Conversion tip: Don’t let bad error messages cost you sales

Writing error messages is not sexy. In fact, it’s incredibly tedious and boring. But don’t confuse tedious and boring with unimportant. Often, the quality of an error message can be the difference between a sale and an abandonment. And a poorly written error message is a needless and shameful way to lose a sale. The good news is that improving error messages has a high ROI as the cost of the investment is very low.

It’s important to remember that our sites are really self service software applications, and they’re very likely not as intuitive as we think they are. Referencing back to one of my previous posts, “Is elitism the source of poor usability,” we have to remember that our customers probably aren’t as tech savvy as we are, and they are definitely not anywhere near as familiar with our sites as we are. So, it’s important that we’re very clear in our messaging when something goes wrong.

So what does it take to write a quality error message?

1. Be specific
   It’s so important that we tell our customers exactly what went wrong. Our developers have to write code for every possible instance, but all too often we resort to generic and vague language in our error messages. Here are a couple of examples:

   

   As a customer, I’m not sure what I’m not sure what happened or what I should do about it. I might try once again, but if I got this message a second time I would be gone.

   

   This either/or scenario is really an example of a lazy error message. Which is it? Is the address improperly formatted or does it contain invalid characters? We need to tell customers specifically what is wrong and tell them how to fix it.

   Here’s a much better example:
   

2. Use clear language
   It’s very important to avoid anything that even remotely resembles tech jargon. Try instead to use short words that are part of everyday language.First, a bad example:
   

   Huh? Customers understand “password” but “authentication credentials” are certainly unclear and sound kind of scary, frankly.This one is much better:

   

   This is both specific and written in clear and simple language.

3. Strong visibility
   Error messages need to be extremely prominent. Use color and other symbols, such as exclamation points, to help the error message stand out. It’s also helpful to separate the error messages from the rest of the page with white space. Include the message prominently at the top of the page and also at the specific field, if it’s a form error.Here’s a good top of page error:
   

   I would like to see more white space around the error message, but otherwise this is really good.And I really like this way to highlight a particular field where the error has occurred. It may not be pretty, but then it probably shouldn’t be. It should stand out, and this does. Even better, we get a very specific message telling us exactly what’s wrong with the field.
   
   

4. Be polite
   Whenever an error occurs during our customer’s experience with our site, we’re in danger of losing her if we don’t handle it well. So, let’s be as courteous as possible. The cost of courtesy is zero, and it allows us to come across as friendly as possible.Here’s one that is both specific and polite:
   
   

   Here’s one that goes the extra mile to suggest calling Customer Service if there is still a problem. This is a very nice touch that will go a long way towards saving the sale.
   
   

5. Provide examples for how the information should be entered correctly
   It’s very important they we’re not only specific in defining the problem that occurred but also specific in explaining how to correct the problem. If the customer has entered his email incorrectly, we cannot assume that he knows what he did wrong or how to enter it correctly.Here’s an error message that explains the format pretty well:
   

   However, the customer may not understand what “domain” means. It may be be better to also use a real example with a well-known domain like “name@aol.com.” Even better, incorporate the information the customer entered, if possible.For example, the error might say something like:

   You entered “kevin” for your email address, which is not a complete address. Please enter an “@” symbol followed by an email provider after your email name. For example, “kevin@yahoo.com.”

Even better, be proactive. Stop the error before it occurs.

I really love how Restaurant.com handles their form fields. Upon entry to a form field, a dialogue box dynamically appears next to the field with some helpful information. The movement that occurs upon entry really draws your attention to the helpful information, which I find considerably more effective than help text persistently present under or next to a field. It’s far easier to ignore static text than something that appears when you enter the field.

Additionally, the folks at Restaurant.com have included some great help text that provides important information. In this example, they’re letting us know the address must match the billing address on our credit card. Excellent!

And here, we get some specific information about the value of our password and the basic requirements for the password. And we get some nice politeness to close it out.

Save those sales. Give error messaging your full attention.

Error messages should get just as much attention as any other site functionality in the requirements processes for our sites. We should give error messaging as much attention as we give to marketing copy. It may not be sexy, but it’s critically important if we want to avoid needlessly losing sales.

What do you think? How much time to you put into error messaging? Do you have examples of particularly good error messaging? Would you add anything to the list of quality error message attributes?

