Building a better mousetrap is not just a matter of making a device more effective for customers who hate mice. The trap has to be used, not just be usable. A major difference between usability in the sense of functionality and usability in the sense of creating things that are used has a lot to do with people’s expectations. Indeed, expectations affect usability in several ways. Let me state three ways:
- Ambiguous—or worse—misguided expectations lead to non-use.
- People’s expectations affect the way they experience a product when they do use it
- A product or service that delivers on people’s expectations will, by definition, be more satisfying than one for which expectations and experiences do not align well.
Let’s consider some examples.
Misaligned Expectations and Use
As a ”lifer” in the marketing research industry, I’ve managed qualitative and quantitative research on the client side (Amoco, Hamms Beer, Coca-Cola) and the vendor side (most notably as president of C&R Research) for almost forty years. In retrospect, perhaps the single, most pronounced source of product failure I’ve encountered during that time has to do with products not matching people’s expectations.
In a short article on the Financial Page of The New Yorker magazine ( May 28, 2007), James Surowiecki said that the “… spiral of complexity, often called ‘feature creep’ in many consumer products, costs consumers time and… businesses money. A recent study by Elke Den Ouden, of Philips Electronics, found that at least half of returned products have nothing wrong with them. Consumers just couldn’t figure out how to use them.”
Surowiecki continued by noting that part of the problem is due to the “ internal-audience-problem: the people who design and sell products are not the ones who buy and use them.”
I am tempted to note, here, that Surowiecki’s comment corresponds to one of my own observations, which is that usability research has historically been prone to conducting studies among its own (or among people versed in like-minded disciplines) so that results are not as readily projected to the end-user populations as they ought to be. I believe marketing research has a somewhat stronger record on this issue, but that will be the subject of another article.
Surowiecki went further to point out that consumers appear to be as enamored of bells and whistles as engineers are. (Most of us would love to see what a cold fusion bug spray could accomplish.) So we are easily seduced into purchasing devices more complex than we need, but subsequently are unwilling to learn how to use.
It’s not a new problem—recall that in the 1980s, having a videocassette recorder that incessantly flashed 12:00 was a common occurrence in houses where owners had no idea how to program the correct time into their gadgetry. And pity the point-and-shoot photographer who purchased Photoshop just to print the vacation pictures.
Clearly, manufacturers are, to some extent, conducting their usability work with people who are more expert than the garden variety consumer who ends up buying the cockamamie thing.
But in fact, the failure to manage expectations is at least as big a culprit as the failure to make products with practical usage profiles.
The problem is one of marketing the sizzle of über-technologies to the point that people expect to be wowed by the blinking lights, but instead, are only stymied by their complexity; expectations and reality just don’t coincide. Somehow companies need to offer potential customers an appealing but (ital)realistic primer on what to expect in a product before they buy it—lest they be disappointed by what they get.
Here’s a good example. A meat packing company marketed a canned “beef stew” that never did well in the market place. It was a good product—easy, nourishing, good tasting—but many consumers faulted it for being too thin or insubstantial to be a really good stew. The company’s first reaction was to make a heartier product—stew you could eat with a fork—but that wasn’t a very practical solution.
Instead, they repositioned the product such that people expected something a little less viscous. They renamed their stew “hearty soup” and it started to fly off the shelf. Same can, different label—one that created expectations that the product more than met. As soup, this stuff was really good.
Clearly, managing expectations can help people find satisfaction in a product by keeping their expectations and experiences in moderate proximity.
Expectations Affect Perceptions
Another important thing about expectations, beyond the fact that they contribute to one’s satisfaction with a given product or service, is that they have a direct bearing on how people actually experience the things they use.
The food industry is rife with examples.
A cheese company wanted to know if they should package their cheese in highly branded, uniformly shaped blocks with sophisticated graphic packaging, or if they would be better off offering their product cut into blocks of random shapes and sizes with more hand-made packaging and brand designations such as one might find in the deli-section of the grocer store.
To answer that question, we asked naïve respondents to tell us which of two cheeses they liked best on a number of dimensions: depth of flavor, pleasant texture, rich and mellow aftertaste, etc. We did not ask people anything having to do with what they thought of the packaging. From their point of view, the study was a taste test.
The two cheeses, opened in front of the respondents, had in fact been cut from the same wheel, but one had been packaged in the major brand motif while the other had been packaged in the deli-counter style. Same cheese from the same wheel, but the deli-counter product won on all counts. The promise of deli-packaging was so strong that it literally gave people a different experience—incidentally one that justified a price sufficiently higher to cover the added cost of random packaging.
