User Experience Blog Posts

Designing a SaaS application that can sustain growth, deliver new services, and establish on-going customer relationships, requires bringing together several levels of skills: experience, innovation/creative , and empirical/validation.

• Experience refers both to know-how regarding interface design and human behavior but also to understanding the unique demands of SaaS delivery and how it impacts the entire customer life cycle.
• Innovation is something I have touched on before.
• Empirical is about using data to understand how users/customers think and behave.

Empirical also means using data to build consensus and get beyond biases that we all bring to the table.

Biases that skew design priorities

One bias common to everyone, from UI architects and UI designers, to CEOs, product managers and development teams, is the availability heuristic - "if you can think of it, is must be important". Basically, stakeholders skew the importance of a feature/workflow scenario based on how easy it is to think of an example of someone (possibly themselves) who would use it.

The "typical user" we each know

It's common to be in meetings where the CEO of a company was just browsing a website that had feature X and so now thinks "users" of their application will need X. Or to find a product manager who clearly remembers an an email complaint about a product at his previous company and now makes the statement "users want Y". Or a UI architect who just worked on a mobile project and thinks that all "users" distract easily and can't focus on tasks. Or a visual designer that is really designing for one user they know really well - themselves.

The availability heuristic as a bias in application design is ubiquitous. One impact is that it can impede building consensus among stakeholders around what features to implement and the order in which to implement them. Fortunately there are a number of tools we use to reduce these biases and produce actionable information that maps to bottom-line profitability.

Distilling behavior patterns

Personas is one of these tools. In simplest terms, a persona is a distillation of a behavior pattern into an archetypal character. Personas tell you what people care about, what they do. Every SaaS application will likely have several personas. These might represent different demographics, different market segments, different work roles, different experience levels, or different conceptual models of how things work.

What a user wants to accomplish (behavior) and how they want to feel (emotion)

Personas are derived both by looking ethnographically at very objective behaviors and then seeing correlations that distill into a behavior pattern, as well as by observing how users "feel" (e.g. secure, confident, delighted, engaged, exploratory, etc.) and modeling these into a pattern. These data-derived behavior/motivational/emotional patterns are converted into characters, with a photo and name (e.g. Joe, Veronica) so that they can take on a life of their own, be empathized with and define the boundaries for the application design.

Disambiguity and prioritizing release cycles

Deriving personas is not some touchy-feely, creative writing exercise. but rather an effective way for design AND development teams to focus attention, remove bias and examine usage contexts and the opportunities they present. Instead of teams asking "How would I use the software" or "How would 'a user' use the software", they can ask "How would Joe use the software", "How would Veronica use the software". This reduces ambiguity and "squishy", "elastic" users. Proposed feature and design assertions can now be tested and validated against more closely defined user characteristics and goals.

We put these personas to work in scenarios in order to elicit key requirements that serve as the initial roadmap for application functionality and design. We can also plan priorities for SaaS release cycles. For example we might say: "Here is management's presumed list of 10 must-have features. But the personas that are the most important to our market, won't really care about features 3,7, and 9. So maybe we delay these features until the next release cycle."

Personas are just one in a toolbox of UX tools we use to remove bias, build consensus, identify functionality/design opportunities and tailor the SaaS environment to various user distinctions where the differences will have a meaningful effect on services and where service differentiation makes financial sense.

Remember David Byrne from the rock band Talking Heads? I recently listened to a talk he gave at the Ted conference, on how architecture helped music to evolve. Byrne's premise was eloquently simple. Artists create a style of music that works for the environment in which it will be heard.

Artists evolve musical styles to fit the playback environment

Byrne gives the example how West African music, with its intricate rhythms works outdoors in the open, but would be a disaster in a Gothic cathedral where reverberation would confuse everything. On the other hand, music that works in a Gothic cathedral (long notes, little rhythm) sounds flattened in outdoor environments.

Bach wrote music for playback in a room that was smaller and not as vaulted as a Gothic cathedral. Consequently his work could be more intricate and could change keys without risking large dissonances. His music was an evolution from previous styles, but also something completely new, designed to fit the context of the room.