The Case for an E-Commerce IT Org Change

As multi-channel retailers move more and more towards implementing cross-channel strategies, organizational structures need to change to support those new strategies. I am a huge proponent of breaking up most e-commerce silo  organizations and integrating online and in-store marketing and merchandising teams to ensure a common vision and voice across channels. For IT, though, I actually recommend the opposite approach. I believe technology  professionals who work full time (or near full-time) on the e-commerce site should report directly to the head of e-commerce.

While at a high level is seems like technology should have the same kind of continuity as marketing and merchandising, I believe a close look tells a different story.

Here’s why e-commerce IT is significantly different from traditional IT:

In e-commerce, the business is technology

Traditionally, IT creates tools that help employees be more productive and efficient. However, in e-commerce, IT is actually creating software designed to generate revenue. E-commerce “stores” are really self-service software  applications designed to help customers perform a service — in this case it’s to buy the products and services we sell.  Intuit has Quicken; Microsoft has Word and Excel; retailers have our e-commerce sites. We really need to think about  our sites more as software products and organize our teams in a product management type of structure.

Also, a particular pet peeve of mine is when IT folks refer to those in other functions of the company as “the business.”  Just that reference alone insinuates that IT is not a crucial part of the overall business and creates a separation that  frequently leaves IT coming across as second class citizens, which they are not. While I’ve never liked “the business”  reference in any circumstance, it’s doubly bad in e-commerce where success absolutely depends on technology team  members actively working as part of the business.

Self-service applications require a different mindset

Working on an e-commerce application that is designed to be used directly by customers requires a very different  mindset than what is typically required when working on applications that support employees. Even when the  underlying technology is similar, the mindset required is substantially different. New employee applications usually  come with training programs and manuals. Moreover, employees are ultimately forced to use the app; they get used to it and get incrementally better at using it over time through daily usage. Customers, however, don’t get the benefit of manuals and training programs. They’re on their own. And if the experience doesn’t satisfy them, they give up and the sale is lost.

Site functionality and customer experience are major components of the e-commerce business strategy

The website application is a key differentiator for the business, and customer experience is hugely driven by site functionality. While functions other than technology certainly contribute to an e-commerce site’s success or failure, there can be little doubt that the quality of the technology is a massive contributing factor.

E-commerce is 100% dependent on technology to be open for business.

While technology is critical in all areas of the business, most retailers have offline contingencies for stores so they can  continue to make sales even if the system is down. Websites obviously don’t have an offline mode.

Web businesses are still immature and need considerable agility and flexibility to mature as quickly as possible

For many absolutely legitimate reasons, most IT organizations at multi-channel retailers have significant (and some might say onerous) processes in place to manage technology requests and roadmap prioritization. Because requests for technology improvements come from all corners of the company, it’s important for CIOs to ensure they are spending their resources on work that is thoroughly vetted and likely to generate the highest return on investment for the company. But given the absolute dependence of the e-commerce business on technology, typical IT prioritization and allocation processes are too slow for e-commerce businesses that need to be able to adjust quickly to issues that arise with customer experience.

The e-commerce competitive marketplace innovates far quicker than the brick & mortar marketplace

The CEO of a pure-play e-commerce company is in basically the same role as the head of e-commerce at a multi-channel retailer. If for no other reason than there is no alternative, the CIO of a pure-play reports to the CEO. This reporting structure gives the pure-play leader a leg up in agility and the ability to react to customer needs. In a multi-channel retailer, the CIO must split time between many functions of the business, and I find e-commerce often gets time allocated in a ratio roughly equal to its financial contribution to the business. While such an allocation is understandable given everything on a busy CIO’s plate, I believe this lessened focus can lead to stunted growth and lost ground to competitors such as Amazon who are more devoted to improving their software application and increasing their customers’ satisfaction with their site customer experience.

I believe if a head of e-commerce is to be truly held accountable for the success of the site, he or she should have  appropriate authority over such a major contributor to the success of the site.

So why should the head of e-commerce have authority over e-commerce IT and not e-commerce marketing and merchandising?

To me, it’s all about what faces the customer and what doesn’t. A brand should be clear to its customers about who it is and what it stands for, so continuity in marketing and merchandising trumps silo control over those aspects of the business. Site functionality has no parallels in other parts of the organization. Since it is both unique and customer facing, I believe the head of the online channel should maintain the authority to develop and execute the technical strategy for his or her business unit when it directly affects the customer relationship.

