ARCast.TV - Dr. Neil Roodyn on Software First Impressions

The Discussion

  • User profile image

    The download link does not work.

  • User profile image
    Sorry about that - I fixed it.
  • User profile image
    Ron Jacobs: So we're on tonight here on Sydney Harbor, just having a nice little ride on a water taxi. It's a beautiful day. And Sydney is a city that's built around the water. Just a gorgeous place to be, like a lot of great cities. We're later on going to be climbing up on the bridge behind me here. Should be an exciting day.
    Announcer: It's Tuesday, April 10th, 2007, and you're watching ARCast TV.
    Ron: That's right! Welcome back, friends. This is your host Ron Jacobs. I once heard a wise saying that went like this: "You never get a second chance to make a first impression." It's funny when you think about all the things we go through to make a good first impression, make ourselves look nice, wear nice clothes.

    I'm not against any of those things, by the way. I'm all for them. You should brush your teeth. But what I'm thinking about is, what is the first impression of your software? Recently when I was down in Australia I met with Dr. Neil Roodyn, who is just an amazing guy. He's been a guest on our cast many times; I think he is close to the record holder, this was his fourth episode.

    We were at this beautiful site called South Head, where we talked about first impressions. So let's welcome Dr. Neil Roodyn.

    So I'm here at Luna Park in Sydney. And I was thinking about what this face behind me says. Part of the user experience is the experience of the eyes. It's what you see. And although you probably can't read it, the theme of Luna Park is "Just for Fun." So this face is trying to tell us this is a fun place.

    Now, it also tells us something about the style of the park. It looks like a theme park from the 40s. It looks kind of retro. And maybe that's very intentional. I think there's something about the way you use visuals that says something. It gives a message to the eye. And that's part of the user experience.

    Ron: Hey, this is Ron Jacobs and I'm here in Sydney where I'm joined today by Dr. Neil Roodyn. Neil, we're at this place called South Head. Tell me a little bit about this place.
    Dr. Neil Roodyn: So there's two Heads that are at the edge of Sydney Harbor. There's North Head, which is just over there, and there's South Head, which is where we are. And as the two Heads come close together, where they nearly meet, that's the gap to get in and out of the harbor from the rest of the world, from the sea outside.
    Ron: Ah, I see. In fact, one of the early explorers of this area proclaimed that Sydney had the best natural harbor in the world.
    Neil: That's possibly true. [laughs] There are a few good natural harbors around the world, San Francisco Bay is obviously one of the great harbors in the world, but Sydney's harbor is pretty outstanding. As you can see, across here, this is all in the harbor, and see right across to the central business district, where everything happens in Sydney. I live over on the other side, on the North Side, further north of North Head.

    It's a pretty amazing place. This is just an outstanding harbor, and then an outstanding ocean, just an awesome ocean to play in as well.
    Ron: Well, it's kind of interesting when we think about here at South Head. This is like the place where the first impression of Sydney Harbor would be made. When you're coming along from the ocean, this is your first impression. We were talking earlier about software and how important it is to make a good first impression.
    Neil: Yeah. It's interesting, isn't it? I guess the first impression most people have of Sydney and Sydney Harbor when they come nowadays is the opera house and the bridge, which are pretty amazing artifacts.

    But the first impression of software is, I think, something that's often under-considered. It's not really thought as an important aspect of software. And as you said earlier on, the setup program or the installer oftentimes is left until the end. On software teams I work with, the first thing I make them do is build the setup installer.

    I think that first point of contact with a piece of software that you build is critical. As you install software, the first thing that an end user or consumer sees is your installer. And if that is a poor experience, then the rest of their experience is tainted by that. And, in fact, I've seen pieces of software that do pretty much what the user wants, but the installer was so appalling that forevermore they tell you that the software sucks.
    Ron: [laughs] It's funny, because I think if you ask most teams, "What's the first thing your users will see?" they will imagine what happens when you run the program after it's installed, that first screen, or something like that.
    Neil: Yeah.
    Ron: But I don't think many would say the setup program.
    Neil: Right, but it is, especially for consumer application. And even when you're talking about enterprise applications, the install process is what the admin sees. And what you want, if you're a software developer, is the admin to be your best friend, because they're the guy who actually has to do first line support most of the time. You want to be chummy with those guys. You want those guys to help you solve problems if there are problems in your software.

