Marketing Expert's Corner
This article written in 2009
There's a concept in software user interface design called the Out of Box Experience. It has nothing to do with out of body experiences, or thinking outside the box. It's all about what the user sees and has to do when they first open your product's box: unpacking the stuff, finding 4 "read me first" slips of paper, not finding the right cables, trying to install the software with the 37 dll's and patches missing.
You know, all the stuff that Apple gets perfect every time and most other companies can't do right for more than about a month at a time.
So...we don't have any boxes any more. How can OOBE matter? Extend the idea from "your product" to "your company." Substitute the words "initial customer experience" -- or even better, "initial contact experience" (for people who aren't yet customers). Guess what: Apple gets this perfect every time, and too many other companies can't seem to do it right to save their lives. This isn't just a matter of attractively designed bags and boxes...it's less to do with design than with execution and attention to detail.
If you plan on keeping customers, you need to design interactions as carefully as you design products. Whether you're in retail or consumer or IT or professional services, it is almost impossible to be too good at OOBE. And little things you're doing sloppily are turning customers away every day. Due to layoffs & cutbacks, this issue is getting worse. Read up.
Imagine you're trying to contact a company for the first time, either with an innocent request for information or to find out what their warranty returns policy is.
- You call them up on the phone and are greeted with an automatic call director (ACD) or interactive voice response (IVR) unit. You punch through the menu hierarchy, or you talk to the voice recognition unit, and get lost. You hang up and try again, only to find out that the thing you're trying to do is down a blind alley, where you get caught in an infinite loop from which there's no escape to a human operator that can help you.
- You then go to the website and are greeted with an impossible navigation structure. Maybe the thing you're trying to find is there, maybe it isn't. You look for the site map and there isn't one. You use the search feature and it bounces you to a bunch of irrelevancies.
- You go to the "Contact us" form to try to get someone to talk to. No phone number other than Sales. You fill out the "call me" form and it's got eight items to complete. You fill it out, click "submit" (did you really want to "submit" to anything? why does it always ask you to submit?) and the form replies with an error message about one of your entries. You follow the link to fix the entry and -- tada! -- all of the fields are blank. You enter them again...same error message. After the third time you give up.
- You go to the web site's "contact webmaster" form to let them know about the bug in the form. That
form also doesn't work. It also erases all your fields
after giving you the error message.
So now -- are you EVER going to do business with this company? Not very likely if you have the choice.
The Road to Hell is Paved with Good Intentions
From the first moment a customer, prospect, or curious tire-kicker tries to interact with your company, you have the opportunity to impress them...positively or negatively. Unfortunately, it's all too easy to put people off, even though your intention is to give them the best possible customer service. Many an horrendous customer experience has occurred in the name of cost control, expediting, or responsiveness. Despite the good intentions, fragmentary execution means a lot of people fall through the cracks.
To avoid these problems, you've got to think through the entire contact cycle as if you were an uninformed customer. This isn't just a "site design" or "information architecture" exercise. This is the look and feel of your entire company. Here are the steps:
- Design the customer's (or prospect's) first contact. They're not going to be contacting you by accident, so think through who is likely to contact you, what their goal is, what their state of mind is, and what channel they'll be using to connect. Use personas and roles, just like you would with any other user interface. Think about the timing of the customer contact, and what their emotional state will be. This is the clearest example of my mantra, "design your customer before you design your product."
- Develop workflows for the top 5-10 expected customer interactions. These workflows will never be totally automated, but they have to be thought through systematically so that there are clear paths (with signage and options) at every stage so people can't get lost or trapped.
- Workflow #0 is the catch-all exception. Make it easy for them to get to a human being quickly. Use phone, IM, and email. Use Twitter if you have to.
- On the phone, most companies will use some sort of ACD / IVR / VR technology. Make the menu trees as narrow as possible (anything more than 5 choices at a time is ridiculous), and provide an exit-to-operator option as early as possible. Always supplement the voice-recognition system with keypad responses, so the 10% of callers from noisy environments or strong accents don't get frustrated. Yes, this is going to cost more.
- For the customer support reps who will eventually be responding to callers, make sure that their headsets aren't broken, their VoIP delay isn't too long, their accents aren't too strong, their energy level isn't depleted, and their speaking voice is clear and pleasant. The drive towards always-lower call center costs has really taken its toll -- a remarkably large number of phone reps have trouble with verbal communication. Test each of them, monthly.
- Website forms should ask as few questions as possible -- less really is more, if you want responsiveness and un-frustrated customers. For initial registration on your site, it should just be name, email address, password, and country. Use progressive registration to get more profiling information from web visitors every time they ask you for something. Never ask people to fill out long questionnaires (in the last year, I've seen registration forms with over 50 questions on them -- this to download a video!).
- The website needs to be designed for quick navigation -- follow the standards of your industry unless there is a very very specific reason not to. Make sure there's a site map, and the link to the site map is at the footer of every page. Improve the site search (probably with an "advanced search page") so that result sets are small and relevant.
- Use multiple landing pages, each tailored to the roles and goals of people who are visiting the site. Every workflow from step 2, plus every ad campaign you have, should have a separate landing page. Optimize these pages for the 15-second rule: an uninitiated user should never have to spend more than that amount of time on a landing page, and their goal should be achieved with one click.
- Every path through every workflow needs to be tested for potholes and wrong turns. The tests must include email auto-responses, to make sure they're sensible and helpful. The testers must not be the designers, or even members of the design team. The testers need to know enough about your industry to make sensible choices, but not know so much about your company or its jargon that they can avoid traps by guessing how the company thinks about things.
You never get a second chance to make a first impression. So don't skimp on this stuff! The business case for investing in OOBE is sledgehammer-simple:
- How many prospects can you afford to lose?
- How many customers can you piss off without effecting profitability?
The answers here won't be "zero," but will be very small indeed.
Phone Us +1 650 326 2626