Far too often, important application design decisions
are based on conceptions of users and
their tasks that are imagined by different
members of your project team but have little
basis in reality.
The result? Users struggle with an application bloated
with unnecessary features, their real needs
(requirements) go unfulfilled, and they are faced with awkward,
inefficient, and confusing methods to accomplish their
goals with your application.
Application development is expensive. Rework required
from the guesses and mistakes about users and their needs adds significant
time and cost. You need the right information for application design
to ensure that your project starts on the right path -- and stays
Understanding your applications users -- their
various capabilities, expectations, wants, needs, and goals -- is
the foundation for creating a successful application.
Real expertise and experience for your user-research
I am a Ph.D.
Human Factors psychologist, and former
of research design and statistics:
I know the academic side of research methodology. But with 30
years of real
experience conducting user-research for technology projects,
I keep a sharp focus on getting the information your project can actually
And as a user-interface
designer, I know what information your team really
needs to develop a usable application.
Identifying and prioritizing a good set of technical-,
business-, and end-user requirements is the foundation of a successful
project. It is well known that misunderstood or missing requirements
are the primary reason for cost overruns in software projects. Dont
waste time and money on developing features that your target
market won't use, or don't work the way they should.
I have significant experience in identifying and
prioritizing user-requirements directly from representative samples of users from your target markets.
If your project has little time or budget for comprehensive
work with end-users, I can do my best to gather and prioritize requirements
from your business's subject-matter experts, business analysts,
"super-users", or product trainers.
Given my focus
on business applications, much of
the work I do involves deriving business-requirements from business stakeholders at your company. Most business requirements have direct
implications for end-users and for your application's user-interface design.
Your design team needs a good understanding of
the individual tasks users need to do to achieve their goals (i.e.,
to get their work done) so they can design an application that will
make it easier to accomplish those goals. Knowing your target end-users'
current set of tasks is important because they embody important
details of various systems and business processes that may participate
in the future solution provided by your new application.
A task analysis
identifies all the various tasks that users perform (including conditional
tasks) as they work to accomplish their specific "goal".
A task analysis may identify several error-prone,
delay-prone, and time-consuming tasks in the user's the current
ways of accomplishing their goals. The analysis finds opportunities
for technology to streamline inefficient
processes by reducing the number of individual tasks involved.
- In typical business environments, a particular goal (e.g., "complete a transaction") involves a series of tasks
that must be completed by a number of users,
usually with a number of different systems. This is the current process
- "Tasks" may involve
things like: a user reading their e-mail to locate requests from
other groups in the company, locating clients' account data on a legacy system, logging into a transaction system to place an order, printing an approval form to route a request for approval, await the response, and then post order information on another system.
Ultimately, a usable application will enable users
to accomplish their goals more easily, more quickly, with fewer steps and fewer errors, and they will be more satisfied with
the experience (in some
cases, the application may also strive to strengthen and deepen
your company's relationship with the user).
in a Task Analysis
- Determine the goals and scope of the development project.
- Identify an initial list of tasks from Subject
Matter Experts (SMEs), process specialists, and
process documentation. If necessary, validate the task list
and gather additional task data from actual end-users using
contextual inquiry, focus
groups, and user surveys.
- Create a process-flow diagram
that depicts overall task-flow (or "workflow").
are a useful format for software projects.
- Identify tasks / steps in the current process flow that
could be improved, streamlined, or
eliminated (automated) by user-centered application
design. Consider the process flow with modified SWOT
analysis to prioritize which tasks to focus on
A task-analysis is an important input
for your application's user-interface
designer when designing a usable application.
details that are documented for each task/step during Task Analysis:
- What event(s) triggers the user to start
- From where/who does the triggeroriginate?
- How does the user receive the notification?
(are they told or do they need to notice?)
- Are there ever problems?
- Who is responsible for completing this
step (the Actor)?
- What needs to happen for the task to be
completedand the next person in the task flow triggered?
(locate information,process information
- What systems are involved (IT systems,
- How often do your have to do this task?
- How long does it take to complete the
task? (duration, total time. Are there interruptions?)
- How much of your job is occupied by doing
- How important is this task relative to
your other work?
- Are there errors that typically happen
during this step?
