All About Website Usability Blog – Holly Phillips

Additional types of usability research
September 24, 2009, 3:55 am
Filed under: usability testing

The line between market research and usability testing is a very fuzzy one.  In addition to the types of usability tests I’ve discussed earlier, the following types of market research is also often used to help get a richer picture of customer needs, thoughts, and actions. 

Different tools and techniques are required depending on what the research is trying to achieve. To try to understand what customers feel about a certain ad, what value they get from training courses, or why they believe that our products make their lives easier, the first three types of research (qualitative) listed below may be most appropriate. If the question is what impact a new ad will have on our sales, how many people will take a new training class, or how much we should charge for a new product, the last two techniques (quantitative) are more appropriate.

  • Customer Visits:  Anecdotal / qualitative, not projectable to the general population, may be Agilent-biased.  Exploratory and experiential, not definitive.
  • Individual Interviews:  Anecdotal / qualitative, not projectable to the general population, may be interviewer-biased. May be conducted by phone or in-person.  Exploratory and suggestive, not definitive.
  • Focus Groups:  Anecdotal / qualitative, not projectable to the general population, may be group-biased.  Exploratory and creative, not definitive.
  • Descriptive Surveys:  Quantitative / projectable, objective. Descriptive, not predictive.
  • Choice Models:  Quantitative / projectable, objective. Predictive of trade-offs and purchase behavior.
  • Controlled ExperimentsQuantitative / projectable, objective. Predictive of trade-offs and purchase behavior.
  • Customer Satisfaction or Loyalty Studies:  Quantitative / projectable, objective. Typically done longitudinally (repeated over time) to provide trend information.  Many well-accepted series of questions to obtain measure of satisfaction or loyalty; often extrapolated to predict purchase intent.

The next time you’re asked to run a “usability” study, take a close look at the research questions and see whether one of these types of research might just be more appropriate.


Nice reminder from Orbitz
September 15, 2009, 4:31 am
Filed under: confidence, visual design

I was shopping for plane tickets today and ran across a great feature from Orbitz.  They must know that people check lots of different websites for plane tickets (witness the rise of travel site super-aggregators who act as intermediaries to submit your travel dates to the next level of aggregators like Orbitz).  Anyway, I wasn’t ready to buy just yet and closed the Orbitz site.  This is what came up in its own window once I closed the Orbitz site:


We’ve all seen the last-minute marketing efforts upon exit of some site (rethink your order and save $10, today only, etc).  But what strikes me about this one is it isn’t trying to force a decision from me.  Instead, it’s simply trying to stay top-of-mind as I continue my shopping.  The wording is friendly and not pushy (“Still shopping?  Keep this window open to hold onto your trip info”).  As a customer, I actually appreciate this type of message — clearly enough to blog about it, and also enough to actually do what they wanted me to do (keep them in mind as I continue to search).  An additional unwritten implication is that they must stand behind their prices to do this, since they know I’m price-shopping and they still want to have their deal front-and-center for comparison purposes.  Nice.

Catalog page filtering: Make sure your filter levels match your customer needs
September 10, 2009, 4:49 am
Filed under: form design, usability basics, usability testing, visual design

I need (ok, maybe not NEED) to buy a new TV, so was poking around to find the best reviews and deals.  I need the TV to be a very specific size to it will fit inside an existing cabinet — not an unusual request these days.  But many of the sites I looked at only let you filter down to a range of sizes, not a specific size.  Depending on the number of TV’s offered, this can range from annoying to downright tedious and unfriendly.

In my case, I can take a 40″ maximum screen.  I’d like to look at a few 37″ ones too.  Here’s my experience:

  • Amazon
    6,584 TV’s (they’re not called Amazon for nothing!).  The closest I can narrow down is either 30-39″, or 40-49″.  There are 1200 TV’s in the 40-49″ range, 1358 in the 30-39″ range.  I don’t really want to narrow it to LCD (I’d like to be open to LED and Plasma at this point), but I will.  Now there are 679 in the larger screen range.  Better, but still way too many for me to scan through to find just the 40″ ones.  In fact, I’m stuck.  What do I end up doing?  Scanning the first two pages for 40″ TV’s and stopping after I’ve seen 4 or 5.  But I’m wary that I may have missed the “perfect” 40″ TV because it’s buried in the pack somewhere.
  • Best Buy
    Similar experience:  61 TVs in the 40-49″ range.  If I actually go through the whole tedious exercise of looking at the entire list, I see there are only 13 40″ TV’s.  Sure wish I could have narrowed it down that way.
  • Samsung solves this problem in a great way – they use a slider for screensize that lets you set the field as wide or narrow as you want.  Kudos to Samsung!
  • NexTag gets the award for the most annoying screensize selector:  they have a 30-40″ category and a 40-50″ category.  Great – now I have to look in BOTH categories to find the 40″ TV’s, because I don’t have confidence which one they’re in.

Lesson:  Samsung has done their homework.  They know their customer, and know that people often need to shop for TV’s by screensize, so have provided an easy way to do that.  In B2B applications, if you know your customers often need to shop to meet very specific specs, make sure those specs are easy to narrow down.

Eye tracking: Bing vs Google
September 3, 2009, 4:52 am
Filed under: usability testing

The folks at UserCentric just did a cool eyetracking study to see if people look different places on the Bing search results page vs the Google search results page.  See the details and heatmap at UserCentric.

They discuss the similarities and differences and it’s definitely worth taking a look at their findings. 

Thig brings up an interesting question for corporate internet sites:  if people scan Bing and Google pages very similarly, has that become the norm?  Ie, are people conditioned to scan a search results page in this way regardless of whether it’s Google, Bing, or  If so, would it be better for corporations to model their search results pages after this norm instead of trying to come up with something better?  (Where the web is concerned, sometimes intuitive is better than best.)  My hope is to test some of our corporate search results pages (which have some distinct differences from this pattern) and see.