Interview with Dr. Deborah J Mayhew, Internationally Acclaimed Author, Lecturer & Usability Consultant
Among Dr. Mayhew’s many accomplishments, she is the author of several influential Usability books and has been applying her significant usability engineering expertise for Fortune 100 companies since 1986. Dr. Mayhew holds a Ph.D in Cognitive Psychology from Tufts University, an M.A. in Experimental Psychology from the University of Denver and a B.A. in Psychology from Brown University. Her books, “The Usability Engineering Lifecycle” and “Cost-Justifying Usability” are in my opinion two classics of Usability and User-Centered design.
1. What’s your background? Where did you go to school, what subjects interested you?
I majored in Psychology in college, discovering Cognitive Psychology late in my college years. Other subfields of psychology had interested me, but when I finally encountered Cognitive Psychology, I knew I had found my calling. What could be more fascinating than the internal workings of the human mind?!
I then applied to graduate school in Experimental Psychology, went to the University of Denver for my Masters where my initial Advisor was a researcher in intelligence testing, and finished my PhD at Tufts University.
My area of specialization was human problem solving, but I also studied general cognition, perception, developmental cognitive psychology, statistics, and neuro-psychology among other subtopics. My research assistantships involved work at the Colorado State Penitentiary doing IQ testing, and work in a primate lab tracking the social behaviors of macaque monkeys. Both my Masters and PhD theses were on topics in human problem solving.
2. How did you get into the usability field?
My graduate education was almost entirely funded by grant money. But one summer while at Tufts working towards my PhD, there was no grant money available for a stipend, so I needed to find a summer job to pay the rent. I had learned some programming skills in graduate school, so I looked for a programming job for the summer. I landed a job at a small high powered contract development and consulting firm in Cambridge, MA started up by some recent MIT graduates. This was 1978, and a PhD candidate in Psychology could easily get a job as a programmer.
At summer’s end, I liked it so much I decided to stay on. I got permission from Tufts to work full time while finishing my doctorate, which at that point, was mostly just completing my thesis. I worked for almost 4 years as a software developer and IT management consultant till I finished my PhD.
Then I took a job as probably one of the very earliest software usability professionals, at Honeywell Information Systems. That was in 1981. I was at the right place at the right time with the right set of dual backgrounds: software development and cognitive psychology. The field of software usability was just emerging.
I worked at Honeywell for a couple of years, then at Wang Laboratories a couple of years, and then went back to academia. I got a position as an assistant professor in the College of Computer Science at Northeastern University in Boston, where I spent a couple of years teaching usability courses to both undergrads and grad students. In 1986 I launched my consulting business, probably one of the first full time independent consultancies in software usability in the nation. I have been an independent consultant ever since, now in my 24th year.
3. What is it about usability that you most enjoy, or find most rewarding?
Although I have been working in the software industry since 1978, I still mainly consider myself a cognitive psychologist. I love learning and thinking about how the human mind works in the intellectual, information processing sense. In particular, human learning, memory, creativity and problem solving, as well as the more recent connections of those phenomenon to brain science, fascinate me. I enjoy the problem of applying what we know about cognition to the design of software user interfaces.
I enjoy all aspects of usability engineering – I like doing the research to understand different kinds of work people do and how they think about it and organize it. I really love doing user interface design. I also love applying what I learned about the scientific method in grad school to designing usability tests that will yield valid and useful results, and provide insight into how people approach software.
In addition, I love being a consultant because of the opportunity it provides to sample a lot of different industries, hardware/software platforms, application types and user types. The variety keeps it fresh for me.
I also love being independent. The only job as an employee I ever enjoyed was my first one as a programmer for a small firm of 50 people. Large corporations and academia were very frustrating work environments for me. Relative freedom from organizational politics and control of my lifestyle are very important to me. I live in a very rural area in a community I have been a part of since I was born, which is a lifestyle I love. Being independent means I can live where I want and how I want. The internet has meant I need to travel less and less to work with my clients, who are all over the nation and the world.
4. Your early research with usability and user-centered design methods (for example, “The Usability Engineering Lifecycle” from 1999) has been standard reading for anyone interested in an education in usability, how did you come up with this approach, what were your inspirations?
It’s a little hard to remember! 
Certainly all my methodology was built on the work of others. I taught my first tutorial at CHI in 1986 (the very first CHI conference was in 1983), and the basic framework of my Usability Engineering Lifecycle was already there.
I think my main contribution with the 1999 book was synthesizing a lot of well established individual techniques into a lifecycle process. There is very little that is original in the techniques I describe in that book – what is unique is how they are all pulled together in a logical sequence where each one feeds into the next, as well as the templates, examples and war stories I included.
I never planned to write a book. I was teaching at Northeastern University in 1985 when a recruiter for Prentice-Hall came around my department hoping to talk professors into signing contracts to write text books. They put the idea in my head, and I started writing my first book, Principles and Guidelines for Software User Interface Design, which they published in 1992. I had originally planned for a single book to include both principles and methods, but the principles half got so big I decided to defer the methods for another book, which I then published in 1999.
