User Pathways

This image was shown during Peter Moville’s talk about IA 3.0. What is interesting about it is how he linked this to Christopher Alexander’s text about design in architecture and also Peter Merholz’s essay Metadata for the Masses. In which he highlights ‘desire lines’ how paving is built once you see the paths that people tread.

If we look at online behaviour, user paths give us a solid idea of routes to content, where they return to and where they tend to go next. Human behaviour tends to follow patterns, see this article about mobile phone usage for an example of how predictable we tend to be.

Engagement is Conversation

This pattern of behaviour is something that good web analytics managers can see by interrogating their data. Alec Cochrane and I recently presented on the subject of building and measuring engagement and he summarises our talk here.

Metrics can tell you the most popular paths to frequently visited pages and we can change the interface accordingly. This can help us formulate our persuasion architecture but of course there are other things to consider…

No place like home

The homepage on a site suffers in a different way as it’s a starting place, a returning place and contains areas created as a diversion (meant for conversion) by many different stakeholders.

Add to this the fact that many users from search, email or RSS may never see it and it is an area that seems to be in a state of struggle. Serving the purpose of communicating a message from the site owners as well as serving the user.

Of course the more routes you have to your content the better the overall measure of the engagement your site has with your community of users.

Elements of engagement

A concept model can describe elements of engagement that need further exploration

But if we look at this in another way and think of the home as a place that should react to user’s wishes regularly, then it would be far easier to create an engaging experience.

Personas can help here but going back to the concept model would bring the greatest clarity about what needs to be presented. The concept model describes the whole site but also captures the key elements of engagement, and the parts that need representation on a homepage.

Taking this into account a typical homepage needs to offer;

1. Fresh content that is regularly updated with visible changes aided by the use of images. (Users check back frequently for news)
2. The use of themed areas and a cluster of stories around a subject area
3. If Jobs are offered they need a better promotion with an idea of the amount of job types displayed to the user
4. A consistent placement of sign-up areas throughout the site, this will enable conversions to occur and conversations to begin between you and your users.
5. An improvement of the graphics and images used on the site needs to be made to convey quality and of being a trusted authority. (Maximise promotional areas and do not give too much emphasis to ‘most popular’ widgets. These regions can take up large amounts of screen area and though they drive some traffic, without accurate measurement we can never be sure of the effectiveness of these devices)
6. Research material (white papers or case studies) need to be placed in an obvious position with more ‘evergreen’ content in the form of useful research that is graphically promoted.
7. Use blog content in a way that enriches themes or offers other angles on an existing story (also helps in changing of content on homepage). Opinion pieces also help in conveying authority and tone at the place where some users are visiting for the first time and also engage with returning users looking for an opportunities to converse.

Context, placement and pathways

The points above are based on news sites that offer different elements and in turn different modes of user engagement. Obviously looking for a job is not the same as researching and looking for case studies. So context is always the major factor when we think of what elements to display.

Put together the user’s path through the site, the context of where they have come from and what they are interested in and you have the ingredients to build the best engaging experience. The tricky part is to have an adaptive interface that will allow you flexibility.

By being aware of context, placement of content to reflect interests and user’s paths it will give a road map for engagement success.

Luke Wroblewski - Content Page Design Best Practices

One of the talks at the IA Summit was by Luke Wroblewski, author of two books and various resources published on his site. If you can see/hear the presentation at this location, I would urge you to do so. There will be something in there I have missed!

The content he shared, was an insightful window into how we design pages and how the business requirements of a page may actually work against it. It really reminded me about the mechanics of persuasion, and he highlighted some insights explicitly. The following observations were made by Wroblewski.

The web ecosystem

Conceptually, although we illustrate site maps as tree-structures, we know the pages exist with specific unique relationships to other web content.

We also know that the paths to content are becoming more fluid and not just about search. They are becoming more about authority recommendations from trusted sources or conversations around subjects with peers.

