SharedBook: On Demand

Abandonment Issues

Working in eCommerce, I’ve always been trained to believe that abandonment is a bad thing.  Our goal is always to increase sales and make the most of the traffic we have.

The logic is straightforward.  If you look at the page-to-page progression through a site (the “funnel”), the more users you can get to move on to the next page in the process, the more sales you’ll have.  Focus first on the pages that have the highest abandonment (thus the lowest retention) and see if you can increase the number that progress forward.

But I am starting to believe that this approach is a bit too simplistic.  It occurs to me that this concept borders on treating each page like a beauty pageant.  The user (“judge”) gives the thumbs up approval for each individual page design by agreeing to progress forward.  Or, he or she shows their disapproval by exiting the process altogether.  Eventually - each page is designed and redesigned to gain the highest overall approval rating (aka conversion).  But does this create a cohesive design?

I’m determined to come at this from the user experience standpoint.  As a result, I’ve folded two of my own concepts here.

First one is that no matter what your product is, there will be users who don’t want it.  For them, the best experience is to learn this as quickly and painlessly as possible.  Thus one of my key measures is how soon people abandon.  If the majority is half way through the process then either that page has a problem, or it is the first time your tell the user a key piece of information that tells them the product is not for them (like, say, the price).

Second is that no matter how much a user wants your product, they won’t always be ready to buy it right now.  These folks usually abandon on the last step (the payment page).  The offset is that you can frequently entice them back (ever get a coupon for something they carted but didn’t ultimately buy?).

So my hybrid recommendation is to focus on the middle of the process.  Assume these users are interested and willing to go along.  If they abandon half way through, check to see if you are disclosing some key fact at this stage.  If so, try to disclose it earlier.  If not, then you’ve probably found the page that needs your design help.  Improve the conversion of that page and then repeat the entire process.

What's in a Name

Unfortunately, I cannot skip the opportunity to quote Shakespeare for this entry.

“A rose by any other name would smell as sweet”

Of course it would!  But it probably wouldn’t sell very well on your Web site.

You can sell roses on your Web site, so long as folks know what a rose is.  It’s all about setting proper user expectations.  Seems simple enough – just call it a rose and you’re set.

But suppose they’ve never heard of what you’re selling.  Suppose it is something that never existed before.  You’ll have to spend some time explaining the item well enough that the user not only knows what it is, but also what it isn’t.  It is important to get the user to understand all of the benefits and uses an item will have, but not overstep the mark and have them think it does more than it actually does.  As an example, if you were going to explain a rose and how wonderful it smells you might need to ensure the user knows that it has an expiration on it.  A 3 week old rose, by every name, no longer smells sweet!

So what if you are not trying to sell an item, but rather a process?  What if you are in charge of the information design for a Web site and it is your job to get a user to understand what happens next?  What if you are trying to sell the user on clicking a button?  It’s all about setting the user's expectations which, turns out, is almost no different from selling anything else.

The first rule is to try to tap into common terminology so that you don’t have to explain too much.  A very common example is the “Add to Cart” button.  Surely the very first Web site that used this phrase had a LOT of explaining to do.  Who back then would have been thinking of a grocery store shopping cart on the Internet?  But now that it is so common, most users know exactly what to expect when they click such a button.  Tap into the user expectations that this phrase evokes by simply using it if it applies.

But not every site does common things and perhaps your site does something unique.  What if you do need to give the user a button that is unlike any other Web site?  This requires more effort and diligence.

Of course, you’ll want to explain it.  You can put some text next to the button saying exactly what happens (and perhaps what won’t).  But know that the average user will not read it.  So now what do you do?

1 - Give it a name.  Try your best to be descriptive without being wordy.  Hope that it will stand on its own and users will understand what it does – or at least be curious enough to try it. 

2 - Keep it consistent.  No matter how much you want to try using different terms in different places on your site – fight the urge.  If you keep changing the name, no user will ever know what is happening.  If you discover it needs to change, then change it everywhere, all at once.  And if that is going to take some time, then let it take time.

3 - Test it.  You’ll be amazed at how an average user interprets the text on a button – no matter how clear you think it is.  Put it in front of someone (anyone you can find who’s game) and ask them to tell you what they think will happen when they click it.  If it turns out you set the wrong expectations (and chances are you have), then ask them, “What would you think it does if it said ____ instead?”  Use this to find a better term and then make a change – but make it pervasively.