    And if their experience of setting up the software and installing it sucks, then forevermore they'll think the software sucks too. And so they'll hate you. They'll be like your worst enemy; because they'll be like "Do you know what was involved in setting up that software? I don't even want to deal with it anymore." Especially when you ship version two or the patch, right?
    Ron: Well, I was thinking how there are a lot of things in user experience that have an analogy in the real world. If I were to buy a large appliance for my home, and the manufacturer or the store where I bought it said, "Well, we'll send out a crew to install it for you." If these guys showed up shabbily dressed with dirty shoes, making a mess on my carpet and all that, I might have a very negative feeling about them.
    Neil: Absolutely. You might think, actually, that negative feeling goes on into the product. You think, "Oh, that damn product that I got. You know how much paint it was to get the carpets cleaned afterwards, just for that one product?"

    Actually, another really good example of that first user experience is the experience, do you have an iPod?
    Ron: Something like it.
    Neil: Something like an iPod. I'm thinking specifically of the Apple packaging, before you even get to the device. I actually think the packaging is better than the device.
    Ron: [laughs]
    Neil: Because when you get the package that the iPod comes in, it's so slick, it's so shmick, it's so well-designed that you just want the thing inside it. You take it out and the packaging opens up like a flower and there's the iPod in store for you, ready to take out and start using. That experience is a beautiful experience. I think that experience is actually better than the experience of using the device. But that's another story.

    What Apple has done has really worked out. The installer experience begins before you even get to touch the device. It begins as you open the box. And if you're shipping consumer products and you're actually shipping real CDs in boxes, you want to think about that. It's not just the installer, but it's actually how you present that to the customer.
    Ron: So you've done a lot of design work in your career?
    Neil: Well, design in terms of software design, for sure. But in terms of artistic design and creative design I've made a point of working very closely with design houses and design companies, because I realize the importance of that first experience, the importance of the usability experience throughout the whole of your software, the importance of discoverability. It's something that's underestimated often.

    There's actually a couple of companies, but one in the UK that I've work with extensively over the last ten years, they do a lot of the design for software that I work with. I often bring them into companies that I'm working with to help with getting that first user experience and getting the software so that it's very discoverable, the features that the user needs to find are very easy to find. It actually becomes a rich experience and enjoyable experience for the user.

    One of the things these guys say is that when you build a piece of software, when you build a product, it needs to become a place that the user wants to go back to. Unfortunately, in this industry we often build places the user never wants to go to again.

    Ron: But they have to.
    Neil: But they have to and they feel forced to. If we can build this lush little place that the user goes, "Oh, I really want to go back there because that was such a great experience." There definitely are some applications out there that are a bit like that. You probably know those applications in your area which you keep going back to because you think, "It's a really nice experience, I just want to open it up because the way it opens up is super-cool."

    Ron: Let me close that box again, all right. [laughter] OK, if we are talking to people who are learning to become architects, what's your advice to them about learning about this user experience of setup, of installation, of opening a box, so to speak?
    Neil: I think the first thing is, make sure that you consider it early on in the process. In fact, I would consider it first. It is the first thing the user sees, it is the first thing you install, the first point of contact for your software. Put it on top of the list. That is what you need to think about, what is that experience going to be?

    Secondly, most software architects and people who want to get into software architecture are not designers, they're not creative. They're engineer typically background. So, align yourself with someone that does have that creative capability. Find some people that do have that creative capability and work with them so that you can make sure that that experience through the whole life-cycle of your product is a beautiful experience, a place the user wants to come back to.

    And so, understand where your own weakness is and align yourself with other people that can fill in those weaknesses.
    Ron: Most engineers that I know avoid the setup process.
    Neil: Right.
    Ron: They don't want to be connected to it; they want to leave it till the very end. If they could give it away or outsource it, they would if they could.
    Neil: Right, partially because it's hard work. It's hard to get it right and it's hard to make it feel like a shmick experience. It's hard to make it feel real super-slick and work really nicely. And you got to keep interrupting the user with all these questions like, "Do you want it on the C drive or the D drive", right?


    Real boring stuff that the user actually doesn't care about, and so you've got to think about how's this going to work, what's the best way of getting this package. This is, again, very much about smart client and rich client stuff, but how do you get it onto the user's machine so they can use it?
    Ron: Well, in fact I was thinking about, if you were to sit down at the beginning of a project and say, "What are the risks to the success of this project", you might start listing things like, "Maybe it won't perform well", "Maybe it will have some kind of security problem". Few people would say, "One of the major risks to our project is that people will not be able to successfully install it".
    Neil: Right, or use it.


