If there is a large quantity of , either intentionally or not, this will be diminishing the search performance potential for that content. The idea is to identify and explore those pages that have untapped potential, and either re-link them to the website structure, or remove them entirely to reduce the continued risk of and . The main goal for SEO is to drive and improve search engine visibility for relevant searches, and we recognise that having a vast number of pages unlinked to any aspect of our website is negatively affecting the domain performance.

Search engines may still decide to index orphaned pages, this can be a result of other websites linking to it or if we have submitted it to search engines ourselves, but search engines still deem those not linked pages as less important than other pages that are internally linked.

How do we find orphaned pages? 

������Ƭing orphan pages is key to solving our problems. From our crawl reports and analysis, we have realised that a sizable proportion of our content is classified as orphaned, some of which is indexed.

There were a number of revealing results from our audit including:

1. a large number of and old redirected pages in sitemaps.
2. news, events, seminar pages dating back to 10 years ago.
3. many pages were purposely taken out of the navigation (content owners using this as a way of archiving their site or hiding pages) without deleting the page itself.
4. some old landing pages which could also be removed.

The majority of those pages which are unlinked can be classified as expired content, these pages are considerably contributing to our index bloat problem. A total of 36,000 unique unlinked urls (that are indexed in search results). The majority exist under old news articles in the directory /news/ and then a large number of instances under the schools (faculties) information architecture. Reflecting on our results from the crawl analysis, it was evident that we need to reduce these sections and proactively focus on outcomes to enhance performance.

Figure 1: detective work looking and finding orphan pages

It’s clean up time!

If you love crawl visualisations like we do at the University of Southampton you will notice that orphan pages will show up here which definitely provides perspective as well as creating an unusual piece of artwork to display at your workspace!

More seriously, how can we as SEO specialists prevent orphan pages from diluting our website’s performance, but also adding to a lot of digital waste. Firstly we need to understand what impact this has on key performance metrics. For example:

1. orphan pages have a low ranking capability
2. organic traffic is typically low
3. and crawl waste
4. content dead-ends and poor user experience
5. occupying valuable bandwidth, storage without driving traffic or conversions
6. overall affecting domain score
7. environmental burden to our digital estate.

In reality it may be a difficult task for any large website to have zero orphan pages. However, what matters is that we focus on creating a framework of structuring content in a user-centric way and aim towards a stronger internal linking structure. Orphan pages should be a minimal exception that proves the rule, rather than being treated as standard practice.

Maximise our efforts 

Crawl waste is very common across large domains that contain out of date or obsolete content that is not updated or removed. Leading to hundreds or even thousands of pages that do not need to exist, weighing the website down. Crawl waste exists where bots regularly crawl pages or broken links that they shouldn’t have to. That is why it is so important to address expired content, and have pages and links return an indexable status or at least a . We are already making gains with this kind of investment in web maintenance, this will really pay us dividends in the long term.

Caption: we keep looking for index bloat and orphan pages

Reduce our bloat

Continuing to hoard orphan pages that are tricky to find signals to search engines that a large portion of our content is not relevant enough to warrant ranking. The good news is it is simple enough to fix and we are already working hard on that front! Removal of low quality pages provides a better chance for more important pages to improve their search visibility in Google. We strongly believe having a plan for when to retire content is one of the most important parts of a content strategy.

Next steps for the University website

It is imperative for the future of the website that we provide users and search engine crawlers the best possible chance to discover our most important web pages. Removing expired content is another step forward in optimising our crawl budget, relieving some of that index bloat, and getting closer to improving overall user experience.

That’s not the end of it though. As part of OneWeb there was a strong focus on collaboration with the wider University and digital stakeholders. So if you hear from us with regards to minimising orphan pages, please help us to achieve some of our intended outcomes:

1. identify large content areas for removal, such as news, events, seminars.
2. clear up the domain by identifying errors in our sitemaps or pages resulting in a 404 due to being deleted.
3. identify if there are any high value content areas that are at risk.

We are working on a proposal on how to retire this vast amount of content, and will share more with you as we progress. In the meantime, we would like to invite you to share any ideas.