Wroblewski breaks this ecosystem down into;

• Communication – Instant Messenger, Twitter, Email

• Display Surfaces – Facebook, Netvibes

• Content Creators – Blogs, websites

• Content Aggregators – Digg, Slashdot

• Search – Google, Yahoo, MSN etc

This ecosystem will be different for every site, and every user has their own network (whether its explicitly or tacitly known).

Presenting the user with clear CTAs

Calls to action are often ignored and yet it’s the reason for a user’s site visit. On news websites, users want to read content   and yet commercial pressure tends to crowd the content area.

This slide highlights the NY Times content area

Wroblewski showed some examples of where advertising and irrelevant page content took up 76% percent of screen real estate leaving 24% for content. He compared that to the New York Times which had a 90% focus on content.

Surely a no-brainer as to what we need to do? Obviously the business context needs to be considered here but a few charts made me think about the rationale used in designs.

This slide outlines the access areas that a page needs to display, note that it must be adaptive and that the technology of the site doesn’t prescribe a treatment for the interface.

Instead of showing everything available on the site on every page, we need to be more targeted about what we present to the user. Wroblewski backed up his observation by urging us to do some thinking for our users.

Why do they bounce

He cited the book, The Paradox of Choice: Why More Is Less by Barry Schwartz. The upshot is that when confronted with too much choice, a user will use the back button, the most simple choice – and bounce rates are inevitably increased unnecessarily. More bad news is that the peak value for a user conversion was between 2 and 3 seconds.

This slide states the attention deficiency of users in general  

The idea of giving the user a tailored experience, and not making the user think is exactly what helps define a good design and fulfil its business objectives.

Design principles for engagement

From what Wroblewski said about good content page design , I took his comments and also added some of my own to form some overarching principles.

1. Content – build for focus, deliver on what you offer, short, concise and easy to scan, forms the best platform for engagement. Build bespoke channels for visitor flow but ensure they are flexible. Flexibility, or adaptable user paths are key to an engaging and versatile site that can accommodate changes in site structure and user needs.
2. Calls to action – think about CTAs and their presentation, give clear choices and make sure they are not too numerous. CTAs (and even the necessary evil of advertising) will be welcome if relevant.
3. Context – Maximise credibility through visual design, this helps build trust. Build credibility through visual hierarchy in the minimum space possible and an appreciation of a user’s situation (where they came from, the origins of the traffic). Make sure the user gets easily grounded on arrival  and can orientate themselves.

I think Luke Wroblewski deserves great credit for crystallizing the thoughts of many designers and getting the statistics together to back up his common sense approach to the problems designers face.

By stating methods to enable engagement, Wroblewski has furthered my thoughts behind the arts of persuasion and conversion. For this reason I have bought his latest book, Web Form Design and will be posting a review on the site, later in the year. Form design is a major hurdle where so many users fail to convert, and I believe this book will be invaluable in addressing that.

Mathew Milan - The Information Architect and the Fighter Pilot

If you click the image above you will go to a response to the presentation by Mathew Milan, that contains the presentation slides with audio and numerous comments from readers beneath it.

From my point of view this was the most thought provoking of the presentations because it touches on elements of my design education, that of reflective practice. But it is really important because of the ramifications of Milan’s observations, and the ensuing discussions…

I commented on the post:

….The fighter pilot story was an excellent way to re-frame how we think of ourselves as design practitioners. The OODA model (Observe, Orient, Decide and Act) is very similar to the reflective practice model talked about by Cal Swann.

He said that ‘The design process is iterative. It can only be effective if it is a constant process of revisiting the problem, re-analysing it and synthesizing revised solutions’

In effect it is a PAOR model - Plan, Act, Observe and Reflect. Its the reflection part that is important (perhaps the differentiator to Boyd’s) that feeds into the next cycle. In reflection we evaluate and prepare to synthesise our next move.

What we also need to remember is our ability to garner tools, techniques or tricks from experience. In the fighter pilot’s situation that may be through training or combat, in our experience it is through many different types of work.

Then we gain tacit knowledge , the ability to intuitively know when something is right and looks good. That is what defines great designers (or fighter pilots, sports stars, dancers etc).