As for the “Cost Justifying Usability” books, I had an “aha!” moment after reading an early article on the topic by Marilyn Mantei, then gave a talk at a conference on the topic. Randolph Bias attended that talk, and then decided the time was ripe for a book on this topic and contacted me to collaborate.
5. In 2001, you had argued that Discount User Experience methods would not sufficiently reduce project risk and insure Return On Investment. You had further stated that paying for Spot usability checks by “Gurus” would not either. In today’s environment, with AGILE software development processes, do you still hold to that belief, why or why not?
That is a very interesting question! 
Yes, I do still hold that belief, perhaps even more so.
I view an application user interface as a coherent, unified system, that cannot be effectively designed piecemeal. Its like an automobile, a huge bridge, the human body, or any other complex system – all the parts have to work together in an integrated, smooth way. Thus, any approach that does not consider the whole is not going to be very effective.
I cannot look at a few individual functions, screens or pages and give effective advice on how to design them, without understanding the complete set of functionality across the whole application, not to mention a lot of key things about the users, the tasks they are doing, and the environment they are doing them in.
For instance, I cannot recommend using bold or a color as a cue for something like required fields, without knowing all the things that need to be consistently cued across the application and what cues I am going to use for what. Not to mention the fact that there needs to be a coherent information architecture that can only be designed based on an understanding of users’ tasks and the full set of functionality being offered in an application.
So for me, any approach that does not consider the entire application is going to fail to achieve consistency and coherence, the fundamentals of usability, and any approach that does not incorporate detailed information about users, tasks and environment is going to miss opportunities to tailor an interface to those requirements.
I think you truly get what you pay for.
There may be simple little page design issues that can be addressed without the big picture, but they are not going to buy you all that much. The most bang for the buck comes with higher level issues of information architecture and conceptual model design. So, the briefer and shallower the analysis, the higher the risk of missing key usability issues and opportunities.
As for Agile, I think there is an inherent conflict between that approach and sound usability engineering. I advocate a top down approach to UI design, that starts with the information architecture (in turn premised on an understanding of the full set of intended functionality and user tasks), and then a conceptual model design which is a set of standards for the visual presentation of and interaction with the IA, followed last by page detail design standards. That is, big picture established first, details to follow.
While I am not a student of the Agile method and have not actually practiced it (so far I have not worked with a client who has adopted it), I learned a little about it while consulting with a firm building software tools to support the Agile methodology.
While there is much about the Agile approach that is attractive (lots of iterations, feedback from the customer throughout, discovery of requirements as you go along via concrete prototypes), I think there is only so far you can get by putting a system together piecemeal, without considering the overall framework.
I wonder about the Agile approach because it seems to me that just like UI design, in system architecture design a top down approach is also most efficient and effective. The only way I can imagine the Agile approach working effectively is if a great deal of rework is done along the way as the discovery process evolves.
You cannot build a patchwork UI and have it be usable (or easily maintainable for that matter.) Thus, if you are going to approach it in a piecemeal way, there is a point at which you are going to have to throw out a lot of work and start from the top and design a coherent framework, in order to impose order on chaos.
One could certainly take an approach to UI design that would be compatible with the Agile philosophy and approach – I am just not convinced it will be either effective or efficient.
6. In your book “Cost Justifying Usability” you made the case that you must align your cost justification to the audience. How should a usability practitioner respond if the audience is a CFO, and that CFO expressly desires a “guaranteed” Return On Investment?
Great question!
And one I have thought about a fair amount. I have indeed been asked on occasion, and recently more often, to “guarantee” that my work will improve usability in some measurable way.
At first blush it seems like a perfectly reasonable request – I do after all claim that there can be a dramatic return on investment from expert usability work. However, the question is, how can the impact of usability work accurately be assessed?
Let’s consider an example. Imagine we are consulting on an e-commerce web site. The goal is increased sales, or “conversions.” A client might say to me, how about if you guarantee that your work will increase our conversion rate by at least 1% (which could be cost-justified, given my quote), or we will not pay. There are a number of problems here:
First, unless the client is willing to implement all of my redesign recommendations, exactly as specified, I cannot be held responsible for the results. This may seem reasonable to the client, but for many reasons – some technical, some political – they rarely do this. Who is to say if they have come close enough to implementing my recommendations for me to be held accountable for the results?
In many if not most of my consulting jobs, the client only takes some of my recommendations, or implements them in a somewhat different way than I specified. If this is the case, I cannot take responsibility for the results.
Second, many things impact conversion rates other than usability. For example, suppose a client took and completely implemented all my recommendations exactly . . . . just prior to the dot-com bust, or just prior to the recent recession? Who is to say whether it is the economy or my design that has impacted conversions?
Third, such a guarantee can work against my client too. What if they implement technical improvements which make their website work better on more browsers, or the economy picks up, at the same time they also implement my recommendations? Even if my recommendations are not responsible, other factors may raise conversion rates, and they would have to pay me regardless of whether I earned it or not.