Opera houses like La Scala, helped evolve yet another kind of music. Two hundred years ago, opera patrons did not listen in hushed silence. Instead they would eat, drink and yell out to people on stage and to each other. The dramatic and repeating melodies of operatic music evolved in response to this opera experience.

Today's music reflects the environment very clearly. Microphones enable soft personal sounds and complex inflections. Pop music is written for the MP3 player, capable of extreme detail but a limited dynamic range (too much dynamics could blow out your eardrums or at least force you to adjust the volume).

All of these examples illustrate the point that the context for playback determines the kinds of music that works and that evolves.

UI Design firms evolve information & tasks to fit the presentation environment

Now love of music aside, why did this all interest me?

Potential clients often approach us about developing mobile apps for the iPhone/Android phones or for the iPad. This is a high growth area for us. Most of these requests are from clients who already provide desktop or Web-based SaaS solutions and now want to scale this experience to a smaller screen and capture a slice of a very fast growing market.

Many UI firms quote the mantra "simplicity, consistency and usability are the keys to great mobile applications". While this is indeed true, it is not sufficient. Like music that evolves to fit its playback context, the user experience must evolve to fit its presentation context. This means looking at what people do in that environment and changing the very nature of the application and UI so that it takes advantage of the context.

Mobile application design as extensions to existing apps, not stand-ins

Good mobile UI design is not just about creating simpler and more streamlined interfaces (such as virtual keyboards that change depending on the kind of data you are entering), it is also about understanding how the context changes what makes for effective activities.

For example, on mobile devices, people are in extreme multi-task mode. The UI design needs to shift from being entirely focused on linear task completion (as in traditional desktop applications), to applications that are easy to enter and exist. They need to keep the user's mental and physical attention by being clear, visually rich and integrating gesture, touch and sound into the interface.

Mobile devices also provide unique behavioral and sense information such as motion and orientation via accelerometers, GPS data (location awareness), and the physicality of gestures. These can be used to refine the experience, anticipate needs and deliver options that are contextually relevant.

The most effective mobile applications act as extensions to existing web or desktop apps, not as stand-ins until the user can get back to a "real" computer.

The music of Mozart was not a stand-in for Gregorian Chants

So back to the David Byrne Ted talk; the music of Mozart was not a stand-in for Gregorian Chants performed in Gothic cathedrals. The musical styles were completely different to take advantage of the unique aspects afforded by the context.

The same is true with user experience design across different devices. The interface enables unique and valuable behaviors that can engage users in completely different ways. Whether it is a fit-in-the-palm-of-your-hand smartphone, a gesture-based tablet like the iPad, an instrument control interface like a fuel gauge, a desktop SaaS application or a speech-recognition-controlled security system, each context offers unique opportunities and the user interface should evolve accordingly.

I’ve written extensively about user experience characteristics that define a successful and profitable SaaS application.  I’ve talked about performance, connectedness, conceptual models, monitoring, on-boarding and the essential requirement to design for the complete customer life cycle.

I’d even argue that the user experience of a SaaS application can be more important to its ultimate profitability then the features provided by the core application. Because the SaaS business model relies on service subscriptions (which are perishable!), continued customer retention is essential for profitability.  Continued customer retention depends on user experience.

This is in contrast to traditional on-premise software. With on-premise software, profitability is defined by the intellectual property of the code and the value that it can command on a per-seat license. The vendor’s goal is to sell as many licensed copies as possible. The list of features is what makes the sale. Once the software is sold, the revenue is recognized. Retention or even long-term use or user satisfaction, is not the primary focus (the 10%-20% maintenance fees traditionally charged by big enterprise on-premise software vendors does not change this focus).

With SaaS, value is defined by the user experience that leads to customer retention and a predictable recurring revenue stream.The engagement process (sign-up, provisioning, social networking), the ease of use, the timed/frequent updates, a focus on high value scenarios, user monitoring, agile pricing, etc. are what engenders sign-ups and retention (i.e. the recurring revenue stream for the application). Modular, reusable UI and mutli-tenant design enable scalability without linearly increasing costs.

This radically alters what defines successful application design.  In on-premise software, the sale of the application license defines the value.  With SaaS, a scalable user experience around the entire customer life cycle defines the value.