Setting the right expectation during a process is critical.  And, in my experience, no one gets it right the first time. 

So what’s in a name?  Sometimes everything.

The Goldilocks Principle

So I've always been curious about the story of Goldilocks and the Three Bears.  It fascinates me that there is a story for children about a girl who breaks into a home, rummages through the family's possessions, and indulges herself in selecting "just the right one" of anything she wants.

The saying goes "beggars can't be choosers," but apparently intruders can do whatever they want.

Nonetheless, the story is quite enduring.  And it does appeal to me on a user experience level - the concept that everyone seeks the product that fits them "just right."  Unlike, say, clothes that are tailored to a hundred different sizes and shapes, the Goldilocks story divides everything into three manageable categories - the two extremes and the one right in the middle.

When it comes to designing Web sites, I try to follow a similar triage.

Say you are trying to build a design for a particular process.  There are probably hundreds of choices and different designs you can try.  How do you narrow the field?  I usually look at extremes. 

First, take the process and suppose you have a "surgical user" - one who knows exactly what he/she wants and will drill right to it.  What would be the fastest and most direct process?

Compare this with the "accidental tourist."  Someone who happened upon the site and is, for the moment, interested in seeing what you have to offer.  What would be the most engaging process?

Now apply the Goldilocks principle and aim toward the middle.

Next, apply the same principle to compare the "devoted reader of every footnote" versus the "60 mile-per-hour click demon."  How much text do you put in the page to satisfy the extremes?  Now - find the middle.

Next you could compare the "technically savvy" daily Web user versus the "Sunday driver."  How much do you try and impress with the latest rich media without intimidating?  Again - aim for the middle.

Depending on the specific application, the important "extremes" will vary.  But by applying this principle, eventually a design that balances everything will emerge.  It will likely blend the best and worst of everything and be the most generally appropriate site for your users.

Of course there is a down-side to applying this principle.  During the discovery process you become very aware that you can only hope to please most of the people most of the time. 

If Door Does Not Open – Do Not Enter

This past Christmas my parents gave me a day calendar of "The Stupidest Things Ever Said".  A week ago, the day's quote was from a sign posted on some door somewhere:

"If door does not open, Do Not Enter"

Should I be concerned that this sign makes a lot of sense to me?  I know, I know ... the funny thing about it is that if the door doesn't open, then you can't enter.  Thus there is no choice in the matter. 

But choice is at the very heart of user experience.

I consider a very analogous web design (that is getting more and more use) - the grayed-out button concept.  You'll see it any time you install new software on your PC.  There is usually a set of software terms that they want you to agree to before you are allowed to play.  So they show you a big scrollable legal text block and a 'Continue' button at the bottom of the page.  But the button is gray, and clicking it yields no action.  Only when you notice the small check-box above it that reads "I agree" and click it that your continue button becomes active.

But suppose this were the first time you stumbled upon this design?  You might think there was something wrong with your PC.  You might think that there was something wrong with the software.  You might think all kinds of things, because there are too many possibilities for something being broken or wrong.

Telling the user that the button is intentionally deactivated is vital.  They have to understand what is happening in order give their acceptance.  Otherwise, they are lost and the software goes unused.

A good rule of thumb is to bear in mind that nothing is universally clear.  And when a user is on a mission to get what they need, sometimes they need a little extra help.  Even if it means stating the obvious. 

 

Should Subway Trains Have Windows?

For those of us who work in the NY office, riding a subway train is a common thing without much to consider.  However, I come from the New Jersey suburbs and it was something I had to get used to.  If you'd asked me a few years ago to imagine riding a subway train, I'd probably picture a train parked in the station with its doors open.  But now that I ride them every day, I see them from a different perspective.

The majority of time spent on a subway train is inside a tunnel.  There isn't much to look at -  you can clearly see lots of conduits and other functional "behind the scenes" stuff.  Mostly what you see is dust and dirt from a lack of upkeep.  Surely no one is assigned to clean a subway tunnel and they certainly aren't anything for a Transit Authority to be particularly proud of.