    Your project is a total failure if people can't successfully install it, because no one can ever use it.
    Ron: That's true.
    Neil: So, it is actually a real high risk. You have to actually think about that early on, and what you need to put that on your high-priority list of things to accomplish is to make it real easy to install. Lower the barrier to entry to your software. The barrier to entry to your product is the installer. If they get an installer that they can't get all the way through, or they get through but install wrong and the package doesn't function correctly, then you're in trouble.
    Ron: I was noticing the Office 2000 setup for the beta release I installed.
    Neil: The 2007 stuff?
    Ron: Yeah. It actually did a very nice job, I thought, of simplifying it. It just shows up a nice big dialog box with a big button that says "Upgrade" and then a small little button down there that says "Customize" in case you want to choose different options about it. You click the one button, that's all it ever asked it. It didn't do anything else. It just went away and installed it, and came back and it was all done. I thought that was brilliant.
    Neil: That's a good example, I guess. I haven't gone through that process; I've just used images to install it.


    I've just taken disk images of Vista with 2007 already installed and dumped them onto a hard drive. But it's a good example. That's what you want. You want a real low barrier to entry, just that one button "Please install it", bang, and then it goes away and does whatever is needed to do. There's nothing worse than a piece of software that comes in a ZIP file with a README that says, "Please unzip this to C drive, C colon slash special temp drive. Copy these exes into this folder, register these 15 DLLs, copy this DLL into the Windows System folder, run this exe, close it, start it again. Now you're ready to go."


    "Oh, and if you want shortcuts on the Desktop, do this."

    Ron: Well, you've pretty much guaranteed that only people who are engineers will be able to use your product.
    Neil: Right, and even most of them will get pissed off with it pretty quick.

    Ron: That's true. So, that makes a lot of sense for desktop applications, but I think one question we ought to ask about the install right from the beginning, is there any need to install anything at all, or can we experience where there is no installation?
    Neil: You can. There's the concept of a web application or a web-based application. You take, for example, what Flash does. Flash is running an application, there's something on your machine, but there's something online that is streaming as well with some interaction model. That's quite a nice model and a lot of people use that. Of course, it does require on having that always connected scenario for it to work and for it to function for your users.

    But even if you go to a complete online application, there is a first user experience. There is that first time that you register for the site or your register for the service. There's the first time you hit the home page. All of those initial experiences are part of the install process for that type of software. I'm trying to think of an example, but say you just create a Livemail account.

    The first time you go to Livemail to create a new account, you have to go through the registration process. That's the install process, essentially, that you have to go through and so that needs to be thought about: how do you lower the barrier to entry to make that as smooth and enjoyable as possible so that when you've finished going through the process you'll remember what your password is. Maybe it gets emailed to you in some way or the dialog actually says "Please write this down somewhere so you'll remember it" or "Make sure you remember this password". Then it leads you into actually using the application. So, it is something that needs to be considered.
    Ron: It's kind of funny when I think about that. The hotel I'm staying at here in Sydney, when I start the Internet browser in the hotel room, of course they want you to register to pay for your Internet use.
    Neil: Of course.
    Ron: I go to a page where it prompts me to select a plan, whether I want to purchase one hour or 24 hours of access. Then, there was just this text field that said "Time Units Purchased". I couldn't make out what they meant by that. I already said if I wanted 24 hours. Did I put 24 here, or what did I put?
    Neil: So you bought 24 days of 24 hours, did you?

    Ron: No, fortunately. First I left it blank. Then it came back and it kept saying I had to fill this in. So, I took a guess and put 1, and then it worked out on the next page that I was purchasing one unit of 24 hours.
    Neil: Right.
    Ron: It was a very uncomfortable experience, because, yeah, you could easily make that mistake and put 24 in there and purchase 24 days.
    Neil: Actually, in that example you didn't have an option to go somewhere else. But imagine you had had two choices of which Internet provider to use, and one make it slick and one was asking you a question you didn't know what the answer was. Which one are you going to go for, right?

    That's what happens in the real world. There's a ton of options for most users to use different online email accounts, different online services, different online retailers. Some of them make it easy and some of them make it harder. So, you want to make it as easy as possible so that the users come to you.
    Ron: Sometimes these kind of sites, when you register, they'll ask all sorts of information that you don't want to give up. What's really annoying is when they make it required that you tell them all this information.
    Neil: "How many days a week do you shave?"

    Ron: Yeah, these kind of things, or how many children you have, how much money you make, where you live, things like this. They're obviously wanting to sell this information, but I don't appreciate being required to give it. So, I'll oftentimes not register because I don't want to give that up.
    Neil: Right, and so that company has lot another user. To think how many times they lose you because people just don't want to go through that painful install process. And yet, if they were smarter, and I've seen companies do this, they can get you into the site, they can let you register, and then as you start using the product they can start gathering information about you.

