The important thing here isn't that a company implemented guest checkout. It's that when they did it, because we could see the problem in our research, they found $300m in revenue. Guest checkout won't work for everyone, but doing research like this likely will.
I doubt it was that simple. The UX flow changes if you can purchase as a guest.
More importantly, I expect when the system was built, the developers asked if all orders had a user, received the answer 'yes', and accordingly built that into the system.
Then one day some consultant comes along and pronounces that 'not all orders have a user now, please implement a guest checkout'.
So the developers probably needed to recode the whole flow, possibly remove some foreign key constraints, allowing a user to only partially exist (which needs to reconcile with a real user later if the email matches an existing user later).
And what if they don't create an account at that time, but they do later, should all orders to a given email then get attached to that account?
Are there privacy issues around this? Can we assume an email address belongs to an individual, I am not sure you can.
All this for a system that is processing millions of dollars in orders. So regression testing and deployment, would have to be done very carefully.
This is all just what I can think of off the top of my head.
So, with respect, you did not just change a fucking button.
It wasn't simple on the back-end. Your estimate of "man-weeks" is off by an order of magnitude, because of the back-end system complexity.
As the article states, it started with the button, which, as you quite rightly point out, dominoed into a lot of changes and thinking about edge cases that didn't exist before. The point I was trying to make was that it started with the button.
The big story here isn't that adding this particular button will yield $300m in revenue. The story was that, by watching users, we saw an opportunity to reap $300m. And we took it and it worked.
The original draft had more detail. As did the backstory piece we wrote. Editors cut it down for page count. (It was originally a foreword to a book on web form design, which was all about the buttons and the fields.)
Yes, good software should abstract the complexity away from the customer. That does not mean the complexity does not exist.
My post was motivated by the fact that the article is written as if this was a trivial change - "The designers fixed the problem simply" - and it really was not (for the developers).
"We took a simple idea and made $300million more dollars" - Yes.
"We did something simple and made $300million more dollars" - Not even nearly.
I mean, sure. But this is, like, a day of work + however much testing you want to do to make sure it works right. It clearly cost a lot less than 300 million dollars.
I'm sure it took much longer, but the "core" work here (in other words, what you'd need to do if this were just your project and you weren't too concerned about uptime) probably didn't take weeks.
Read "Don't Make Me Think" a few years ago. It really put into words what I hated about certain designs, whether it was physical or on the web. All the extra steps, things that doesn't make sense, things that are convenient for the business but not for the user.
Wish more places invested in design. I had to help my dad with Massachusetts new design for their tax homepage. Just awful. You go there to pay taxes monthly, to come close to accessing that page is 3-4 buttons deep.
I recently discovered I could increase my real sales conversion rate by about 30%(!) by switching from a multiple page checkout process to a single-page one. About all I could figure was that each additional "action" in terms of page changing gave the user more time to have doubt about making the purchase. Edit: just it case it wasn't clear, I actually got the increase, it wasn't just a hypothetical possibility.
I don't see what's wrong with IKEA's layout. I mean, it's bad if you came there to buy a particular something (though in that case, you can just skip the gallery section entirely and go straight to the well-organized shop and warehouse sections.) But it's exactly right for what I go there for: looking through the gallery as a 1:1-scale-model catalog, deciding what you want/need from that browsing, and then picking it up if you still want it by the time you see it again in the shop section. It's destination shopping, in the same way a bookstore is.
I hate Ikea for this reason. The last time I went, I was looking for a bathroom vanity/sink. By the time I found them, I was so annoyed at being forced to navigate the serpentine displays that I walked out. Yes, there are cut-throughs, but they aren't labelled. Time is too valuable to be wasted walking around furniture I don't want or need.
If you went there alone, then you're not really the target customer.
Most customers come to Ikea with family or spouse. The store's design provides a space of reflection on the state of their own home against an unfolding sequence of idealized Scandinavian nuclear families. At the end of the pilgrimage, ample opportunity is provided to make amends with cheap impulse purchases.
To me it's irritating when I know what I want and don't want to browse all day. I may want to see the qualities of an item upfront but I don't want to go thru the rest of the maze to get to the warehouse to pick out my items.
But it seems to work for most of their shoppers due to a combination of price and utility.
They have discreet shortcuts for if you do know where you're going. It's very cleverly designed so that the "natural" path through the store takes you on the grand tour, but you can get straight to specific places if you need to.