It is rare that the only thing that changes from one release to another is the UI design. Usually many other things, both internal and external to the web site, change as well. In anything but a well controlled laboratory study, it is very hard to determine the sources of changes in measures of success or failure.
Fourth, if I could take controlled, laboratory-based before and after measures of usability in a project and use those as the measure of my contribution, that might at first blush seem reasonable. But, as we all know, UI design is part science and part art. Few of us could reasonably guarantee an improvement if not afforded the opportunity to iteratively test and redesign.
It’s rare that a client is willing to pay for multiple iterations of redesign and testing, or even before and after measures, so the criteria to determine if the the guarantee has been met or not is simply not cost effective in most cases.
In general, there are just too many other factors in play besides UI design that impact measures of usability. Thus, I don’t think its unreasonable to decline these sorts of requests for guarantees.
Finally, full time usability employees (or employees in any field for that matter) are not asked to guarantee a specific payoff from their work in order to get paid their salary, and I do not see any reason why consultants should either.
Unlike employees, clients simply do not have to rehire us if they are not happy with our work. They generally have to live with poorly performing employees. Consultants survive (or not) by establishing credibility in other ways – word of mouth, references, publishing well regarded books, etc.
What I offer my potential clients instead of guarantees is success stories.
For example, a large government client from a couple of years ago recently contacted me to work on a new release of an intranet I had redesigned for them previously. They had implemented most of my previous recommendations, and told me that the new design had been a big success:
User complaints had dropped from 10% to 0%.
Many more users had adopted the intranet as a work tool.
They had received an internal award for innovation.
Their intranet was now considered an example of best practices in usable design within their agency.
They found they had much more clout in recruiting other internal development projects to adhere to some of their standards.
When you can cite stories like this, and give references from clients who have benefited from your work in these ways, this is good enough to be expected to be paid for your time and expertise.
7. Social Media, with its almost instant communications threshold has seen rampant growth in the past few years. How are the Social web and instant connected communications like Twitter, FaceBook, Instant Messaging and the like changing usability and usability testing, or are they?
I wouldn’t say social media is actually changing usability or usability testing – they are just new types of applications, where all the same old principles and methods of usability still apply.
When the web first came along, people said the same thing, they seemed to think everything we had learned about usability in the context of traditional desktop applications had suddenly gone out the window, and we had to start all over again.
It just was not true.
People have not changed much in the last 50 years (or 50,000 for that matter) and all of the principles of usability are premised on understanding human information processing capabilities and weaknesses. Usability methods are similarly premised on how to measure and interpret human behavior.
While changes in technology do open up new possibilities, or place new constraints on the design of human-computer interaction (eg, limits of handheld mobile devices), most of the design principles and methods are universal and platform-independent.
For example, human memory constraints, and the advantages of consistency, are the same regardless of whether you are designing for a large screen, or a handheld device – you just have to apply the principles to different technological constraints and capabilities. Similarly, testing a design on either type of platform will follow a very similar if not identical methodology.
What I do find interesting about social media in the context of the business world is the opportunity to consider incorporating the concept of social media into corporate software to build community, share corporate knowledge assets more effectively and promote professional networking for the organization’s benefit.
I have just been helping a client consider how to use these new concepts to support their business goals through their intranet, and the possibilities seem to have a lot of potential. There will certainly be usability issues in designing these capabilities and usability testing will be just as important as ever. There is just not a whole lot of new principles to be revealed or methodologies to be devised. The same tried and true usability approaches will work just fine.
By the way, my opinion (and the opinion of many of my peers) is that Facebook has an abysmal user interface. They keep changing it, but apparently not based on any usability science, as it remains flawed, just in different ways with each modification.
Social media applications need usability input, just as any other type of software application.
If someone produced an application with the same functionality and a superior user experience, Facebook would be in trouble, just as IBM was in trouble when the Apple Macintosh came out.
8. In your opinion, what’s next for usability? What should usability practitioners and web site designers be focusing on for the future, say the next 2-3 years?
I think designing good UIs for small mobile/handheld devices is going to be a big area for our field, and applying social media in ways like I just described above as well.
In addition, I think we need to welcome and learn to work effectively with other professions and skillsets required to create a truly optimal User Experience: graphic designers, eCommerce “persuasion” experts, SEO experts, pay-per-click experts, etc.
The most powerful and successful user experiences require all these elements (and probably more), and it’s amazing what we can create working in interdisciplinary teams, that none of us could produce alone.
9. What are your plans for the future? What are you looking forward to doing next in your career?
Well, actually, I am working on a very exciting new project which I cannot talk about yet. Ask me again in about 6 months!
Thank you Dr. Deborah Mayhew!
It’s refreshing to hear that even with all the changes in terms of communications and the near-instant aspect of the new Social Media, the same basics of good usability and user-centered design apply.
For more information about Dr. Deborah Mayhew visit her web site at:
