There are lots of conversation talking about which option is better, a long page containing full article, or to break it in multiple pages? Many webmasters are confused about this and hence are not scoring good in search ranking. Here I am clearly explaining when to use a single page architecture and when to do it in parts.

Recent observations show that content in a single page is more preferred by search engines because they think that it will enhance user experience without bothering about clicking on "next" link to load the next part of the article. This reduces the latency every time a page is being loaded.

On the other hand, if the single page that you have, contains a lot of images or other bulkier contents, then it is better to divide them in parts. That's because no one is interested in waiting for a long page to load.

So it is important to balance the page size depending on your content. Here are few points that will help you to decide what to do.

  • Latency time: If you think that the single page will have very high latency then break the page in multiple parts.

  • Content should not be broken arbitrarily because it will not add any value to them. Instead try to break pages in meaningful sections with appropriate sub-headings. This will focus on a particular theme of the section.

  • Use standard HTML rel="next" and rel="prev" elements to specify a sequential relationship between series of pages that are part of the same content. This will help crawlers to understand the content flow. Never use the rel="canonical" tag to point to the first page of the series or to any other sub-page of the series. This will mislead search engines.

  • If you are providing both, i.e. a single page having full content and the same content broken in parts, then use the rel="canonical" tag in all sub-pages to point the parent article page (the full content page). However if you don't want to include the single-page content in search results then don't include rel="canonical" tag to the sub-pages, and mark the parent page as "noindex" using any of the standard methods.