Designing for emergent systems relies on tacit knowledge and no process or methodology will give you that.

It was a wordy comment to make I admit. But a presentation like this is meant to inspire and start discussions and Milan has a knack of conveying complex concepts in an easy to understand way.

The upshot is, we need to be able to adapt, those that do will survive. He states that we need to embrace the Chi (the unexpected or unorthodox) and the Cheng (the expected, orthodox idea).

Deconstruct to reconstruct

We need to synthesise our ideas faster, use our solutions appropriately. To construct effectively we need to deconstruct, break down to build up. Milan suggests we are too fixated with structure, we work to ‘permanent grids’, we are bad at ‘unstructure’.

This element harks back to Jared Spool’s presentation. How we become entrenched in methodology without the ability to move onto using techniques that perform well in emergent situations.

Milan states that we are only half a discipline. He didn’t really explain this in the presentation but he does in this post.  He states that our interaction design qualities are addressed easily (wireframes, flows etc) with visible elements. Knowledge management areas such as taxonomies and metadata are also a well-documented.

However his concern is the ‘deep IA’, the value of what we do beneath the deliverables.

What is the value of ‘deep IA’?

Well maybe the answer lies in the arts of persuasion, understanding user paths and designing with empathy of context. Emerging and adaptive systems are chaotic and unpredictable, but so is any design process.

There are always mental leaps that occur that can not be rationalised or explained. But we must embrace these relationships, that may be tacit and intangible, this is what we must be able to do effectively - use tools that bridge these chaotic unknowns.

I personally think the tool to do this effectively is the concept model. Its versatility to show various problems and relationships, its freedom and ability to show complex layers, make it a core tool to help exhibit fundamental elements of ‘deep IA’.

A fascinating talk that will help define a new direction, it was the meat between the bread of Spool’s keynote and Hinton’s closing plenary.

Persuasion Architecture has been around for years, Bryan Eisenberg (and his brother Jeffrey) founded the term and has been successfully establishing it as a concept and a measurable process. However, in a recent post, he states that after 7 years we still must be aware of usability and optimising the user experience. Regardless of the passage of time, sites still struggle to be successful.

I tend to agree about this, but then its because when dealing with websites, there are a myriad of elements that will ensure a successful site. Not least, good IA and interface design and of course the ability to convert your customers or users, to enable interaction that is intuitive and easy. But importantly, you need metrics to prove the KPI’s are changing and that all the money spent on IA and this persuasive design is worth it. Its the clearest way to get ROI on IA.

Good IA should always include persuasion architecture within its remit. Its a part of the process. After enabling findability you then want the user to either buy, sign-up, comment, post, bid, sell in the easiest way possible. Of course this will never be straight forward and requires constant testing to ensure the funnel is doing its job.

It comes down to the user’s paths through the site. For each conversion we need to look at the user flow. Wireframes should be drafted for each stage of their interaction. We need both user scenarios and personas to gain a complete picture of who we are building for. We then need to A/B test the pages in the live environment and measure that effectiveness. Google’s free website optimiser is a great free tool for this.

All this should come down to a team of people;

• the UX team to gather personas and create user stories that will help inform the interaction designer
• brand marketing to help define the variety of messages that need to be conveyed and the tone of the proposition
• the IA and interaction designer for the wireframes
• the web analyst to track and monitor the A/B testing and funnel analysis of conversions and to measure audience engagement
• the web development resource to set up a user test environment with appropriate tracking

The metrics that follow will show uplift in certain areas and will encourage the stakeholders to back these projects by investing in the testing and optimisation of their sites.

In the coming month’s I will, with my colleague Alec Cochrane, investigate engagement in more detail. What excites me most about persuasive design is that it is the logical conclusion to the application of the initial information design and a way of measuring that design’s success. Persuasion is not forcing a user to interact but its a way of helping them make choices that the user has already shown interest in. Check back for updates and read more on the subject from Alec here.

Also try Omniture’s excellent whitepapers on the subjects - well worth the read…