One might look askance at such research as being bent on deception, of facilitating the manipulation of consumers in the hopes of pursuing a greater profit. But, I might remind my reader that restaurateurs have, for centuries, understood that the elegant plating of foodstuffs is an essential ingredient of success—not because it hoodwinks a diner into believing they are eating something better than what is being set before them, but that it makes the food more appealing to one’s sensibilities and is appreciated all the more for that fact. Positioning one’s product in a light that makes it taste better is not, in my mind, anything short of good management of expectations.
The point is that expectations don’t just define people’s approach to usage; they help define the experience itself. It may seem absurd, but the truth is that people literally tasted things differently when they approached a product with different expectations.
Managing expectations can be a real balancing act. Promise too much and people are destined to be disappointed. Promise too little and people won’t try your product.
Sometimes the “promise” can be off center. A large hospital redesigning its lobby found one design that was clean and efficient but was also so clinical as to be deemed “heartless,” while one design that was elegant and lush was viewed with suspicion as promising excessive bills and a mentality of form over function. Thus, the hospital began the balancing act. They settled on a modern but comfortable presentation that seemed to assure people of a good measure of warmth without undue concentration beyond medical functionality. Did the lobby provide better care? I suspect that in some small way, it did—people who are at ease with their medical environment heal faster than people who are nervous—for whatever reason.
Expectations in Usability Research
Effective usability research should have an expectation component to it— something that takes into account the variance between the promise of a product and what it actually delivers. Armed with that information, sponsor companies can find ways to tweak their products, or the implicit or explicit promise of those products, so that people’s expectations and experiences align.
Consider three different visitors to a hotel website: a conventioneer, a vacationer, and a stranded commuter. Each will approach the website with different needs and, accordingly, different expectations. The conventioneer will expect to be told about meeting rooms with good acoustics; the vacationer, easy access to tourist sites; and the stranded commuter, a quiet room and a decent sundry store.
These occasion types, not people types, define visitor needs and expectations. The same individuals may be commuters or vacationers or conventioneers at different times (just as they may want hearty soup and stew at different times), and their needs and expectations will change considerably from one occasion to the next.
If you had a typology of user types, each with an associated set of typical needs and expectations, you should be able to ask people the nature of their situation (conventioneer, vacationer, or commuter) and branch them directly to that portion of your website designed to address the sort of questions they are most likely to have, rather than having them wade through a decision tree that requires them to winnow down their options from a vast, all-purpose array of possible issues. To be sure, assumptions made as to one’s needs under a given usage scenario must not preclude people from finding unique solutions to their own personal requirements (the entire tree needs to remain accessible), but the order in which particular amenities might be presented may well be designed to facilitate finding the optimal solution in fewer steps.
So, occasion (or need-state) segmentation—the process of understanding the needs for which people are most likely to seek satisfaction under common usage scenarios—can be very helpful in designing websites that are more efficient and, accordingly, more user-friendly, by helping the website itself anticipate users’ most likely expectations.
At the very least, test participants should be asked what they expect from a subject product before they begin the research process, and again at the end, how the product performed vis-à-vis those expectations. The promises implicit in the product at hand may then be better understood, and perhaps adjusted, to help manage people’s expectations more appropriately.
A Practical Reason
Even if you don’t buy the substantive character of this argument, let me leave you with a political one. Like it or not, usability and marketing research are inextricably linked in the corporate scheme of things in most institutions, all too often with usability being a second sister on the team, at least when it comes to carving out the corporate budget.
My sense is that usability is a necessary but not sufficient part of the process of bringing something new to market (or fixing something that isn’t working well). It may logically be a predecessor to marketing research given that unusable things are destined to be marketing failures. But it should not be conducted without regard to the marketing issues it raises and the learning that travels inductively and deductively along the process.
One of the most important charges of marketing is that of managing people’s expectations. (What else is advertising supposed to do?) To the extent that marketing people can be brought to understand that good usability research speaks to the issue of managing expectations, it will be seen not as merely an adjunct to product design, but also as an important tool for understanding and enhancing the process of effective marketing.
I would love to see these two disciplines work more closely together—each respecting and utilizing the techniques and substantive insights associated with their pursuit. The AMA (American Marketing Association) should have a usability division and the UPA should have a marketing division. I believe the first place I would begin concentrating the concatenation of these two disciplines is in the arena of measuring and managing expectations.