Contents:
- Letter from the Editor: Learning from E-Commerce
- Upcoming Presentations - Feature: The Customer Sieve - More Resources
- From the Mailbag
--> Letter from the Editor: Learning from E-Commerce
Greetings,
We've been very excited about the research we've been doing on e-commerce sites. While e-commerce only directly affects a small portion of the sites on the Internet, lots of different sites share common traits with e-commerce.
For example, on an e-commerce site, a shopper has to:
- first find where the site keeps the interesting products, - then figure out which product they want to purchase, - then determine if the product actually meets their needs.
Sites other than e-commerce also have a similar flow. For example, when a senior citizen in New York City is interested in finding a government-sponsored home meal delivery program, they have to:
- first identify where the NYC.gov site has the home meal programs, - then figure out which programs actually deliver to their area, - then determine if the program has everything they need.
Much of what we learn from e-commerce sites will have direct application to many other sites, like NYC.gov. This issue's feature article, The Customer Sieve, is no exception. By studying where users are "filtered out" from making their purchases, we can learn what causes users to leave sites before completing their goals.
So, even if you don't directly work on an e-commerce site, I hope that you'll find the featured article and our other new research interesting.
As always, we want to hear your thoughts on our work. You can send them to uie@uie.com.
Jared
- o - o - o -
-> Upcoming Presentations
Here's where Jared Spool will talk about User Interface Engineering's latest research in February:
Berkeley, CA, February 5, 7:00pm Designing for Revenue: Using Research to Fulfill Business Goals. Sponsored by BayCHI-East (Open to the public) http://www.sims.berkeley.edu/~sinha/baychieast/spoolAnnouncement.html
Burlingame, CA, February 6, 8:30am Brain Sparks: (2 presentations with Erik Ojakaar) Being Content with Flash (TM) Feet, Yards, or Smoots? Exploring Web Analytics Today http://world.std.com/~uieweb/sparks.htm
Seattle, WA, February 13, 1:15pm You Had to Be There: Just-In-Time Information on the Web. Sponsored by Winwriters http://www.winwriters.com
Newton, MA, February 28, 8:30am Brain Sparks: (2 presentations with Christine Perfetti and Erik Ojakaar) Debunking the Myths of Usability Getting Your Content in Front of Users: Categories vs. Search http://world.std.com/~uieweb/sparks.htm
- o - o - o -
-> Feature: The Customer Sieve
We've learned that using a web site is a progressive process. Each user transitions from one stage to the next, as they work to accomplish their goal.
The most pronounced transitions we've seen are on e-commerce sites. When we watch shoppers focusing on buying a product, we can clearly see each stage and when the transitions fail or succeed. By understanding the stages and how they work, we can learn a lot about building better sites.
The stages act as a sieve: each stage inadvertently filtering shoppers out before they reach the next stage. By focusing on this filtering, we can see more users accomplish their goals.
To really see the sieve clearly, we study those shoppers who are completely intent on buying a product. They know what product they want, the web site has it, and they are ready to purchase.
Theoretically, every one of these shoppers should end up making a purchase. However, in our studies, sites frequently prevent these shoppers from completing transactions. By looking closely at the sieve, we can see where things are going wrong and get clues on what to fix.
The sieve can be divided into six stages:
1) The Home Page Stage 2) The Location Stage 3) The Product List Stage 4) The Product Evaluation Stage 5) The Checkout Stage 6) The Receipt and Acceptance Stage
[We should note that, while these stages are all specific to e-commerce sites, there are equivalent stages on other types of sites. We're currently researching what the similarities and differences are for non-e-commerce sites. Stay tuned!]
Let's look at each stage:
1) The Home Page Stage
When a purchase-ready customer comes to a home page, the goal of that page is to get them to the product they desire. And in our studies, the home page typically does a good job at this.
Users who know what they want are typically faced with three choices on an e-commerce home page: (a) use the search engine, (b) choose one of the featured products or (c) use the list of categories.
The very few people who choose (b) in our study go straight to the Product Evaluation Stage. We see only 1 in 237 home page visits result in a purchase this way.
Everyone else either uses search or the categories and progresses on to the Location Stage. For almost all of our purchase-ready shoppers, the home page does it's job quickly and efficiently.
2) The Location Stage
In this stage, the user either uses search or categories. Occasionally, they'll bounce between the two.
We group search and categories together because, from a behavioral perspective, they are essentially the same. In both cases, the user is trying to get to a list of products to choose from.
We've found that 9% of the users stop at this stage. If 100 users started the process, only 91 will continue from this stage.
When users can't identify the right category, they'll often go into search. When search returns a "No Results", they are stuck. (Users rarely try multiple searches. As we described in our article, Users Don't Learn to Search Better [http://world.std.com/~uieweb/Articles/not_learn_search.htm], multiple searches don't help.)
3) The Product List Stage
Shoppers who successfully transition from the Location Stage end up here. The activity for the user changes substantially at this point. The user is faced with a list of products for which they need to isolate the one they are most interested in.
About 8% of the users stop at this stage, failing to move forward and make a purchase. Of the 91 users who made it this far, only 83 will continue on.
It is here that we see the behavior we call Pogosticking. Pogosticking is when the user repeatedly visits a product description page, then hits the back button to return to the list.
The more pogosticking we see in a clickstream, the less likely that user will buy a product from that site. So, the best sites prevent pogosticking by providing as much information as they can in the product list.