    They can gather information either by asking questions, or just by watching your behavior on the site. If it's an online retail site, they can pretty quickly tell whether you've got money or not got money. They can pretty quickly tell where you live because you're going to be sending products there, right? So, they'll be gathering that information anyway.
    Ron: I think sometimes also there's a concern about what they're going to do with this information. About once a year I establish a new personal email address because the spammers have caught up with my old one and they've gotten increasingly clever at defeating the spam filters. So, you wonder how people are getting this address, and obviously some sites I'm registering with are selling it.
    Neil: Yes. That's a reasonably well-known business. People are selling details. If you didn't go to those sites, Ron, and we won't mention what sites are, then you wouldn't have this problem.
    Ron: [laughter] What they should do, if you're a reputable site, though, you should be very clear about your privacy policy so that your users don't think that you're out there selling their information.
    Neil: Right. It is questionable as to why you even need that information as part of the install process to get going using the site. You want to ask for it when you need it and I think that a lot of those sites ask for it way too early on. They're so super-keen to gather all your information and put it all in a database so they've got more information and more user details. Yet, they actually don't need it just for you to get started and start browsing the product, start looking at what's on the site.

    Then, when you buy something, of course you're going to need to put your credit card information in, and when you want to send something to your mum, you need to put in the address to send it to her, and all those kind of things. I think there's big mistake in trying to gather too much information too early.
    Ron: So when it comes to designing these experiences, how do you design an experience that's appropriate for the kind of user that you're going to be targeting?
    Neil: Well, I guess the first question is, what is the user? And a lot of people, even when they do think about it, make the wrong assumptions or they think about it so hard that they think, "Oh, we just need to be general case and we need to make it usable for everyone."

    That doesn't work, and in fact one of the interesting things is that the expected user is not always the real user. A good example of this is, there was a website, I'm trying to remember what it was, I think it was For CEO, and it was a CEO-focused site. They thought the kind of people who are going to register here are CEOs, they have this much information they're prepared to give us; this is the kind of stuff we want to collect off them, etc.

    After a couple of weeks of having launched the site, they realized that actually no CEOs came to their. The people who went to the site are people who wanted to be CEOs. It was a different type of market and a different group of users that they were actually targeting. And so, yes, the site became reasonably successful, but based on the fact that they could move reasonably quickly and start focusing on wanting to be CEOs, not actual CEOs.

    I think to a certain degree, especially when you're online, there's the ability to shift real fast, and starting learning who your real users are as you launch, and go in to launched without being too "Oh, we know our user is 14 to 18 years old." You may well find that actually college kids love your site and your user base is 18 to 25, in which case you can change some of statistics you're collecting. You may change the questions that you ask people.

    You may also have to change the orientation of your site. It may have previously been a very child-friendly site and now it's more teenage/adult friendly, or young adult-friendly, and you need to change that focus. Especially online you need to be prepared for the people you expect using it not being the people using it.

    There's actually another big business problem with that, and that is that I've seen businesses who previously have said, "No, they're not the right users. I know they're signing up, but we don't really want them." Well, I'm sorry, you just got 200, 000 users. "Yeah, but they're not the right ones, they're not the ones we're targeting." Yeah, but they're prepared to pay you money, go with it. Jump on board that, and get onto that business.

    You've now created a business. It's not the business you thought you were going to create, but follow that business and follow that path, because that's the direction you're now going. I think too many business are so focused on following their business plan that they write on day one. It's the same with architecture, they build this architecture on day one and they try to follow it all the way through. You need to be flexible, because the world changes and the way your product gets used will change.

    [music starts]
    Ron: Well, first impressions count. So, thanks so much Dr. Neal for sharing with me today.
    Neil: You're welcome. Thanks, Ron.
    Ron: Yes, Dr. Neil Roodyn, and once again, ladies and gentlemen, we're learning about how important user experiences, in so many ways in that first impression from the point when you run setup.exe or you go to register with the site the very first time and they ask you all kinds of ridiculous questions.

    I'm telling you it's really annoying. I went to one site one day when they were obviously trying to make money from a lot of partners, and they made you go through page after page and uncheck these boxes. No, I don't want that newsletter. No, I don't want this special offer, no, no, no. I hate that stuff. But they don't know the kind of impression they're making. Am I going back to that site? No way. I don't know about you, but that's the way I feel. Thanks for listening.

Add Your 2 Cents