- Is there anyone else your collaborate
with on your task? What do they get and what systems do your
use to collaborate?
- Are there particular problems, annoyances,
errors that occur while doing this task.
- Considering everything, how satisfied
are your with completing this task?
- Output goes to where; to who?
- How do they receive it (which system)?
- Does it need to be acknowledged
to the originator or other users?
- In which task-steps are delays, inefficiencies,
errors, stress occurring in the workflow?
Why are they occurring?
- Does workflow status need to be monitored?
SWOT Analysis (Strengths,
Weaknesses, Opportunities, Threats)
A SWOT analysis provides an effective way to prioritize your
application's support for user-task requirements
in a competitive situation.
Nearly every new application has a competitor, and that
"competitor" is often an awkward
and inefficient set of procedures and applications that users are currently using to get their work done. The "competitor"
may also be the current version of your own application. Often,
users are familiar enough with their current ways of doing things
that any change must clearly address their needs head-on.
The modified SWOT analysis begins with a comprehensive
candidate set of of user- tasks/requirements, identified through
business analysis, task-analysis and earlier user-research. Then,
a representative sample of your target end-users rate both the importance
of these various tasks in their work, and their satisfaction
with their current methods of doing those tasks.
The average importance and satisfaction ratings
for each task is used to populate a SWOT chart. The chart portrays,
from your end-user's perspective:
- The Strengths of your application's competitor:
for your own application to be competitive it must at least
match its competitor on those tasks that users say are important and are well supported by the competitor.
- Your application also has the Opportunity to beat
its competitor by delivering a better solution for
those tasks that users say are important but are poorly supported
by the competitor.
There will also be a number of user tasks/requirements
of low-importance relative to others. Your project will be better
off by setting those aside and instead, focus its resources on the
strengths and weaknesses of your competitor application.
In Contextual Inquiry research I interview your
application's target end-users about
their work, directly in their work environment,
to understand the context in which they will use your application
and to identify requirements for a useful and usable application.
These studies are especially useful when your application will provide
an online solution for a "hands-on" paper process, or provide an integrated replacement of a number of disparate online systems.
The importance of this understanding is
obvious for specialized workplaces like the shop-floor, but even
office environments provide clues that reveal
requirements for an interactive solution
that would not otherwise be identified. These are often important details
that are so much part of the user's everyday
work that they wouldn't think of mentioning it in a discussion.
of a Contextual Inquiry (CI) project
- A typical contextual inquiry project will involve 12 to 24 sessions,
lasting about 1.5 to 2 hours each.
- Typically, the contextual inquiry session begins with the user-interviewee
at their daily work space (usually, their desk) and then simply
stepping-through and describing for me what they do when they come
in for a typical day at work: what
they check first with their morning coffee and what they do with
what "system" (telephone, e-mail, walk down the hall to
see a colleague, schedule a meeting, check a custom business system)
when they need to act on something.
- Throughout the target-user's recreation of their "day at
work" I hear about their motivations, priorities and problems,
I see the data they need to work with, the various people, systems
and procedures they deal with, and how they pass information between
them. I refer to my own checklist of issues to ensure that all the
relevant points for my client's design project have been covered
during the sessions.
- Under a non-disclosure agreement (NDA), I usually audio-record
the sessions, take brief notes, and collect screen-shots and print-outs
of significant work artifacts and photographs of the work environment.
I review the data and extract the key observations from that session
for input to detailed analysis.
- Affinity grouping - There will be
hundreds of key observations collected during the CI sessions. To
derive the key findings, I tag each
observation with values that classify it according to the specific
topics it relates to (e.g., "Analyst complains about incomplete
e-mail requests" may have been identified in 12 of 16 CI
sessions). The individual observations are grouped together according
to their various classification-tags so that major themes can be
Unlike the imagined nirvana of today's
electronic-office, real users do things
like prioritize their work by how high the stack of couriered-packages
piles up on their desk, mark important information with Post-It
notes, and struggle with antiquated legacy hardware, software, and
is a generic term that refers to a number of appropriate individuals
(i.e., your application's target end-users) gathered together
to participate in a moderated discussion of a particular topic. A well run focus group can provide insightful feedback
on a variety of products and services from consumers.