In my IKEA (in Edinburgh), there's a shortcut from the front entrance meaning that to get to the warehouse, you only need to go through Cooking & Eating, Home Decoration and Home Organisation to get to the warehouse, skipping almost the entirety of the store. You may want to check your local store's map on the IKEA website.
And this is why we shop ikea backwards... walk in the warehouse door, pick up a coffee, find the thing in the catalogue, and skip the showroom altogether.
Unless you're doing it because you enjoy the coffee, you can just find the right thing on the website, reserve it, walk in and pick it up. They do support that kind of workflow if that's what you want.
Nothing to a similar subject as design. I recommend "I Will Teach You To Be Rich" and "4 Hour Work Week" a lot. They're not design books, they're all about human psychology and implementation. These traits are what made "Don't Make Me Think" stand out to me. I Will Teach You To Be Rich is a quick read with things you can implement right away. 4 Hour Work Week is longer, it reads as a story, but has detailed information about product testing (skip to these sections if the motivation part doesn't interest you).
I watched Objectified, design movie. It was okay, but didn't give me much more than Don't Make Me Think. It probably would have made more of an impact if I watched it before reading the book. What these books/movies do is change your perception of the world. Once that's done and you see something similar you'd be like "of course" instead of "I never thought of it that way".
I like to put things into perspective: their revenue back then was around 40B a year, so this represented about a 0.3/40= 0.75% increase in revenue. Not shabby at all, especially because their web business wasn't anywhere close to 100% of their business.
When I see a enter you email field, the first thing I think is "God, do I really want to receive more spam?".
When it is a enter your password field "God, I need to store somewhere this password so I don't forget it, and I can't give them a common password as these guys will almost certainly leak it".
And when it is enter my credit card details, I find myself almost never buying from a website I do not know if I can't checkout with paypal.
Between gmail and sanebox, I don't think about spam.
With 1Password, I don't think about passwords.
With Amex, I don't worry about fraud.
I think more about time. Often I want to find what I'm looking for, pay, and be done. Like running in and out of a shop. Like the article says, I'm not here to start a relationship. I've walked out of B&M stores when they do the survey thing at the register (what's your ZIP? What's your email? Sign up for a credit card?)
While this may be true seven years ago, today (2016), if you look major retailers like Amazon, Walmart and overstock.com, they all ask for login before proceeding with payment. Best buy has a both guest checkout and login page before payment. Something has changed over the years and is not yet captured here.
I've actually backed out of making purchases on walmart.com because I didn't want to have to create an account. I always use the guest checkout if available, and only create an account if I really want a product on a specific website, or if I've purchased from the same company a few times already.
Amazon has many other services (including Prime) that require an account, so it makes more sense for them to require accounts than it does for any other online retailer.
I do the same for smaller sites on my second or third time buying from them. But the first time, if it's something I really want, I generally use a temporary email account anyway -- again defeating the purpose.
The thing is most people aren't Amazon or Walmart (I don't know overstock.com). People do have a relationship with Amazon, so their login makes more sense, simply because of the amount of repeat customers.
Most sites aren't Amazon though, they only get the customer for this one particular purchase, because they happen to be the cheapest, or the one with the item in stock.
If anything has chanced it's that your checkout flow needs to be even simpler, so it fits better on phones.
A few years ago (but still a few years after Jared Spool's article) I noticed that zappos.com required registration for purchase. I was super-curious about this because at the time, Amazon had recently bought Zappos and at the time, Amazon was not requiring registration.
I asked Zappos about it, and eventually got to someone who could talk to me about why--the answer was that they get so many returns that the overall customer experience is better if they know all your details up front. (I'm not sure why that's true but that's what they said).
They didn't give details (and I'm sure they wouldn't), but when I think about how I use Amazon now, I return a lot of stuff.
I personally dislike forced login/register but I've seen it work for some businesses.
In particular, the strategy I've seen work was to put registration early (but not too early) in order to get a contact email address and to follow up with people that drop out of the sales flow. Often people just need a reminder.
I enabled "email on incomplete checkout" that in a shop I own, and the response was very mixed. Some people were seriously offended, that we sent out a very friendly worded reminder email, that they left something in their basket (and had gone part of the way through checkout).