I’ll add this final point: I’ve lived through many different org structures surrounding e-commerce IT, and the only times I’ve found the pros to outweigh the cons of an org structure have been when e-commerce IT was part of the e-commerce operation and reporting to the head of e-commerce.

What do you think? Am I completely misguided? What structures have you seen work and not work? What structure do you think is ideal?

The Tree Stump Theory

Since I mentioned it in my eTail presentation last week, I’ve received a number of requests to expound on my Tree Stump Theory in this space. So, here goes:

As truly amazing as the human brain is, it’s not able to re-process everything we see anew every time we see it. So, our brains take some shortcuts by basically ignoring things we are very familiar with, and that can cause us trouble any time we have interactions with people who don’t have the same level of familiarity with something as we do. I usually talk about this in reference to website usability, but it actually applies to many areas of our lives. To illustrate the concept, I have my Tree Stump Theory…

Imagine if someone brought a big tree stump into one of your conference rooms. The first time you saw it, you would say something like “Hey, what’s with the tree stump?” Someone would give you a compelling reason why it was there, and you would go on with the meeting. The next time you entered the conference room, you would notice the tree stump but not ask about it. After while, someone might throw a tablecloth on it or dress it up in some manner, but it would still be there. You would no longer ask about it or think about it. Frankly, you wouldn’t even really see it. You’d just arrange yourselves at the table in a way that worked around the tree stump and go on with your meeting. Meanwhile, anyone new coming into the room can’t help but see the tree stump and find it to be an obstacle.

We all have these types of “tree stumps” on our sites and in our lives. I bet you could think of something like this in your house right now. They manifest themselves as obstacles to good web usability, but they’re also our biases, our stereotypes and any other set of assumptions we rely on, usually unconsciously, to drive our daily actions and decisions. Sometimes they’re relatively harmless, but more often than not tree stumps prevent people from buying on our sites, or they are the unspoken roots of disagreements and miscommunications in our daily interactions both at work and at home.

So how do we get rid of our tree stumps?

1. The first step is to recognize the fact that tree stumps are everywhere, even when we can’t see them.

If you’ve read this far, you’ve probably made it to step one.

2. Next, get some help finding them

The very nature of tree stumps makes them difficult to self-identify. If you’re dealing with web usability, try the steps prescribed in this post. If you’re concerned about tree stumps in strategies, policies or general decisions, seek some input from someone who is outside the general team and who has a different background from you and your key decision makers. Ask them to openly question everything.

3. Specifically call out assumptions, preferably in writing

Assumptions are the roots of tree stumps. We make assumptions so often that we don’t always realize we’re making them. Listen for statements or reasons that hint of tree stumps. The most obvious is “That’s the way we’ve always done it.” If you hear that one, sound the sirens. But there are other, less obvious comments like “People want…” or “Based on my experience…” or “In a previous life we…” Don’t get me wrong, some of these statements could be perfectly accurate and valid. But whenever someone is applying past experience to a currently situation, he or she is assuming the two situations are similar enough to warrant the comparison. That’s potentially an assumption fraught with problems because the number of potentially important variables in any situation is massive. Writing down those assumptions and then testing them on the current situation often brings bad assumptions to light.

Also, on the web usability side, remember that while your internal reason for a tree stump may seem extremely valid to everyone in the company, your customers don’t know those reasons and even if they did, they probably don’t care. Common explanations that won’t hold water with customers include:”I’m not in charge of that area;” “It doesn’t matter because people don’t use that anyway;” or the time-honored classic, “That’s due to the limitations of our platform.”

4. Schedule regular reviews of your own assumptions

This one in some ways is a repeat of #3, but the point here is to specifically and methodically question yourself. This is really hard to do, of course, but it has a tremendous amount of value. One technique I’ve used in various situations is to write down my first impressions of important situations so that I can regularly review them in the future after I’ve learned more. I recently talked with Shop.org about this technique in reference to starting a new job. Beyond that technique, it just takes practice and discipline to think about your own biases and assumptions to see if they still apply.

I also find it helpful to constantly look for new ideas. I read lots of business and science books. I don’t always agree with everything I read, but new ideas cause me to question my own ideas. I also enjoying reading thought-provoking blogs, some of which are listed to the right, and I follow interesting people on Twitter. More than anything, though, I love to spend time talking to people who think differently than I do and are willing to share their perspectives. (And I hope you’ll share your comments on this post and others.)

