The effect of bad adverts online

Online ads: log in, tune out, turn off

Economists have been researching how successful online banner advertising is. It’s bad news for commercial website owners.

The study paid subjects to complete tasks online, with adverts being displayed between each task. The researchers analysed the drop-out rate of the experiment to come up with an effective cost of displaying adverts.

It was found that annoying adverts effectively come at a cost of $1.15 per 1,000 views. An advertiser may pay only 25 cents per 1,000 views.

The good news is the study found that it is possible to reduce this effect by making the adverts less annoying. Unsurprisingly, animated adverts were found to be the worst examples of online advertising.

However, a ‘good’ advert initiates a drop-out rate that effectively costs 38 cents per 1,000 views to the website owner, for an advert that may pay $2 per 1,000 views.

BBC News responsive redesign

BBC News switches PC users to responsive site

Since the last newsletter, BBC News closed down its old desktop website design and switched all users to the responsive design that was already being displayed to mobile visitors.

As is customary for a BBC News redesign, and indeed any major website redesign, the change divided opinion. Many users are understandably unsettled by the change. However, the reduced costs of responsive design amid the requirement to support an ever-growing variety of devices makes it the only viable decision.

All change again for the BBC News website

New media veteran Martin Belam has seen many a ‘big bang’ relaunch from the inside. He has a checklist of common remarks that users make in the midst of a redesign:

Having gone through several big redesigns at the BBC, Guardian and the Mirror over the years I have a mental checklist of the feedback I’m expecting to get each time: “If it ain’t broke don’t fix it”, “It looks like something Fisher-Price built”, “Did you let the work experience kid design this”.

I can confirm that the digital communications team has received these very comments while we gathered feedback on the new design. This is exactly why we ask not for feedback on visual design, as we are more concerned about usability and functionality issues.

Very vocal criticism online often reflects a real hardcore minority view though.

A site will judge whether a redesign has been a success on metrics like speed, availability, cost of maintaining code, ease of making changes and of course audience growth. Those things won’t always be apparent to the general user.

In 2008 Martin Belam analysed feedback on a BBC News website redesign and found the same comments cropping up.

Collaborating on content

Create better content by working in pairs

Pair programming is a common way for software developers to work. The idea is to split up the tasks, with one person thinking about the strategic direction of the work (navigating/observing), while the other focuses on the tactical aspects (driving).

Would the same concept work for content writing? This article makes a persuasive case.

It turns out that a pair-oriented, exploratory, collaborative approach is extremely valuable for crafting content, too. Several benefits soon emerged:

  • The team thinks before publishing.
  • It forces authors to stay focused.
  • It helps colleagues form a mutual understanding of their content.
  • It results in a more uniform tone.
  • It allows authors to share best practices in regards to writing for the web.

Why we collaborate

More collaboration equals more value

Part of the agile methodology the University is adopting for project work is about building collaborative, multi-disciplinary project teams. Gerry McGovern highlights why collaboration is so important.

For years, scientists have become more and more specialized in their areas of study and research. But if they are to achieve anything of worth they must collaborate with their colleagues. In 1978, the average number of authors per academic article was less than 2. By 2010, it was almost 5. You can’t solve complex problems on your own…

The Norwegian Cancer Society used to have 45 people contributing to their website who hardly ever interacted with each other, resulting in content duplication and 5,000 pages. Now they have six professionals who work very closely together. They have 1,000 pages and all of their key metrics of success are surging ahead. Less people, more collaboration, greater value.

Ethical social research online

The trust engineers

I enjoyed this recent episode of Radiolab, which looked how the internet has given social scientists an unprecedented ability to conduct social experiments, which then helps us improve our designs, in an ethical manner… or does it?

I include it here because a lot of the concepts described in this podcast are similar to what we are trying to achieve with our user-centred approach by looking at data, conducting usability tests and so on.

The BBC’s accessibility case study

Accessibility originates with UX: A BBC iPlayer case study

This is an interesting case study in the challenges of creating an accessible design for the BBC iPlayer.

It may not seem immediately obvious, but visual design can have a massive impact on users who cannot see the page. I often find that mobile applications and websites that are problematic to make accessible are the ones where the visual design, by dictating structure, does not allow it.

This does not mean that standards and guidelines are redundant — far from it. But what we have found at the BBC is that standards need to sit within, and inform, an accessibility framework that runs through product management, user experience, development and quality assurance. As such, accessibility originates with UX. Most of the thinking and requirements should be considered up front so that poor accessibility isn’t designed in.

Getting the content right for millennials

Millennials say the right content drives brand loyalty

The headline sounds positive, but the context is that the wrong content drives young people away. This is a reminder of the challenges we face communicating with our younger users who have grown up with digital embedded in their lives.

Of the respondents, nearly two-thirds (62%) said that the right online content does increase their brand loyalty. However, most are being turned off by the current content being offered to them. Just 32% believe modern brand communications are of any help to them. The type of content they are currently receiving is perceived as being too long, too sales-driven and ignores their individual cultural interests.

