The prerequisite for a successful new website is that the technical foundation of the website is in order - i.e. that there are no technical errors and inadequacies that reduce the quality of the website and prevent Google from indexing the website. It is also essential that existing content from any old website is retained, expanded or optimised in line with the search habits of the target audience.
In this post, we describe a number of recommendations and considerations when developing a new website. The recommendations apply to the development of a new website, regardless of the choice of CMS system (WordPress, Joomla, Shopify, etc.).
For a more in-depth guide, download a copy of our migration log. Also consider downloading our SEO book, in which you get a more general guide to SEO.
Wireframes
Wireframes are graphic illustrations of the structure and content of a website, which are prepared before the actual programming of the website begins. Typically, relevant key people review the graphic wireframes to ensure that the requirements and goals of the new website are met through the design. It is also important that any wireframes are also reviewed from an SEO perspective.
It is both easier and cheaper to make adjustments to the website during the design process, rather than once programming has started, and if you have to do a lot of reprogramming afterwards, you risk pushing back both the budget and the timetable for the whole migration. Therefore, resources spent on reviewing and adjusting wireframes are well spent.
When reviewing wireframes from an SEO perspective, pay particular attention to the amount of content that can fit on the pages. As a general rule, the amount of content on a page should be greater than the amount of content competitors at the top of Google have on their corresponding pages. In practice, this means that a template - usually used on many different pages - should be able to accommodate a body text of a very variable length depending on the individual page.
You can map the content of competitors' pages by doing a relevant search on Google and manually counting the number of words on the top-ranked pages. Alternatively, you can use a browser plugin such as SEOquake to automatically count the number of words.
When reviewing the site's wireframes, it is also important to ensure that the mobile version of the site does not contain less content than the desktop version. It is the mobile version on which Google's indexing of the website is based (so-called mobile-first indexing), and therefore the mobile version must not be a reduced version of the desktop version.
Indexing
While the new website is under development, it is important that search engines do not indexes the website. This is partly because strangers should not be able to access the website while it is under development, and partly because it can have major consequences both for the visibility of the existing website on Google and for the visibility of the new website once it is launched. Search engines can be prevented from indexing the website in four different ways, the first of which is the one we recommend:
- IP restriction. Allow users with specific IP addresses (e.g. company IP address and selected home addresses) to access the website and block all other users from the website. Can usually be set up by the company's web or system administrator.
- Noindex-tag. Insert a noindex tag () on all pages of the website. The tag tells search engines not to index the pages. The disadvantage of this solution is that strangers can access the pages if they know the URL.
- Password Protection. Set up password protection on the website. This is an effective access protection, but it has a significant drawback in that automated SEO tools typically cannot access the website, making testing of the website more difficult.
- Robots-txt-blocking. Block search engines from the website via the robots.txt file. The disadvantage of this solution is that Google does not necessarily respect the blocking and that strangers can gain unhindered access to the website.
Side architecture
Side architecture is an expression of the site's menu structure and page structure. If, prior to the development of the new website, a keyword analysis, the page architecture should be determined based on considerations from the keyword analysis. This will ensure that the website is built from relevant pages with relevant content that the target audience is looking for.
For example, suppose the website consists of a number of category pages, sub-category pages and product pages. If you don't know your target audience's search habits, you'll create all these pages blindly without knowing if there's a demand for them - unless you have user data from your existing website that can tell you something about your target audience's intentions and needs.
A concrete example is a webshop selling workwear. Here is a very short keyword analysis, prepared using Google Keyword Planner:
The analysis shows that there is almost 80 times more demand for 'workwear' than 'workwear' and that 'workwear' is therefore by far the best choice as a category page. The analysis also shows that there is a high demand for safety shoes, work shoes and work shorts, which would therefore benefit from their own category page on the website.
Internal links
Internal links are links on your website that link to other pages on the website. These links need to be updated during the migration process to make sure they point to the correct pages on the new website. In order to update the links, you need to have an overview of all the pages on the new website and know their URLs.
Examples of internal links are:
- menulinks
- text links in body text
- knaplinks
- picture links
We recommend that links are written with relative URLs (e.g. '/specials/seo' instead of hhttps://inboundcph.dk/specialer/seo'). By making links independent of the domain, you can change the domain without having to correct the links afterwards. If you already use relative URLs and do not change the URLs of the pages during the migration, you probably do not need to update the internal links on the website.
Note whether the URLs of the new website are written with or without a trailing slash (e.g. 'inboundcph.dk/seo/' versus 'inboundcph.dk/seo'). Only one of the two versions may be used as the web address (and users must be automatically redirected from the wrong version to the correct version using a 301-redirect). If the URL of internal links is wrong, users are redirected and the time it takes degrades the user experience.