So I was thinking - why do subway trains have windows?  Wouldn't it be a better user experience if we didn't have to look at a dirty tunnel?  Trains could use the extra wall space for advertising and other revenue generating opportunities.  These MUST be nicer to look at than a tunnel.

Once you arrive at a station, the doors open - so the rider would be able to see outside ONLY when they needed to.  This is such a win-win - why do subway trains have windows?

The answer is that this entire concept is in direct conflict to one of my personal rules of user experience - NEVER hide answers from users just to shield them from the truth. Users are curious human-beings. They want to know about their surroundings. If you take away their awareness they will not be pleased (not to mention a bit motion sick).

This is quite similar to a request made to me many years ago.  A Web site I designed allowed users to find locations worldwide.  For the most part, services were available in every country, but there were some countries without.  I was asked to remove those countries from the drop-down list - the justification being user experience.  Why let a user tell you they want service in a particular country if you are only going to tell them you don't have it?  Isn't it better customer service to show them a list with only positive results?

It is the same issue as windows on a subway train.  Without a full list, users become disoriented.  They don't see the country they are looking for and wonder if there's been a mistake.  Perhaps it is listed somewhere else?  Maybe it's spelled wrong? 

The answer was to leave the full list and tell users that we didn't offer service in those countries.  In other words - full disclosure.

The best experience is, at least, an honest one.

Lessons in Patience

Working for an Internet start-up like SharedBook is always exciting.  At any one time there are so many projects underway and new products about to launch. Even more so as a Product Manager - it's like a candy store!  So much to choose from.

But I have to remind myself to have patience.  Though change is good and may be exciting, too much change too quickly doesn't always create a good user experience.

Take, for example, our Checkout process.  Last week we released our new PayPal compatibility (Mor already wrote about this - The First Tester).  This represents the third round of changes to our checkout process since I started a brief 6 months ago.   And there are more changes in store for our checkout process coming soon.

So why not do ALL of these changes at once?  Wouldn't it be more exciting and dramatic to reveal everything as one major upgrade?  My enthusiasm says YES!  But my experience says different.

First off, making major changes in an established application has associated risks.  There are so many things to consider and changing lots means testing even more.  So it is better to take smaller steps and ensure each one is on solid ground.

Secondly (and more importantly), users don't necessarily want to see everything suddenly change.  Has it ever happened to you that you were on a Web site one day and you worked your way to the very end only to realize you forgot something?  Imagine you come back the next day to finish and nothing is as you've left it.  I've had this happen before and it breaks my stride.  Suddenly my confidence is reduced - regardless of whether the change was an improvement.  As a user my first reaction is, "This isn't working right."  I usually stop and regroup.  Sometimes I'll even decide to come back later.

Thus we have the several smaller steps we've taken.  And soon we'll have the new ones too.   Each step brings us forward and makes our user's experience better and better.  It just takes a little patience.

Navigating the Globe?

So I'm ever conscious of the lingo that people acquire as part of their work.  I wrote about learning the lingo back in August.

One very important term (and even more important concept) is what we call "Global Navigation."  Is this some kind of GPS?  OK, I exaggerate.

I guess the term Navigation is well know by people.  The term "Global Navigation" applies to any of the navigation for a site that appears globally on every single page of the site.  "Home" is the most common global navigation element.  Everyone has one on every page in their site.

What seems to be less known is the absolute rule for global navigation - keep it consistent.  It seems that a common mistake for Information Architects is that they take each page in a site and design it as it stands alone.  This means that quite commonly the global navigation of the site changes position, contents, look or even all three.

Perhaps this yields very smart looking individual pages.  Maybe it allows different sections of their site to have a different look and feel to suit a different marketing segment.

But the bad side to it is that it confuses users.  When designing an entire site, there is a very critical item that must be kept in mind.  Users won't know exactly where things on your site are.  As a matter of fact, they likely won't even know if your site has what they are looking for.  Users will look around, try different things to find what they are looking for.  Most of the time they won't find it and so they will want to try another option.  Imagine what it's like if the options keep moving around the page, or magically appearing and disappearing from page-to-page.

I always think of it like this.  Suppose someone came to your house and reorganized everything.  It's always great to have a tidy home and you might be thrilled.  But when you first go to your dresser to find the right clothes to wear you might not know which drawer to look in.  Imagine that you start with the first drawer and open it.  You don't find what you are looking for, but you notice that all of the other drawers have suddenly changed their places.  You then try to open the second drawer and the first drawer moves somewhere else.  How confusing would it be?