If they haven't actually A/B tested this, then perhaps nothing has changed and they could also realize an increase in sales if they did the same thing. Because they are big they probably wouldn't get as big of a boost, but again, without actually testing, it's only guessing.
I was looking for a specific product today that could only be shipped quickly by a few sites (no where I had an account). They all asked me to create an account. I ended up not buying anything.
The company I worked at a few years ago cargo-culted this advice. They believed that the real problem to conversions was when the customer was told to register, and constantly A/B tested moving registration to different points in the process, to little effect.
Great little piece. Stuff like this really makes me want to scream. How could these site designers possibly not understand how annoying login screens are? Do they not use the internet themselves? Forced login is the #1 most annoying thing you can encounter at a website, and I know that it seems as if I have all the benefits of hindsight, but I feel very confident in saying that I could have explained this to somebody in 2005. Simply bizarre that this is missed so consistently.
In my experience, it's not site designers making this decision. Especially on larger sites, many business rules come from other places that are more interested in forcing numbers to increase and don't think the whole thing through like a designer might.
And, let's not forget that UX as a thing is much more accepted as an important thing today than when this article was written. It's still a challenge for a lot of large companies, but there are many that are finally starting to catch on.
Facebook isn't a retail site, it makes sense that you would need to login, because it's what's customize the site to you.
I think the complaint about login are on sites where it can't make any possible difference to what you came to do. Reading an articles or buying sausages should require a login, because it can't possibly contribute positive to the experience to add the extra step.
Honestly I believe that's why pay-walls doesn't work. It's not the at least some people won't pay, it's just that the processes of having to login to access something you paid for is reducing the appeal. I have a subscription for the Economist, I never use the online content, because I can't be bothered to login. I can't remember my username, or my password. It just because a hurdle and in the end I'd rather not deal with it.
This article gave me flashbacks: must make account with postal service to buy stamps.. Not to mention some strange government rules for passwords (10 characters....)
When Jared Spool mentioned this $300M button in another of his classic talks, he emphasized repeatedly how the metrics that determined the abysmal shopping cart completion rate were completely missing. Even though the metrics for how many people are exiting the flow via password reset and never coming back were relatively easy to get from the retailer's db, they had a flawed funnel model that they began with. They were only measuring between 'adding an item to the cart' and checking out. They had no metrics for 'request new password (reset)'.
This may all be dated now with very sophisticated client-side metrics but at the time it was relevant to the mega-retailer.
This is a great example of why design changes should always be justified by measurement rather than simply by artistic opinion. I don't care if you think the new design is slick and modern and responsive and a work of artistic genius. Can you prove that it has a significant effect on whatever is our important business metric?
You may not always have the ability to properly make those measurements... it takes me a week before I can gather enough data to even weakly show the effects of a change, and I run a top 100,000 site.
Also it is much more difficult to measure long term impact. For example, I made a change a couple years ago that quadrupled conversion rates, but irreversibly dampened growth. By the time it was evident that it was -that- change, it was too late to recover.
This stuff is easy to talk about in theory, but real life ain't ceteris paribus.
Required sign ups during checkout is almost certainly a business idea. You'll never find a working designer that wants to include flow-breaking elements like a required registration.
I'm a bit amazed retailers aren't using federated logins whenever/wherever. Just offer a slew of them. There are libraries in most frameworks for most OAuth and OAuth2 providers.
Most of our clients aren't interested. They are as oblivious to what OAuth is and what it does as my mom is. The platform we use only got OAuth2 support built in about two years ago.
More like the $300M blunder. Eliminating all hurdles to completing a payment should be a no brainier. Ask to create an account after a purchase not before.
I feel this article is fake. Catchy title and content to get more contractors. Although the UI/UX logic is worth noting.
May be this is a different way of teaching UI/UX lesson.
So much for so-called recent "Flat" design movement. Nothing wrong with the previous UI design. Sometime people just get visual fatigue and decide to switch to another look and feel.
I wrote a post in 2011, talking about the back story a bit: https://www.uie.com/brainsparks/2011/10/17/the-back-story-fo...
And this is a more recent presentation on using metrics in design. I talk about the $300m button starting at 44m:45s - https://www.uie.com/jared-live/#design-opposed
The important thing here isn't that a company implemented guest checkout. It's that when they did it, because we could see the problem in our research, they found $300m in revenue. Guest checkout won't work for everyone, but doing research like this likely will.
- Jared