The key for the effective use of focus
groups in a design project is to know why
and when to use focus groups, and when other methods would
be more effective at gathering the necessary knowledge.
The main advantage
of focus groups is that several (typically 8-12) individuals provide
their input to the topic in a short-time
(typically 2-3 hours), and the real-time interaction among the various
attendees and the moderator can lead to important
insights that may not have otherwise be identified.
The main disadvantage is that "talk
is cheap": it is easy to collect questionable -- and
possibly misleading -- information that will be used to inform the design
of interactive software applications.
I have designed and conducted dozens of focus
groups for design projects, each tailored to the specific requirements
of the particular project.
- As a user-interface designer, I
know what information is needed to make important decisions about
user-interface design. As a former psychology professor in research
design and statistics I know the most efficient and effective methods
to collect quality data for your own project.
||A project team's
appreciation that there is more than
just one type of "user"
of their application is a necessary step in building a usable application.
There are always differences in users needs, skills, and motivations
for using the application, and these have major implications for
successful application design.
User research will discover a myriad of details
about various types of users. But that wealth information
can be overwhelming for the user-interface designer and the project
team as a whole. Designing an application for "users"
is difficult when there are many disjointed details about various target-markets and end-users to consider
in every design decision.
Personas -- descriptions
of fictitious people complete with pictures, names, jobs, knowledge-levels,
and personalities -- provide concrete examples of your application's
users to which everyone on the project team can readily relate.
This shared understanding of the
different types of users will provide the focus
that your team needs to work together, quickly and effectively,
to make the best design decisions for your application's end-users.
A design project will identify perhaps 3 or 4
primary personas -- the key users that will drive your application
design decisions -- and a number of secondary personas who should
also be considered.
If research data are not available, I develop your application's personas from the
knowledge and assumptions of various stakeholders on your project
A design walkthrough is a focus-group that uses
a methodology to gather feedback from a group
of users regarding the design of your current release, a
competitor product, a best-of-class solution, a prototype, or design concepts before an interactive prototype that will permit hands-on usability testing is available.
In the session, I "walk through" each
step of key user-tasks (i.e., usage scenarios) using a depiction of the application (e.g., screen-shots) in a way that the representative users can, as close as possible,
"experience" the application as if they were using it
in the real-world.
The walkthough participants provide feedback about
the application based on their own prior experiences doing
those same tasks in the real-world. The method also uncovers important
information about the tasks that users perform using their current toolset.
Both qualitative and quantitative feedback is
gathered during the walkthrough on both an individual- and group-basis
to provide for a thorough analysis. Satisfaction and user-experience ratings and comments are collected from a dozen or more users in a single session.
The method serves to provide the hands-on task-oriented
context that the group of users can relate to. Their comments and
questions provides a much richer set of data
than a typical interview, survey, or focus-group.
Real-time group walkthrough sessions can be conducted
in-person in a meeting room, or can be run over the Internet using
participants located around the globe.
it's easy to conduct a survey to collect end-user information, satisfaction
ratings, and other feedback about your application or Internet site.
Everyone on the web has been requested to provide site-feedback
by a pop-up / intercept survey, by an e-mailed link to a survey,
by a link to a feedback survey embedded in the site, or have been
asked to provide immediate feedback using a field embedded directly
in the application. Anyone can write a questionnaire and post
it on Survey
or outsource the hassles of data
collection and then analyze the responses and draw conclusions.
But its also easy to collect survey data that is questionable
or of little value for assessing and improving
your application's design where it matters. And often a
survey is used to provide insight about design when other research
methods would be far more effective.
When a survey would be an effective
research method for your project, I write content and develop
and execute a survey strategy that will provide high-quality
data that can be used to drive
the right design decisions.
do this requires an understanding of your end-users, their goals,
your business goals for the application, and your application itself.
But to do it right also some serious knowledge of psychometrics,
questionnaire and survey design, and statistical analysis.