You can contact the SEO Digital User Experience team by emailing us:

Kath Sellwood: kath.sellwood@soton.ac.uk
Rayne Prendergast: r.e.prendergast@soton.ac.uk
Elise Corbin: e.corbin@soton.ac.uk

For example, voice search is probably something you’ve been hearing more and more about. It’s become more popular in recent years due to the improvements made by search engines. Here’s a quick look at why that matters for us:

Figure 1: Example of using voice search for UoS

Access needs first

Why does it matter?

As a team, our goal is to deliver digital services that meet the needs of all our diverse user groups.

We all know that voice assistants are on the rise and there is a vast amount of complexity in how we connect in a digital world.

As of 2019, there were an estimated 3.25 billion digital voice assistants being used in devices around the world. Forecasts from suggested that by 2023 the number of digital voice assistants will reach around 8 billion units – more than the world’s population.

Voice interfaces are also ‘old hat’ for many people with access needs and there’s much excitement in the accessibility community about voice assistants. Their dramatically simpler interface has the potential to help lots of people who read and consume content through their ears rather than eyes.

For our team, working on voice is an opportunity to meet the rising expectations of users and make our services more findable, accessible and usable.

Source:

Making UoS more understandable to search engines

Voice search allows a user to search via voice commands rather than written. First introduced in 2010 on the Google platform, it’s evolved over the past decade and become easier for users to interact with.

What made voice search a more popular method of finding information was thanks to the introduction of an algorithm change, known as ‘’, in 2013. This focused on ‘semantic search’, which enabled Google to recognise the user’s meaning and intent more successfully, and produce results that would satisfy a voice search query.

This is important because intent is about what users want to do, what they search for, and what you do with that data. We use SEO tools for it and the way people talk, or search can help us as a team to determine what they are trying to do. It also helps us to work alongside our content design colleague and help them prioritise content and determine the language we use to help our users.

An example of this, is the use of question and answer based content amongst our new course pages. What we’ve found is content that features as a structured snippet on Google search is most likely to appear for voice search queries too:

Source:

In this example, we’re using voice search for “what is criminology and psychology”. We’ve structured the content to produce an easily digestible section of information explaining the definition of criminology and psychology.

How does voice search differ from traditional search?

The way we use language when speaking is very different to when we’re typing. People tend to use shorter or more colloquial words and phrases when searching for information with their voices.

For example, a search for the University’s location may be expressed as: “University of Southampton location” when typed into Google. But when speaking to a voice assistant, a user is more likely to more naturally ask “Hey Google, where is the University of Southampton?”

Voice search tends to be very specific.

Source:

Why is voice search so popular?

Well, there’s a few reasons why…

It’s faster and easier to search

Voice search is significantly faster than traditional methods of searching for anything online; making it a very popular alternative to typing into a search engine.

Results tend to be more appropriate and convenient

When it’s faster and easier to find answers, it makes it more convenient for all. Where it becomes more appropriate is the use of structured data. The change in search behaviour resulted in an algorithm change as mentioned previously. This required SEO specialists and web developers to adapt code and structure data in a way that met the needs of search engines.

This means that not only are results faster, they are much more specific and appropriate for the user.

With the era of voice search well underway, structured data has enabled websites to get on board too. This is because it breaks down information in a way that search engines can understand and interpret. You may have heard of before, which is a specific format most websites use. Search engines are now able to read the code and use it to display search results in a specific and much richer way.

It’s more apt for mobile

Smartphones and tablets are everywhere and this makes it essential for anything to be made with mobile in mind. This applies to search engines too; with mobile-first indexing launched a few years ago, it has all worked to enhance users’ experience with their devices. With the ability to speak to your phone, voice search really took off.

How can we as a university adapt this into our SEO strategy?

So, with all that we know about voice search and the wild and wonderful workings of search engine algorithms – we’re left with a question: “Hey UoS, are we ready for voice search?”.

The answer is: yes.

As part of OneWeb, we’re looking at a strong focus on collaboration with our development team and UX disciplines to work with templates that allow for structured data.

You can’t understand content without understanding search intent data, so we’re incorporating and restructuring our content, as well as adapting the conversational language we use with help from content designers.

That’s not all though. We are always striving to improve our website’s user experience, and this is just the beginning. Voice search is a great opportunity for us and is likely to continuously evolve. What we need to discover is whether voice search is ready for university-based searches.