The good news for brands is that only 12% declared their active dislike for marketing communications. Therefore, brands that can engage Millennials through personalized, humorous, intelligent and helpful content have ample opportunity to drive loyalty and sales.

A reminder of the downsides of FAQs

FAQs are the dinosaurs of web navigation

Gerry McGovern outlines some of the key problems with FAQ sections on websites, including the following:

  • If information is frequently asked for, you really ought to focus on it more in your navigation.
  • FAQ sections are not user-centred because a user has no way of knowing if their question is “frequently asked”.
  • They are easy for writers to create, but inconvenient for readers to use.
  • They are often used not as a way of answering users’ questions but as another dumping ground for PR propaganda.
  • Questions are less scannable by readers.

When I wanted to update my address for my TV license, out of desperation I ended up on the FAQ page. The first question was: “Why is it important to pay your TV license?” Right. A definite FAQ…

Sarah Richards, who did excellent work for GOV.UK in creating useful content, has written a number of great pieces debunking the FAQ. “FAQs are convenient for writers,” Sarah has written. “They put everything in a long list; it’s all neatly organised and the ‘Q’ does a lot of work for you. But they’re more work for readers – questions take longer to scan and understand than simple headings and you can’t take any meaning from them in a quick glance.”

How the web is used in China

On China’s bleeding edge: Web design trends 2015

We, along with our colleagues in Admissions and IT Services, have been doing a lot of thinking recently about how we can best reach users in China. Web culture in China is very different to here, and it is even more difficult for us to keep on top of the trends. So this article about where the web stands in China today was very timely.

What surprised me the most is just how dominant mobile appears to be, and in a different way to in the west. For instance, the mobile app WeChat is ubiquitous, and it contains a native browser that has a very large bearing on how websites are designed and optimised in China.

I do a bit of UI consulting for foreign firms that are thinking of entering the Chinese market, and I find myself saying one thing quite a bit: Localizing for China means more than translating your desktop website into Mandarin and calling it a day. It might sound harsh but here’s my advice: Do China right or don’t bother. If you half-arse it, it will be obvious, and you will lose money and credibility. Know who you’re talking to and how they use the web, and open communication via the channels they expect (WeChat, light apps, QR codes), and you’ll be that much closer to a meaningful connection with your users.

History of the homepage prototype project

The prototype homepage aimed at external audiences has been unveiled this week. It is the outcome of over a year of work.

It has been an eventful year for the digital communications team. One year on from the launch of Study at St Andrews, we are now in a very different place.

Following the perceived success of the new principles developed as part of the Study at St Andrews project, and a recognition that the University needed to rethink the way it approaches digital, the digital communications team and the web team merged. (The new name for the team is digital communications team, which brings to mind a certain Spitting Image sketch.)

Since then we have been undertaking a major piece of work to create a digital framework that will help us with our aim to create a simplified, more user-centred website. You will hear much more about that in the coming weeks.

Our main deliverable product to sit alongside this is the creation of the homepage prototype.

The problem with the existing homepage

Transformed digital landscape

Screenshot of the existing homepage from the Internet Archive

The University homepage last underwent a substantial redesign in September 2008. The design did a good job of meeting the required standards of the time. But the digital landscape moves quickly, and the design is no longer suitable for today’s users.

To put it in perspective, the iPhone was still a new piece of technology, and the mobile web was not yet taken seriously in the way it is today. The first iPad was launched 18 months after the University homepage was last redesigned. Mobile and tablet traffic now makes up over a quarter of all our external visits.

Instagram was launched two years after. In 2008, Facebook had 145 million monthly active users. By the end of 2014 that figure had increased to 1.19 billion.

Google Chrome had not yet been released. The most popular web browser was Internet Explorer 7, a browser that many major websites now do not even support.

The digital landscape has transformed completely in the past seven years. Users have very different expectations. If anything, the pace of change is increasing. As such, a rethink of our website is long overdue.

Institutional pressures

Challenges from within the University have also left the existing homepage creaking at the seams. When the existing homepage was launched it contained 52 hyperlinks. Today that same design accommodates 74 links as more and more stakeholders have requested space on the homepage.

Part of the digital communications team’s challenge is to streamline the homepage to improve the focus on our external audiences. We must ensure that the homepage is meeting external users’ genuine needs, as opposed to serving who shouts the loudest across various internal silos.

Stakeholders may see adding yet another link to the homepage as an easy solution to their problem. But every link added dilutes the user’s experience and makes it harder to find the information they are looking for.

We have some difficult decisions to make.

First steps: looking at the data

The first step came before the original digital communications team was even formed, when I was working by myself with Admissions on the new Study at St Andrews section. I had also been tasked with investigating the creation of a redesigned University homepage. So in October 2013 I took the time to analyse the real usage of the existing homepage.

There is a great deal of web usage data available to us, primarily through Google Analytics. This data could be used not just to inform us about how people use the website, but to give us a deeper understanding of how all our target audiences interact with the institution as a whole. But historically we have not effectively used this data.

