Publishing brands Penguin and Dorling Kindersley, both part of the Penguin Group, recently completed a project to relaunch their websites and improve interaction and navigation for users.
The revamp was pretty far reaching - the team took a user-centred approach, with extensive usability testing and planning, and found new ways to think about marketing books via the site.
The group is also set to launch new sites to increase its engagement with customers - one is a youth-oriented site called spinebreakers.co.uk, which is employing teenagers in its development.
Here, Penguin and DK's online development manager Jeanette Angell speaks about the reasons behind the project and the techniques it used.
Can you run through the main aims of the relaunch?
The aims of relaunching dorlingkindersley.co.uk and penguin.co.uk were to improve the processes of finding, choosing and buying a book and to bring to the fore the inspirational content that we produce about our books. Additionally, we wanted to encourage more interaction between us and our readers.
The sites hadn’t had a significant redesign for a number of years, growing organically over that time.
Previously, our focus was on new publications but we wanted to make our long tail more visible. Our navigation and content, due to the organic growth, had become in need of some rationalisation.
Internally, we also wanted to establish an efficient system that could make our full catalogue easily accessible, demanding fewer resources, and enabling us to focus our time on creating inspirational tools and content for our readers.
At heart, our overall strategy was to focus on our readers’ needs and for them to be able to establish a relationship with us; our books, our authors and ultimately each other.
This required a shift in our thinking from being a company focused on ‘broadcast’ to one fully engaging with our readers.
The first step was to involve them in the relaunch process, which we did by adopting a user-centred design (UCD) approach to the project, conducting usability analysis on the current site, analysing those findings and re-testing our new site designs before committing to build.
This is the first wave of our relaunch, focusing on the find, choose and buy a book behaviours. Coming months will see us engaging our readers in a much more participatory way.
Our readers are now at the heart of everything we do, as you’ll see from some of our other initiatives such as traveldk.com, which allows readers to create and print their own travel guides from a mixture of DK Eyewitness Travel content and their own and other’s contributions to the site, and also the soon-to-launch spinebreakers.co.uk, a site by and for our teenaged readers.
How have you adapted your navigation and site search?
We have moved to using a faceted search technology so that readers can now search and filter their results according to relevant criteria – subject, age range, price, format, time period and so on. If users can’t find what they want, they simply leave.
The changes we’ve made to the site enable book product and content to sit alongside each other more harmoniously. Previously, our search function could only provide results for our books; now we can provide results for content also. Our search technology allows us to target our content and promotions according to the searches or filters people are making.
We create many standalone microsites that are devoted to a particular book, author or series. Our homepages now provide us with much more space for us to give people access to this content, leaving the faceted left nav to do the job of giving access to the product catalogue.
The changes have also brought us some challenges; not all yet surmounted.
For example, determining the subject taxonomy and facets to apply for a book. Initially, we’ve focused on the practical facets but now we have that in place, we can now focus on some more inspirational and subjective methods for finding books.
We also have the ongoing data challenge, as does anyone with a vast product catalogue. We spent a significant amount of time working through the product catalogue, classifying to our new taxonomy.
Ensuring 100% accuracy of such a vast catalogue can prove challenging, so we have engaged our audience here too - as Anna Rafferty (Penguin's digital marketing director) mentioned in her bookseller.com blog post:
"If you discover a book in an inexplicable place, we’re sorry, please bear with us and feel free to drop me a mail (I might even send you a thank-you present)."
What usability techniques did you adopt and what were the main findings?
Our first step was to analyse the sites we had; establish what people wanted from our websites and whether we were on the right track with our assumptions; and identify the key usability issues in the journey from home - search results – product page – checkout - order.
Working with Flow Interactive, we conducted a series of one-on-one observation sessions asking people to perform tasks such as finding books, choosing books, buying a book, looking for content they were interested in and so on.
If there’s anyone reading who hasn’t yet sat behind mirrored glass, watching your visitors try and fail to do what seems to you to be the most obvious task, I urge you to experience it. The lessons learned from this experience can be staggering!
Following this we went about identifying our key personas – four people who represent the different types of people who might be looking for books, or looking for information that might result in a book site visit.
We deliberately didn’t include people who were already loyal to the DK and Penguin brands, as we know from our newsletters and subscriber base that we already meet their core needs. Of course, any usability issues that we resolved for people not yet enjoying our brand and product would also bring benefits to our existing customer base.
Taking our personas, we examined tasks that would see them arrive on our websites. From our findings from the usability analysis, we launched an iterative design and test phase, working up a lo-fi prototype of the site that we then put in front of users again.
Establishing the wireframes to test took a significant amount of time, wall space and copious amounts of post-it notes (every web team's crucial tool).
However, changing them to resolve issues found in subsequent test phases was quick and simple. Why? Because we had stripped back to critical, core features and had not yet committed branding, colour, style and more importantly, expensive unnecessary development time. Our testers could concentrate on the tasks in hand without having their judgement clouded by peripheral factors.
By the time we came to build, we were confident that we were building a usable site. The old adage that you can’t spend enough time planning is true – especially when you extend that to your IA, wire-frame construction and pre-design and build testing.