Since 1995 I have conducted online user-surveys for more than
a dozen projects, including small surveys in ongoing design
projects for internal applications and large
national / global surveys for technology applications
in database technology, online stock trading, and stock-market
My background as a Professor in research design and statistics
helps ensure the quality of your results and interpretation
of the data. And as a user-interface designer, I know which
questions will be most useful for finding and solving usability
Card sorts / Taxonomies
Much of the value of Internet technology is the
access it provides to a nearly endless amount of information. But
that information might as well not be on your own Internet/Intranet
site if its users give up before they can
find it. Content is king, and usability is its servant.
But in order for information to be found it needs
to be organized and labeled in ways that makes sense to its various
types of end-users (i.e., its Personas).
That information taxonomy determines
your site's primary and secondary navigation-menu structure, which
is the principal method by which users interact with your site to
locate what they need.
Your site's information
coupled with its integration to your site's search engine and personalization
are all central to providing a usable user experience. These are
all facets of your site's information architecture.
While a designer or Information Architect can make intelligent guesses as how to organize
information, they are business insiders who are far closer to your site's content inventory
than real end-users. They know what information is
actually in the inventory and they know the businesses' terminology,
whereas the user needs to discover what is there, and will
approach the information-seeking task using their own vocabulary.
The best way to understand users' perspectives
of an information taxonomy is to use the cart-sort
- In a card-sort session, an appropriate
end-user is given a pile of index-cards: on each card is a description
(name, picture) of a specific type of information. There may be
dozens or even a couple of hundred cards in the pile. The user's
task is to sort the cards into a number of piles that "go
together." When they are satisfied with their sorting, they
describe the attributes that define each of the separate groupings.
They may then be asked to put together the two groups that are
most similar to one another, and continue doing so until all the
original groups are back in one pile. (There
are many variations of the technique suited to different purposes.)
- Statistical analysis of the data aggregated
over will uncover groupings of items that are meaningful
over the group of users in general, and also for different user
segments. This information is used to help identify an information
hierarchy that would be meaningful to your site visitors.
The card-sort method can also be used to organize
functions of software applications for findability in your application's menu structure.
click-streams, and instrumented
user-interfaces (UIs) provide rich archival data of application
usage. These data-sets hold enormous potential for understanding
the behavior of visitors using your site and for improving your
- For example, research
that studied a large content environment found that in a three
month period, over 90% of available content files were never accessed,
and two-thirds of the remaining 10% were accessed only once. But
a dozen content files were opened 100,000 times or more. The implications
for user-interface design are obvious.
Anyone with a website can collect usage data,
but anyone who has tried knows it is difficult to use them to evaluate
and improve the usability of your site (not just its marketing campaigns)
or your application. Web
/ e-business intelligence solutions
are embraced by all successful pure e-businesses, but still, many
large strategic web properties do little with this information.
I specialize in turning usage data into real
information for improving site / application design. When
analyzed appropriately, usage data can be powerful in identifying
where your design isn't working
as you want, but it cannot directly answer
why it isn't working. I examine application usage data to determine which research, testing,
and design will most benefit your application and its end-users.
I have considerable experience in both academic research and
IT projects working with usage data from desktop and Internet
applications. I use these data to understand which parts of
the application and which navigation sequences are most popular,
which are being underutilized and show that the application
is not being used "effectively", and which user-interface events point directly to user problems. My strong background
in statistical analysis and data-mining
helps me to uncover telling trends from huge data-logs that
may otherwise be missed.
As marketplaces become globalized, many applications
must be easy to use despite regional variations in language, business-processes,
and regulatory environments of its user base. To address this problem,
I will develop a strategy to engage the necessary
global audience in your user research.
- This can involve Internet
collaboration technology (e.g.,
GoToMeeting, WebEx, and others) where I can reach-out and gather
data in real-time from users located around the world.
- When necessary I travel
to locations in North America and Europe to collect data by direct
involvement with the appropriate set of users.
- I can also outsource research
to local expertise and then coordinate
global data collection and unify data analysis
to ensure the findings are of immediate use to your project team.
More Research Methods
The methods and analytic techniques listed above
are just a sampling of the most well-known procedures I've used
many times on usability-focussed projects since 1981. But there
are many other methods that can be tailored to provide the key insights
you need for your own project.
I have a Ph.D.
background in fundamental human-research methods and statistics.
I apply my knowledge and experience in research and IT projects
to provide the most timely, useful, accurate, and insightful understanding
of users and their work for your own project needs.