The Holiday Stork Cometh?

Very shortly SharedBook will be launching expedited shipping options - just in time for the holidays. This feature is a significant upgrade to our shopping cart, allowing customers to choose overnight, second day, and ground shipping with full tracking (in addition to our current free standard shipping).

I feel a bit like the proud father waiting for the blessed arrival. Not just because this is my first major project to be released, but because it is one of those great features that touches all aspects of our business. To really see how impactful this new function is, I tend to look at it backwards.

Obviously the biggest benefit is to the customer receiving their book as fast as possible. So the delivery and associated tracking capability they'll have is vital. Just preceding this is our operational staff who oversee the production of the book and all the new accounting processes to monitor the costs of such a service.

Before this can take place is the development of the web site changes to quote customers the price (which depends on the number of books and where in the world they'll be shipped). And even more complicated - the estimated delivery date for the shipment so users can make an informed decision.

Prior to that is the marketing message to bring customers to our doors. From emailing registered users about the last order date to ensure delivery in time for Christmas to working with our Business Development group to help our partners get the word our about the value of their content as the perfect gift book.

Come to think of it - this feature has many expectant parents. Cigars for everyone!

Simplicity Can Be Very Complicated

This past week has been spent putting together designs and concepts for some of our new partners.  Specifically, creating the bridge pages that lie between our partners' sites (where the content originates) and our Reverse Publishing Platform where a professionally printed book is created.  These pages bridge the gap between the two systems, but serve a key function to our consumers - customization.

The SharedBook platform allows users to pick and choose through many different options on a bridge page to create a one-of-a-kind book built on demand.  In order to develop this process (which is different for each and every partner) we think through what a user might want.  Early in these conversations, we start to list all of the possible combinations - from lists of news articles and blog entries stored in partner databases, to hi-resolution images that will make beautiful covers.  We review them all.

Then comes the time to put together the User Interface that will offer all of these options. We envision what can be comprehended in a simple interaction and determine what is too complicated.  We combine different options together to make nice "bundles" and continue refining things until we arrive at a simple page or two.

This is when I always stop myself and say, "Why did this take so long?"  The resulting page becomes so obvious that it is hard to believe that there was ever a question of how to design it.  Anyone who looked at it wouldn't give it a second thought.

This is when I know we've succeeded.  We've arrived at the goal that is crucial to a user experience.

Should I quote Shakespeare - "Much Ado About Nothing"?  No, it's more appropriate to quote Leonardo da Vinci - "Simplicity is the ultimate sophistication."

To Basic Backs

In my years of information design, I have learned to take designs with a grain of salt. That is, though you may spend many days working out an elegant flow of pages and processes, users may decide to interact with your site in a very different way.

Take the example of any travel site. Every one of them asks the user to give some details about Where and When they want to go, followed by some kind of “Search” button. They click and off go hundreds of calculations to find just the right price for just the right product. Results are displayed.  But then the traveler will invariably want to compare the quotes for going to a different place on a different day. So travel sites always give the user a link to “make a change” and search again. Some sites have four or five of these links to facilitate surgical changes of just the right element for the new set of calculations.

But users, as I’ve observed, hardly ever use these links. They invariably hit their back button and attempt to rerun a new query. One would think this is a simple matter, but it can be very complex.

In many cases Web sites pass data from page-to-page using something called a "POST method."   Using the back button reverses things and browsers like to give a “Warning: Page has expired” message (this scares away many users). This is so common that there are forums that devote much time and effort to assisting developers in dealing with, and programming around this scenario.

Also, Web sites make use of nice interstitial pages to read while the “hundreds of calculations” take place.  This creates an awkward situation where the back button appears to start the whole search process over again.

When things like this occur – it means bad news for everyone. The user becomes frustrated. The Web site server is taxed unnecessarily. And most likely no one gets what they want.

The lesson here is to never lose site of the goal - satisfying your customer's need.  Sometimes this means focusing your efforts in places you never expected.  Sometimes you end up having to become an expert on "POST" and "GET" methods in order to design something that will satisfy the majority of users. 

All because of a basic “back."