This year Catalyst has seen a significant shift in areas of practice and the kinds of applications and interfaces we are engaged to develop. Part of this reflects changes in our own areas of interest, in particular, our work to support and develop sustainability initiatives. But I believe the shifts area reflective a larger move in the industry itself.

Compared to the last several years we have seen dramatic growth in the demand for application design services in sustainability and health care. There has also been continued growth in specific delivery styles of applications, such as SaaS and RIAs.  Desktop apps growth is relatively stable and flat, except in the area of health care, where this is an uptick in growth.  None of this is too surprising.

More interesting has been a rapid surge in interest for mobile application and gesture based interface design, as well as an increase in the requests for instrument control interfaces (perhaps reflecting the growth in embedded processors and remote monitoring).

For my own conceptual understanding of these trends I created color-coded matrix that shows the high growth areas relative to the lower growth areas.  It presents a heatmap-style view of application design trends with red representing hot areas and blue representing cool areas.

I recently attended an event on social media and application design. I got into a conversation with another attendee and asked him what he did for a living.  He paused and then answered “I help good companies let others know how good they really are”.

It was an odd answer, so I asked more:  “How do you do that?”. He then went on to explain how he developed web sites, wrote content, optimized SEO, and did the whole social media thing.  It sounded like all of the mechanical things that so many other consultants and agencies say they do. But his initial description intrigued me because his description gave me his purpose for what he did, the ‘why’ rather than the ‘what and how’.

It started me thinking about Catalyst Resources and the way that we describe what we do.  Sure we design Rich Internet Applications and efficient, aesthetic interfaces. We even refine this further by specifying our focus on mission critical applications and modular, reusable UI.  But this ‘what and how’ is not the motivator that actually makes us wake up every morning and go to work.  It is not the ‘why’ we do what we do.  It is not the essence of our DNA as a company.

Why we do what we do at Catalyst Resources is because we believe that we can fundamentally improve the efficiency and the “pleasurability” of any business activity.  The ‘what and how’ is by optimizing business processes, streamlining workflows, creating user-validated, high-performance, modular user interfaces and designing beautiful, intuitive and streamlined Rich Internet Applications.

When I am brought in on a new project I don’t immediately start thinking about what I am going to do from a UI or an application design perspective.  I think “what does this company need in order to be more successful”.  Coming up with this takes a lot more work and a lot more creativity then coming up with a well-designed set of UI elements.  But it is this goal that is the real inspiration behind every project that Catalyst undertakes.

Maybe we need to come up with a tag line: Catalyst Resources - we make the world better, one application at a time.  Kind of has a nice ring.

Many of our recent engagements have been around designing instrument control for mission critical applications. These interfaces enable workers to interact with equipment via an interface rather than mechanical gauges, switches and levers.

For example, for a biotech firm we designed and validated the user interface that controls diagnostic systems for analyzing tissue sample pathologies and drug discovery. For an energy optimization and management company we designed the user interface for instrumentation that controls and monitors real-time fuel inventories and environmental compliance. For a security system we designed the user interface that controls security cameras and alarm controls.

The user interface for these kind of instrument control is the link between the operator and decisions that affect critical infrastructure or even life and death medical situations.

So what are some of the challenges when designing these kinds of interfaces?

• Displaying data is not the same thing as displaying information
  Just because you think the interface shows it, doesn't mean operators see it, understand it, know how it correlates with their behavior, or feel motivated to take action.
• It isn't always obvious what to measure
  Determining what exactly are the critical pieces of feedback and environmental cues that operators really on can be challenging. You need to translate the operators "intuitive" understanding of the real world process and cues into an abstracted software interface. Since operators adjust systems according to the information they receive, if you're measuring the wrong thing, their decisions may have less impact or even the wrong impact.
• How do you optimally communicate the information
  Which type of displays are more effective for conveying information for particular tasks? Words, flashing lights, animations, charts and graphs, images, auditory cues, etc.? Different information requires different presentation modes and can engender measurable difference in reaction times.
• What are the most effective ways for the user to interact with the information
  Touch interface, gestural input, mouse, keypad, speech recognition? Which is going to be appropriate so the user interacts more with the task and less with the equipment?
• What are the social and collaborative possibilities
  How can the interface be designed so that it is easy to get outside feedback or collaboration. Can comparison data quickly be pulled for "expert" validation?
• How do you make the interface more engaging than the real world
  Instrument Control Interfaces are primarily about increasing productivity. But to achieve a real productivity gain, operators must use the new interface preferentially over existing systems. A poorly designed and validated interface could very easily be viewed as "Yet another damn display". So it is essential to make the display natural, workflow-useful, and engaging. Otherwise operators will find workarounds and the interface will be of much less value.
• How do you make the interface responsive
  If an instrument control interface is perceived as slow to respond, it will not be optimally used. Actions must be cancel-able, screens should load incrementally, the number of actions to complete a task should be streamlined, etc.