Tree stumps are everywhere. We’ve all got them. And as soon as we remove some, more will crop up. It takes a concerted effort and a solid process to regularly look for and remove the tree stumps in our lives and our businesses. But I’ll argue that those of us who are aware of our tree stumps are on a much faster path to improvement than those who go on ignoring them.

What do you think? What types of tree stumps have you run into? How do you go about removing them?

Is elitism the source of poor usability?

Most sites are still achieving single digit conversion rates even though customer intent-to-purchase rates are 20% or higher in most cases. Customers are continuing to run into obstacles to the purchase process that need to be eliminated. The good news is that during this time of limited capital investments, retailers can use low cost means to find and eliminate as many obstacles to purchase as possible.

The first step is to get into the right mindset and remove what I feel is the biggest disconnect with the customers that many retailers have: we’re way more comfortable and experienced with our own sites than our customers are. We use our sites every day, and we know exactly how they’re supposed to work. However, our customers are generally nowhere near as familiar with our sites as we are.

Two weeks ago, I was lucky to be able to attend GSI‘s Connect conference for its clients. I was even luckier to attend a fantastic session by GSI’s Senior Director of Usability, Michael Summers. Michael got the audience’s attention pretty quickly by calling us all elitists…and he had a good point. He asked us how many of us fit the demographic for today’s main Internet users and quickly made the point that we were higher educated, higher paid and more Internet savvy — by a long shot — than the average site user in the marketplace. If that wasn’t enough, he showed some video of average Americans shopping online who had trouble with some of what we in the industry would consider among the most basic aspects of websites.

To solve this disconnect we need to see our sites through our customers’ eyes. There are a number of ways to do this that I’ve found to be effective.

1. Use statistically significant customer satisfaction surveys to get trendable data that will  point to the biggest problem areas of the site.
   The two key phrases here are “statistically significant” and “trendable.” Per my last post, continuous measurement is important to avoid random outliers and uncover the underlying truth. When done correctly, customer satisfaction surveys can be extremely reliable, accurate, and predictive and can tell you not only which areas of a site customers complain about most, but also which areas of the site will actually have the biggest impact on purchase intent and loyalty. This is critical information to provide some some direction on where to focus your usability efforts.
2. Ask open-ended questions to add color to the quantitative information.
   Quantitative analysis is extremely useful, but numbers alone aren’t nearly enough. Numbers will certainly tell you the problem areas of the site, but to really get your arms around what the numbers are saying requires adding some color to them with some qualitative information. Asking more open-ended questions like “If you could make one improvement to our site, what would it be?” are good starters to bring some of the numbers to life. If the numbers tell you that customers in general are having problems with navigation and you see that multiple customers say in open-ended comments they just want to see all the blue dresses in stock, you might start to consider adding color choice to your navigation. Or maybe you already have an option to navigate by color, but the customers aren’t seeing it and you’ll need to find a way to make it more apparent.
3. Watch your customers use your site.
   The absolute best way to add color to the data is to actually watch customers use the site. In the past, I’ve seen great discoveries come from taking a laptop into a store and asking real customers to shop on the site while I or someone on my team watched silently. In these situations, it’s very important not to be too prescriptive in the tasks the customer is asked to do. Ask them to “find and buy a new pair of dress shoes” rather than “go to the men’s tab, then select dress shoes and find a pair of black, size 9 shoes.” It never fails to amaze me in this situation how many different avenues customers will take to accomplish the task, and they’ll frequently run into trouble. These trouble spots are the areas to find and eliminate. Some of the smallest fixes can often significantly improve conversion and customer satisfaction.If the logistics of getting into a store are too difficult or you don’t have physical stores, there are technology alternatives, like Tealeaf’s CX and ForeSee’s new CS Session Replay, that provide the ability to replay customers’ sessions on your screen.
4. Have an expert conduct a usability audit.
   Even after discovering where customers are having trouble, it’s sometimes still very difficult to determine exactly what you should be doing differently to make the experience easier and more intuitive for your customers. In those cases, expert advice via a third party usability audit is an excellent solution. I’ve used trained usability experts in the past to identify specific improvements that led to tremendous business results. Third party usability auditors bring to the table both fresh and trained eyes that have likely seen problems similar to those on your site before and have come up with solutions for those problems or seen how other sites have solved those problems.

Regardless of the mechanisms you choose to use, the key to better usability, better customer satisfaction and the resulting better conversion and sales, is finding ways to see your site through your customers’ eyes.

Are you a usability elitist? Do you watch customers use your site? What have you learned in the process?