Some users will end up going back to the Location Stage because none of the products displayed are what they want. These users are significantly less likely to end up purchasing anything than those users who move onto the Product Evaluation Stage. And many users give up at this stage.
4) The Product Evaluation Stage
This is the place where we see the most filtering of our users -- a whopping 25% stop here. Only 58 will continue after this stage.
While some of them stop because none of the products they evaluate fit their needs, most stop because they can't tell if the products are good enough.
Often, there isn't enough information or the *right* information isn't present. After observing hundreds of shopping expeditions, we've found people want to know lots of different things.
In clothing, for example, the fabric used is important. Is it cotton or a blend? We've had users who wanted to know the thread count for sheets.
Pictures play a big role here. Our initial analysis of some recent data suggests that the larger the picture, the more likely the user will purchase.
We've found that sites with similar product lines have dramatically different success rates at this stage. That tells us that the design of the product pages plays a huge role in whether people continue or not.
5) The Checkout Stage
People reach this page when they add a product to their cart and start the checkout process.
Almost everything that has been written about e-commerce usability focuses on this stage. And, we see a lot of drop outs at this point -- about 13%, leaving 45 people to finish the checkout process successfully.
But, the most interesting thing is that the vast majority of dropouts here come from two factors: required registration and poor shipping charge policies.
While we see usability problems at this stage, often manifesting themselves as input errors in the myriad of data entry fields that users need to fill out, we rarely see these problems causing users to abandon their carts. Users seem content to keep pounding away at the site until it finally relinquishes and processes their purchase.
6) The Receipt and Acceptance Stage
When we first started looking at e-commerce, we thought that after the user had completed checkout, everything was done.
We were shocked to find out that 11% of our users were so unhappy with a product they received that they returned it. In many cases, they didn't receive the product at all or it was the wrong product.
In a recent study, 8 out of 44 users told us they were unhappy with products they purchased but didn't want the hassle of returning them.
Some shoppers told us they returned a product because it wasn't what they expected. While these failures showed up in the Receipt stage, they are more likely failures of the Product Evaluation Stage -- the descriptions didn't set the right expectations.
Out of our original 100 purchase-ready shoppers, only 34 people actually got what they wanted.
As we learn more about the different stages and why users give up, we can hone our craft while, simultaneously, increasing the success of our users and our businesses.
[For linking purposes, you can also find this article at http://world.std.com/~uieweb/Articles/customer_sieve.htm]
- o - o - o -
-> More Resources
If you're interested in learning more about our detailed e-commerce research and how it might apply to your site, we've got a couple of options for you:
o There're still seats available in our upcoming e-commerce course, Designing for Dollars: Discovering How People Buy Online. (February 4 & 5, Burlingame, CA) This course is 2 days that are jam-packed with our latest findings.
o We also will be presenting our latest findings at the UIE Research Forum in Burlingame, CA on March 21, as part of the User Interface 6 West conference. You'll be getting details very soon.
o We put a lot of our findings in reports, white papers, and articles on our site.
Visit http://www.uie.com to learn more about these resources.
- o - o - o -
-> From the Mailbag
Here's something that caught our attention. On the evolt discussion list (http://www.evolt.org), Joel Canfield posted the following inquiry:
Just completed our first ever intranet usage survey (very informal and totally voluntary) and while most of it makes sense, one item made me second guess myself. (Note: I'm making too much of this, but it's been a tough day and I need a diversion.)
One respondent volunteered the opinion that the home page is 'unorganized and naive.' My immediate response is, "It's functional, not aesthetic." But I'd still like your opinion. I've posted a screen shot at
http://www.magisnetworks.com/images/netmagis_screenshot.gif
The 'unorganized and naive' comment jumped out at us. What would a user mean when they give this critique of a home page. Jared Spool looked at the screenshot and crafted this response:
Did your usage survey talk about the functions that users want from the intranet?
When we've looked into intranet usage, we've found that there is a real 80/20 rule that is followed: 80% of the time, users come for 20% of the functionality or content. (This is optimistic: we've recently discovered that on the Microsoft.com site, 80% of the time, users come for 1% of the content.)
My first reaction to the screen shot was that I couldn't tell what that 20% of the functionality and content is. This might be the basis behind the "unorganized and naive" comment.
If you knew what the most frequent/important functions that people needed from the intranet, you could put direct links to those on the homepage. (Don't worry about longer pages -- there's no evidence to suggest that long homepages are problematic. In fact, most of the evidence we've seen to date says the opposite.)
A good example of moving functionality to the home page is the US Airways site (http://www.usairways.com). The three most common functions (Booking/Pricing a trip, Checking on frequent flyer miles, Checking flight status) are accessible directly from the homepage without having to click and go to an intermediate page to start.
The next most important functions are listed as direct links under each category. This design allows about 40 of the most important functions to be immediately accessible.
In an intranet setting, users most often come to the site with a specific mission in mind. Often, its a mission they've repeated many times before. By studying the usage patterns of the functionality and content and by talking to the users about the missions they are trying to accomplish, you'll get a good idea as to what the most important functionality/content is for the site.
We were excited when Joel wrote this enthusiastic response:
BINGO! That's exactly what's wrong. There are a couple areas that are far and away the most used and appreciated, and they fade into the noise. Thanks for the perceptive feedback. [...] Excellent usability and design info.
It's nice when our research experience helps us handle a real-life problem.
To subscribe or edit your listing go to:
http://list.adventive.com/SCRIPTS/WA.EXE?SUBED1=friends-of-uie&A=1