In its simplest terms, building commercial websites is about marrying something the customers want with something the owning companies want to sell them.
It isn’t as easy as that, though. Often companies must speak to widely divergent audiences who have widely divergent reasons for coming to their site. It doesn’t help that everyone within the owning organization has a different idea about how the customer thinks and what is most important.
Mutually exclusive ideas can all be wrong, but they can’t all be right. So how can we know that our solutions are addressing the right problems?
1. Ask your customers what they think
There always seem to be good, tempting reasons to capture the preferences of your customers…. There never seems to be enough time or enough money… Which is why it takes focus and discipline to resist the temptations.
We will almost always know more about our own products and business goals than our customers, which is precisely why our judgment is so unreliable.
The Curse of Knowledge
In Made to Stick, Chip and Dan Heath assert (and go on to demonstrate) that once we know something, it’s difficult to remember what it was like not knowing it.
For nontechnical people, this is why talking to IT can sometimes be like pulling teeth.
And before we’re too hard on IT, keep in mind that we all have interests that are boring and unintelligible to someone.
Similarly, our customers don’t care what we want, either, except when our demands unnecessarily encumber their ability to get what they want.
Knowing what we want and what we do makes it difficult for us to see things from our customers’ perspectives. The proper research can help us get out of that rut.
Doing without research means making sacrifices
With the right experts on staff, a “dead-reckoning” approach to website design can vastly improve the look of an inferior website.
However, the right experts will also know that proper research can do more than inform the design choices – it can be the difference between creating something that people can interact with something and something they actually want to.
Consequently, it can also be the difference between successfully meeting the right business goals and getting hammered in the marketplace by people whose actual (not perceived) wants and needs necessitate a redesign in a few years.
Examples of Preference Research
- Surveys
- Preference interviews
- Focus groups
2. Watch what your customers do
Irrespective of what they say they like, people behave in peculiar ways.
It is not uncommon in usability lab studies for participants who could not accomplish many goals set forth in the study to rate the overall experience highly in the post-test questionnaire.
This doesn’t make surveys or questionnaires unreliable; they are good for their intended purpose. It just demonstrates that data that indicates preference is not sufficient as an evaluative study – and certainly not a generative one.
Usability lab studies are but one method to observe customers. There are many other methods that provide insight into how customers behave both online and in the “real world,” such as log analyses and ethnography.
Examples of Evaluative Research
- Search and log analytics
- Card sorting
- Usability lab and field studies
3. See what you see
French philosopher Emilé Chartier once said:
“Nothing is more dangerous than an idea when it’s the only one you have.” (Hat tip to Roger von Oech)
I once spent a lot of time putting together a solution for a very complicated problem. The solution came in the form of a plan and interface architecture for a Web application. After making some adjustments based on feedback from some colleagues, I was very proud of it.
The problem was, when we put it in front of our audience in a lab study, no one could understand by its appearance how and why they should interact with it.
I tried a few times to “tweak” it, but to no avail. Even though – after we explained what the utility of it – our audience loved the idea and could “learn” it, they couldn’t give us any useful feedback other than to let us know they didn’t understand how it could be tweaked to make its function apparent.
After just a few users, we were reasonably sure that it couldn’t be salvaged by simply tweaking the interface.
So we simply blew it up.
We started over. Only by doing this did we find a solution that actually worked. But for the first few iterations, I was holding out hope for some derivative of my first attempt.
We are all susceptible to falling in love with our own ideas, but there can’t be any sacred cows when it comes to usability studies or any other kind of research. When we discover that people prefer one thing over our better sensibilities say they should, or if they simply cannot understand how to use something we spent a long time creating, we ought not let that get us down.
After all, by ruling out a possible solution through this research, we’ve gotten one step closer to finding the right solution.
What’s more, we can be comfortably certain about that our investments are safe, because we tested our assumptions and were prepared to deal with the ramifications of all of our research.
Related Links
How Do You Spell Succes? (A “Made to Stick” review)
Mental Models: Aligning Design Strategy with Human Behavior
Card Sorting: Designing Usable Categories
Rocket Surgery Made Easy
No Comments | Trackback | Categories: 1000 Ft. POV, Research, Testing | Email This Post
Leave a Reply