Finally, a word of advice about internal links in body copy: the purpose of internal links is to guide users to other relevant pages and to help search engines understand the structure and main pages of the site. Therefore, it is an SEO advantage to link internally to the most important pages of the website, which are correspondingly important for you to be visible on Google.
Content requirements
Content is crucial for the visibility of the website on Google, and therefore it is important that relevant content from the existing website is at least maintained, and if possible expanded and optimised based on keyword analysis. Relevant content in this context is defined as content that has gained a significant number of visitors from Google over a 12-month period. How high a number of visitors is significant is up to you to consider.
You can use Google Analytics to see which pages have received visitors from Google. Click on 'Behavior' in the menu on the left and then 'Site content' and 'All pages'. If you only want to see visitors from Google (so-called organic traffic) you can add a segment at the top of the page - but remember that it's not just the number of visitors from Google that should determine whether a page is retained.
It is important that you measure the number of visits over a period of 12 whole months. There may be relevant pages that only get traffic in certain periods - for example, Christmas-related pages - and you lose that knowledge if you only look at traffic from, say, the last month. If you remove pages with visitor traffic, it stands to reason that you will lose that traffic on the new site.
The number of inbound links (i.e. links from other websites) to the pages on your website should also influence which pages you choose to keep. Inbound links are an essential part of SEO and one of the things that Google weighs most heavily when deciding positioning of a website. Therefore, you should avoid removing pages that have one or more inbound links.
You can view inbound links to your pages using a tool such as Ahrefs.com (commercial). However, a better solution is to use Screaming Frog (commercial) and through this set up an integration to Google Search Console and Ahrefs.com. This way you have data about the pages, their traffic from Google and inbound links all in one place.
Although some content should be preserved during migration, it is rarely possible to avoid the need to make significant changes to existing content. These changes may include, for example, adapting content, moving content and deleting content. I the post about migration we provide a number of more detailed recommendations for these substantive changes.
Technical requirements
From a technical perspective, there are a number of areas your web developers need to address during the development of the new website. Many of these areas are relevant to all types of new developments - whether you are developing an entirely new website or simply moving the site from one domain to another - while others are relevant only in some cases. There are many areas, and it can be a long process to get everything right.
Among the areas are for example redirectsthat automatically redirects users from the old to the new website, a XML sitemapthat helps Google find all the pages on the new site, and a robots.txt filethat tells Google which parts of the new website Google can index. It is important for the redirects that all pages on the old website are mapped in a spreadsheet (for example using Screaming Frog), before the website is shut down. The mapping will be used to test later whether the redirects are working.
The above areas and many others are described in our SEO book. Once again, it is important to point out that this is not a complete checklist for technical SEO, but an overview of areas that are essential for a migration. For example, SEO-important areas such as hreflang and mobile-friendliness not included because they are not directly related to the migration of a website.
Data collection and data flow
Most websites collect statistics about their users through systems such as Google Tag Manager, Google Analytics, Adobe Analytics, Google Search Console, Leadfeeder, Facebook Pixel and LinkedIn Insight Tag. These systems need to be set up on the new website so that data is not lost when the website is launched. If data is lost after the migration, you also lose the opportunity to quickly evaluate and assess the impact of the migration.
It is also important that any conversion tracking in Google Analytics is set up on the new website. If you use third-party integrations (such as Zapier) to automatically collect data from the website and integrate this data with other systems, ensure that these integrations are set up on the new website before migration.
Testing the website
Prior to the launch of the website, the new website will be tested for technical errors and bugs, and these will need to be fixed before the migration takes place. Testing should cover at least the areas in the checklist below.
If you want to ensure that the new website meets all technical SEO requirements, you can carry out a more thorough technical analysis of the website based on the technical chapter of our SEO book. This extended analysis may touch on areas such as load time, mobile friendliness, canonical tags, hreflang tags, page titles, response codes and structured data.
- Indexability. Check that all pages on the website can be indexed by Google, i.e. that they are not blocked via the website's robots.txt file or a noindex tag in the source code (any IP restriction must first be removed immediately before the migration). Recommended testing tool: Screaming Frog.
- Dead links. Check that there are no dead links on the website, i.e. that there are no links that return a 4xx error code (page does not exist) or 5xx error code (server error). Recommended test tool: Screaming Frog.
- Redirections. Check that all redirects is set up on the new website so that users are automatically redirected from the old to the new website. This check is manual until the migration has taken place.
- XML sitemap. Check that all pages on the new website are entered in the XML sitemapand that the site folder is linked to in the robots.txt file of the site. Recommended testing tool: Screaming Frog. Robots.txt is checked manually.
- Robots.txt. Check that robots.txt file on the new website does not block Google, including that Google is not blocked from indexing JavaScript, CSS and image files. This check is manual.
- Data collection. Check that tracking codes for analytics tools such as Google Analytics, Adobe Analytics and Google Tag Manager are set up correctly on the new website and receive data. This check is manual.