We would like to invite you to share any ideas you have on this topic with us! You can contact the SEO Digital User Experience team by emailing us. Thanks!

Search engines have changed the way we find information, conduct research, shop, and connect with others. Content that contains relevant keywords leads to better user experience and brings more of the right users to our website.

Almost everything on the World Wide Web – whether it’s a website, blog, social network, or app – can be found via a search engine. Search engines have become a connecting force and directional guide to everyday life.

Did you know that the majority of our website visitors come from search engines? This is thanks to a mixture of the University’s brand and the efforts of the SEO team. A strong intertwined SEO strategy is challenging but rewarding. It gives us a competitive edge against some of the biggest universities online, which can feed successfully into user experience testing, content writing and marketing in general. By contrast, a poor strategy leads to higher pay-per-click costs, lower on-page interaction, and less traffic to the website as competitors focus their efforts.

SEO is about discovering opportunities and ensuring that we offer the best experience for both search engine crawlers and users.

Definitely easier said than done!

How do users find us on search engines?

This is something that we refer to as ‘search visibility’, literally: ‘how visible are we in Google’. This starts with a specific query a user has searched, for example undergraduate degree course, indicating our variable ranking position (which changes daily!). The majority of users only ever click on links from the first page of search results, so it is hugely important to climb to or maintain a high spot on the page.

To increase our search visibility, we need to ensure we are following best practice. The University competes against some huge websites – other universities, NHS, WhatUni – it’s not an easy task to increase our position by simply changing a few words. It requires a lot of effort, collaboration, and understanding of how our website is currently used.

It’s also worth noting that search positions do not change overnight, it’s gradual and all about trust of an information source. The more users visit our website and take actions on particular pages which are reflective of their initial query, the better the potential position for that query.

Below is a snippet from a dashboard of subjects categorised within Languages, where we’ve targeted 380 keywords to date. It’s worked out by:

1. Taking all rankings for all tracked keywords.
2. Applying an estimated click-through-rate (CTR) based on each ranking position. The CTR calculation ensures that higher ranking keywords are appropriately weighted in the score.
3. Adding all CTRs and dividing by the number of keywords we are tracking in that campaign, which gives a single metric of 0% -100%, calculated to 2 decimal points.

It is typical to aim for around 35% search visibility on average for targeted keywords. By using this indicator, we can measure search visibility over time to understand our impact in Google. We have some way to go, but there is a strong opportunity here and we have an ambitious goal.

*Search visibility can be represented through a scoring percentage using SEO analytics tools.

There are three main ways we are aiming to improve search visibility scoring:

• on-page optimisation
• overall site authority
• targeting more keywords

What have we found in practice?

Whilst working on keyword research for the education section of the website, we found unexplored opportunities due to changing search behaviours. This gave us a chance to focus our efforts in supplying content designers with high volume and highly relevant keyword phrases, to integrate with their content production.

It was and still is no small task, and work is still underway, but we’ve already found rewards within the first week. Content scores have increased from an average of 70 to around 95 and, with the new platform underway, we are predicting these will improve even further.

With a higher score, we have shown more effort for optimisation, adhering to best practice guidelines from search engines, thus in theory paving the way for us to appear not only higher in search engines but for more queries, expanding our search traffic potential. All of which can help with the University’s goals and strategy.

So what’s next for the SEO team?

SEO isn’t a singular project, it requires regular analysis, opportunity sizing, and optimisation – we have thousands of keywords being tracked across education and these will continue to grow, improve and be used as a benchmark for quality. SEO is an integral part of the content design process and content design is a crucial part of the SEO process. SEO is ‘baked’ throughout the entire product development process, to ensure that we are maximising our traffic potential and getting in front of the most relevant audiences.

With a new addition to the team, we’ll continue to provide knowledge and expertise across the University, including collaborating in content cycles and workflows.

Moving into the next phase will mean taking all of our keyword research and insight gathering, and stitching it together. We’ll map where each search query is being sent, uniting the website as a sole function and avoiding any content duplication. From a technical side, the implementation of our new CMS is definitely keeping us busy!

We’re recording our Show and Tell’s and posting these on the .

Please get in touch if you have any questions, and watch this space!

Written by: Kath Sellwood & Rayne Prendergast