Website migration projects (moving an existing website onto a new CMS platform) are notorious for causing short-term and long-term drops in search engine positioning and traffic, especially when you're making big changes. As someone that has been very nervous when approaching these kinds of projects in the past, I thought I would write a quick blog post / case study on a recent migration project that we worked on with one of our clients.
A few months ago, one of our clients purchased one of their competitor’s retail outlet and website – but just the website, the files were not included. We were then given the task of building an exact replica of the site (which was previously on a bespoke .asp system) on Magento. In normal circumstances, this migration would have been a lot easier for us, but we only had 3-4 weeks to complete the whole project - as the website would be taken down once the buy-out was completed.
In order to prevent any temporary or permanent dips in search engine positioning and traffic, we needed to keep as much of the site as similar as possible, basically trying to prevent Google from considering it a major overhaul.
The original website:
The website we were migrating was ranking fairly well for a number of highly competitive search terms, and a large proportion of it's traffic was coming from these key generic terms. The products sold on the website are highly expensive, desirable items - the average sale is worth several thousands of pounds.
The existing website was optimised fairly well and it had well over 500,000 page URL’s in total.
Restrictions we faced:
- No .htaccess file (because it was a .asp system)
- No access to Google Webmaster Tools or Google Analytics until after the launch/transfer of ownership
- Very little time (a lot of work was carried out over the Christmas holidays)
- Unable to access the files on the server
- A large number of URLs were unable to be kept the same (sections that needed to be removed and dynamic query string URL’s, eg: www.example.com.asp?br1234567)
- Parts of the website were very badly organised (so we needed to change the structure)
What we did:
- Focused on keeping most of the existing content the same
- Adopted the same URL structure where possible
- Mapped 301 redirects and wrote re-write rules where changes were unavoidable (dynamic URL’s for some category pages and content pages)
- Crawled the website to identify and deal with any anomalies
- Kept all of the on-page optimisation the same (apart from a few minor changes)
- Began building a few good, natural links to the website prior to the launch (so that links didn’t stop coming to the website)
- Added rules to apply ‘noindex, follow’ meta robots tags to existing and new dynamic filter pages
- Updated html and xml sitemaps
Step 1: Started monitoring the website (rankings, traffic, 404’s etc)
Step 2: Planned URL mapping (redirects for affected pages of the website and actions for all new pages)
Step 3: Started building links to the existing website
Step 4: Started coding the new version of the site
Step 5: Created re-write rules and redirects
Step 6: Launched stage version of the new website
Step 7: Started building category pages (on stage site)
Step 8: Re-built blog and added blog articles (maintaining same URL structure where possible)
Step 9: Started importing product and category page content (to stage site)
Step 10: Checked that all optimisation was the same (where possible) across the website
Step 11: Customised the checkout (stage version)
Step 12: Comprehensive bug checking and testing
Step 13: Launched the new website
Step 14: Constant checking and monitoring (gained access to Google Analytics and Webmaster Tools at this point)
Based on previous experience and the amount of changes, I was anticipating a slight dip in traffic and rankings after the website was relaunched – which I was wrong about. Although a couple of the main keywords fluctuated slightly, we were seeing that pages that we were building links to were ranking much higher than before, meaning they were also generating more traffic.
Around two weeks later, terms that the website was not ranking in the top 50 for previously were now sitting in the top 5 and some of their biggest terms that were already in the top 5 had moved up into the top 2.
Now, a few months on, we’ve seen a significant improvement in both traffic and rankings - with the website now ranking between 1-3 for the vast majority of their target terms.
Our client was really pleased with how the project went and we’re looking forward to helping the website develop more in the coming months.