The above just scratch the surface of the challenges that are unique to instrument control user interfaces and application design. Biotech, manufacturing, and energy production/management are the areas where we are seeing the greatest growth in the demand for instrument control interfaces. We design these interfaces using modular, reusable UI, so that they can be re-purposed to other equipment or enhanced without code rewrites. As more and more devices gain embedded processors, the market for interface control will continue to expand at a rapid pace.

I just noticed that the Velocity Online Conference is coming up next week. This conference is focused on best practices in performance and operations for web applications to improve the user experience as well as a company's bottom line.

I've written previously about the importance of perceived performance in the design of a SaaS application. For SaaS, where customers can easily switch to another provider, user satisfaction is critically important. Low perceived performance can lead to low satisfaction and high customer churn. High perceived performance can result in high customer satisfaction and stability.

But what is the empirical evidence and how much of a change in perceived performance is necessary to have a significant impact? Well I just came across some hard data - metrics for page views, amount of interaction/use and online revenue presented at Velocity 2009 from some data-heavy players.

The empirical data on the impact of web application performance

Microsoft reported that with Bing, a 2 second slowdown in response time reduced the number of searches by 1.8% and reduced revenue/user by 4.3%. That a lot of money left on the table.

Google reported that as little as a 400 millisecond delay resulted in 0.59% fewer searches per user. But perhaps more interesting, even after the delay was removed, these users still had 0.21% fewer searches, indicating that a slower user experience affected long term behavior. While Google did not report revenue directly, fewer searches likely means fewer AdWord clicks.

Shopzilla had the most complete data on the impact of performance on the bottom line. A year-long performance redesign reduced response times by 5 seconds (from ~7 seconds to ~2 seconds). This resulted in a 7-12% increase in revenue.

But it's really perceived performance that matters

This empirical "bottom line" data becomes even more interesting when it is reviewed in the context of the study 'The Truth About Download Times. This study found that users do not rate the download speeds of Web pages based on the actual stopwatch-clocked download speeds. Perceived speed is dependent on how well "users successfully complete their tasks on a site."

In other words, it's not just page load time that matters. It is the time it takes for a user to successfully complete a task that has the real impact. In the case of the Velocity 2009 data, users couldn't interact with a page until it loaded. Those few milliseconds of additional time, prevented users from accomplishing their search tasks. That signficantly impacted the bottom line.

Designing RIAs for perceived performance

So how do you design a SaaS to enable the user to perform real work more productively - to feel fast, responsive and streamlined. Our RIA designers focus on a number of application flow and design areas.

• Determine the high value scenarios and then reduce the number of actions required to accomplish them
• Minimize the need for complete page refreshes and reloads - the waiting between screens creates serious friction with the user
• Incrementally add information and functionality to a page, based on an analysis of how users process the information (e.g. load a modular component only when needed or in the background based on predictive analysis)
• Progressively download data locally to avoid round-tripping to the database (HTML 5 is a great development here)
• Validate designs (visual, information, interaction and architectural) to determine where users perceive adequate performance vs. where they get slowed down and wait for the web application to catch up
• Enable queries or actions to be canceled - it gives the user a feeling of control and lets them move on to something else they consider more important than waiting
• Add reliable indicators of progress on activities (i.e. don't let users get frustrated in front of a screen not knowing how long something will take)

RIAs are about more than features, graphics and nice visuals. Optimizing perceived and actual performance has a real and quantifiable impact on the bottom line. The more productive an RIA/SaaS design, the more users interact with software.