As you can see from the above Google Trends graph, the gap between previous brand name (Demandware) and new (Salesforce Commerce Cloud) is still considerable, however it is closing.

If Salesforce had not handled the delicate migration in such a way, they could have compromised their position for the previous brand.

Admittedly, not all brands are afforded the luxury of months when it comes to migrating a domain and the migration can often be a short affair. It is important however to retain elements of the old brand on the new domain, and engage in offsite and PR activities to create noise to associate the new brand with the old.

Https (Protocol) migrations

Google made it clear at the I/O conference this year that https is a bare minimum security requirement, and data encryption will be even more important come 2018 and the new GDPR regulations.

These can often go wrong due to both http and https protocols being left live, canonicals not being updated, href lang not being updated… For what on the surface seems like a quick install of an SSL certificate and a search/replace in the database can go wrong, quickly.

Platform migrations

Sometimes a platform migration comes about due to vanity, from a business decision, or a need to change CMS to something more manageable.

However platform migrations are sometimes forced upon you without any control. For example, Magento 1 will no longer be supported as of Summer 2018. This news has caused a stir within the Magento community and brands are now starting to migrate to Magento 2 or 3.

Platform migrations are a minefield of their own pitfalls and due to changes in CMS, a lot of websites and databases need rebuilding.

M Subdomain migrations

Google has made no secret that they’re moving to a mobile first index. In 2015 we had mobilegeddon, in 2016 we had the awful second sequel mobilegeddon 2, and then at the end of 2016 Google announced it would be moving to a mobile first index. Google made further noise about this in 2017, and we predict that the index will be in live action before the end of the year.

John Mueller has also revealed on Twitter that heading into the mobile first index with an M subdomain, or separate URL structure for mobile and desktop has the potential of negatively impacting organic search performance.

Server & hosting migrations

Broadly speaking, if you’re changing your website’s host, server, or any other part of the sites infrastructure it’s important to consider the potential SEO implications of this.

These can be down to cost cutting exercises, or to improve performance.

We know how important site speed is, especially in a world where mobile browsing is increasing (if not the dominant form of browsing).

When migrations do go south

Sometimes a migration just goes wrong, and it can’t be explained by technical reasoning. You can follow all the best practices and religiously tick every box on a coveted checklist but more often than not, the reason for the failure of the migration is one of two things;

SEO consultation wasn’t sought early enough in the planing process and was brought in too late for recommendations to be implemented, or worse, the consultation was sought post-migration.

The recommendations made by the SEO consultant have been seen as low priority by internal teams/developers for various reasons and subsequently have been left out of pre-migration sprints, or left until after launch.

Neither of these are often done out of intentional disregard, but they are omissions resulting from internal stakeholders (often those who sought your consultation) not wholly understand the importance of processes and menial, small technical changes.

The role of the SEO in a migration

As an SEO, we can try and elevate our status within the project team and rather than see ourselves as consultants, we are partners, we are stakeholders.

Often we are employed by an individual, or small team within a greater internal team and it’s our job to assist our internal champion in conveying the importance of our consultation and recommendations. Only then will our voice carry sufficient weight at the table to impact key decisions within the project, and potentially impact project timelines.