The main findings from the old site highlighted that our search was quickly abandoned, as it wasn’t getting people to the results they wanted.
The information provided on product pages didn’t always lend itself to users concluding they had found a book they wanted: 'I wouldn't know if this was for me without picking it up'.
Additionally, where we’d created branded sites we’d send people off into dead ends and further, that we could do better in providing clearer information around our checkout process, delivery, shipping and availability. These are all common traps in the online retail space, but we also picked up a lot of information about what they did want in addition to that. The redesign has revolved a significant number of those issues.
An important part of the project for us was to make sure we included the whole online team in the process, from start to finish, rather than simply handing the entire project over to Flow and a digital agency.
We adopted this workshop approach to the project because we believed it would derive the best results, and leave us with the skills and experience to apply our learnings and processes from this project to other work that we do.
What learnings did you take on from the retail research reports we’ve published by Mike Baxter?
A lot of the issues identified in Mike’s research matched those we found during our usability analysis and informed a large chunk of our thinking during the redesign process.
We ran the site through an audit of the recommendations in the Online Retail 2007: Checkout Special Report and were delighted to find that we had already resolved a significant number of the issues identified in our (at that time) soon to launch redesign. Further issues will be resolved as we migrate to a new CMS next year.
Last year’s Online Retail User Experience Benchmarks Report gave us a great starting point for tackling the facets and taxonomy of the project and led to many hours pouring over our product catalogue, thinking like our key new personas.
The first run through ended up with more scent trails than we knew what to do with, so now, post launch, we’re about to do a further rationalisation piece of work.
The principle of Design Patterns made obvious sense to us and we’d love to be more involved in this work going forward, as I believe it offers a valuable set of tools for all web users and teams.
I adhere to the opinion that the Web 2.0 community principles of collaboration and openness should also apply to all website producers. I think we should learn from each other to create better websites.
For us, the practical application of these learnings allows us to think creatively around other projects we’ve not yet tackled. It’s a cliché but there is no need for each website to reinvent the wheel. Site visitors are task focused and want to achieve what they want to do easily, not be forced to learn how to use a site before they can achieve their goal.
Imagine if the accelerator pedal on cars shifted from the right-most to the left-most side with each model. It would be nightmarish.
Standardisation of common website features – navigation, search, check-out processes and so on - would not make for a bland web experience. It would ensure that people are able to do what they need to do, while allowing companies to put their own stamp on the experience.
We aim to provide our users with an easy to use product catalogue that is easy for us to maintain. Then, around it, we can add inspirational content and other, more unusual navigational devices to inspire readers to seek out and discover books beyond their current favourite genres or authors.
Has the project prompted you to change the way you work in any way?
Yes. Firstly, in the approach we take to our content. Previously there would be a tendency for us to be on ‘broadcast’ mode, to use the website as a tool to promote the newest books.
Now, our focus is on ‘would our readers want this? How would they want it? Does it add value to them? How can we engage them in it? Will they want to be engaged in it?’
Also, it’s easy to forget the strength of the content in the book itself. Our strategy now makes this content a fundamental first consideration – if it’s strong enough to be published, it’s strong enough to convey its message in our promotional material, whatever media through which it may be being published.
Further, our approach to projects has evolved. We do much more up front now, even on the smallest projects.
We continue to draw upon our learnings from the usability testing, we consider our personas when evaluating new features. We can’t always go through the process of testing with real users, but understanding our personas allows us to think through their eyes.
With Spinebreakers we have gone so far as to put the key personas of that audience (teenagers) on the project team. They've been involved throughout the development of the site and will continue to be heavily involved from launch.
Were there any major differences in your approach with the two sites?
DK and Penguin’s publishing output is very different but during the usability analysis and design phase we learned that users' goals on the two sites are essentially the same - they're looking for information and/or books that meet a need they have, be if for a gift, their next read or how to fix a leaky tap.
The key differences lie in how we present information depending upon the actual book. It quickly became apparent that we didn't just have one product type, but that different types of books demanded different treatment on the product page.
Our first task was to classify those book types and identity what people needed from us to make a purchasing decision for each type. Penguin carries a mix of fiction and non-fiction in its product portfolio but its product webpage content was focused on fiction – on the extract.
But there is also a huge number of beautiful non-fiction books in there to be considered. DK's books need to be held, opened and gazed at. How do you express that on a website?
We refocused the book pages on the new site to make the page spreads and extracts the very center of the pages the user encounters. This content already existed on the site, but you had to seek it out.
Are you finding more demand for digital content like podcasts and ebooks?
Absolutely. For us to be present where people are, rather than asking them to come to us has become increasingly important.
Additionally, we’re providing more and more video and audio content with our authors; our podcast on iTunes is nearly two years old now and very popular. Our webcast with Jamie Oliver on AOL last year was very successful. As with any content it’s all about the audience; where they are and what they want.
We’re currently spending a large amount of time in the social network space learning who’s where and what they’re into – it is work, really!
One member of the team may be familiar to some of you as Jeremy Neumann. He’s spent the last year practically living in Second Life, running some great events with Neal Stephenson and William Gibson.
We want our readers to connect with us and our authors in the space where they want us to be.