Looking at data on usage of the homepage across an entire year, I was able to build up a picture of how a redesign of the homepage could help us better meet users’ needs.

The most popular link on the existing homepage is to the current students section. However the target audience of the homepage is external audiences. So in order to determine the true relevance of a page to external audiences, we need to look beyond the raw numbers.

Many pages get a lot of traffic from ‘external’ users that are actually internal users that happen to be outside the University network (for example, students checking the website from home).

I looked both at the popularity of a section among external users, and what proportion of a page’s views were from external users.

An analysis of the main navigation options yielded some fascinating findings.

Performance of global navigation menu options

Next page visited from the homepage (global navigation options) –
Data from the year to 22 October 2013
Navigation menu item Rank % of pageviews external
Prospective Students 5 95.89%
Courses 3 98.49%
Academic Schools 7 81.55%
Alumni 50 91.03%
Parents 78 96.32%
Admin A-Z 37 58.09%
Research 67 90.28%
Library 12 33.92%
About us 31 96.10%
News 82 79.25%
Events 53 69.25%
Visiting 26 93.88%

Perhaps unsurprisingly, Admissions, Course search and Schools all attracted a lot of visitors, and a high proportion of external visitors.

At the other end of the spectrum, Library, while popular in terms of raw numbers, only has one third of its traffic from the homepage coming from external audiences. This is the lowest figure I have seen for any webpage.

The News section came out as the least popular global navigation menu link, despite additionally appearing in two other locations on the homepage.

Relationships between external-facing sections

Google Analytics data - traffic between global navigation items

View the data in detail.

I then analysed traffic between different global navigation menu items. Some interesting patterns emerged.

It was clear, for instance, that there was scope to improve the user experience for the Alumni and Research sections, two priorities for the institution in terms of our external audiences, both of which send a significant amount of their visitors straight back to the homepage.

The About the University section attracts a reasonable proportion of its visits from other external-facing sections. Strong relationships were found between News and Events, and Visiting and Maps (which otherwise attracts very few visits from any page, despite being at the top right of almost every page).

Based on all these findings, I put together a report of some recommendations. I re-ran the exercise in February 2014 to make sure that the statistics were not a blip, and I found broadly similar data.

Card sorting

Card sorting for the homepage prototype

Having established how users were using the existing website, the digital communications team sat down with our colleagues in Corporate Communications to brainstorm what information needs to appear on the homepage and who our key audiences are.

We then performed a card sorting exercise. This is a common user experience technique whereby you group similar concepts together. This is another way of helping us see what sections of the website may relate to each other, which in turn informs the navigation.


Persona - Alan the alumni

The brainstorm about our main external audiences led to the development of six personas to help inform the homepage prototype project. Having established the broad audience groups, we worked with stakeholders across the University and the digital advisory board to establish the personas’ individual needs.

Personas are another user experience tool that help us think about the website from the user’s point of view. We plan to develop up to six personas for each major project we take on, so we will be writing more about them soon.

Information architecture

Information architecture diagram

On the back of all this work, we developed a draft information architecture. This outlined the structure of the rethought external website. The outcome was a greatly streamlined navigation structure, while maintaining a key focus on user needs and institutional priorities.

The pages we have built for the prototype so far only include the homepage itself and a new About section. But creating a fuller information architecture has allowed us to conceptualise the new navigation menus and helped us focus on what really needed to be included on the homepage.

Key concepts of the prototype homepage

  • Strengthened focus on external audiences.
  • Streamlined navigation based on user needs confirmed by analytics, and institutional goals.
  • Hero banner with seasonally relevant content.
  • Facility to add temporary messages (for example in the event of an emergency).
  • Flexible area containing a number of current University highlights.

Gathering feedback from internal stakeholders

In December 2014 we began to show an ‘alpha’ test version of prototype homepage to a variety of internal stakeholders — content knowledge holders and those who have been engaged in the process via the digital advisory board.

Throughout the alpha phase we have been taking your feedback on board and have worked hard to include more content improve the quality of the design.

Usability testing

In February we held a number of usability testing sessions. This is the best way to find out about any functionality problems with the website. Again we have worked with stakeholders to help us set up these sessions (we are particularly grateful to Admissions for allowing us hold a focus group with prospective students visiting on an access event).

We held four sessions:

  • Task-based usability test with external audiences (members of the St Andrews community).
  • Task-based usability test with current students.
  • Guerrilla testing outside the library.
  • Focus group with prospective students (local S5 pupils).

The findings from these sessions were fascinating. We will share more about them in a future blog post.

Outside consultancy

Paul Boag working with the digital communications team

We have been privileged to work with Paul Boag for two days in February. Having worked with over 30 higher education institutions, he knows his stuff when it comes to University websites. His input has also fed into the development of the prototype.

What’s next

Now we have launched the prototype to a much wider audience. This is just the beginning of a conversation with our users about how the University website will evolve in future.

We are still seeking feedback. That feedback, usability testing and analytics data showing us how people use the new webpages will help inform future developments.