Menu Search
Jump to the content X X
SmashingConf London Avatar

We use ad-blockers as well, you know. We gotta keep those servers running though. Did you know that we publish useful books and run friendly conferences — crafted for pros like yourself? E.g. our upcoming SmashingConf London, dedicated to all things web performance.

UX Study The State Of E-Commerce Checkout Design 2012

A year ago we published an article on 11 fundamental guidelines for e-commerce checkout design1 here at Smashing Magazine. The guidelines presented were based on the 63 findings of a larger E-Commerce Checkout Usability research study we conducted in 2011 focusing strictly on the checkout user experience, from “cart” to “completed order”.

This year we’ve taken a look at the state of e-commerce checkouts by documenting and benchmarking the checkout processes of the top 100 grossing e-commerce websites2 based on the findings from the original research study. This has lead to a massive checkout database with 508 checkout steps reviewed, 975 screenshots, and 3,000+ examples of adherences and violations of the checkout usability guidelines.

Here’s a walkthrough of just a handful of the interesting stats we’ve found when benchmarking the top 100 grossing e-commerce websites’ checkout processes:

  1. The average checkout process consist of 5.08 steps.
  2. 24% require account registration.
  3. 81% think their newsletter is a must have (opt-out or worse).
  4. 41% use address validators.
  5. 50% asks for the same information twice.
  6. The average top 100 checkouts violate 33% of the checkout usability guidelines.

In this article I’ll go over each of them and explain exactly what’s behind these numbers, showing you some real life implementations of do’s and don’ts when it comes to checkout processes.

The Average Checkout Process Consists Of 5.08 Steps (But It Doesn’t Influence Usability Too Much) Link

The average checkout consists of 5.08 steps, counting from the shopping cart to the step where the order is actually placed — often a “review and confirm order” step. The shortest checkout process is one step (including cart) and the longest being a massive nine steps.

Average Number Of Checkout Steps

Above you see the distribution among the top 100 grossing e-commerce websites in regards to the number of checkout steps they have. Note that only a single website had one step (including cart), and the “average” for this one website therefore shouldn’t have been given too much weight.

Score As A Function Of Steps

Above, we’ve plotted the websites grouped after the number of checkout steps, moving out from the x-axis, as the groups average checkout usability score moves up the y-axis. As you can see, we’ve found that up until six checkout steps there isn’t a noticeable relation between the number of checkout steps and the quality of the user’s checkout experience. This matches the test subject’s behavior we observed during the checkout usability test back in 2011. What matters the most for checkout experience isn’t the number of steps in a checkout process, but rather what the customer has to do at each step.

With that being said, there does seem to be an upper limit to the number of steps practically achievable in a checkout process before it begins to hurt the checkout experience. The websites with eight or nine steps have accumulated a significantly lower score in checkout usability than the rest of the checkout processes. This is often a result of required account registration (which typically induces more steps and is bad for checkout usability) as well as the fact that websites that end up with over eight checkout steps simply have more chances available to screw up the experience for their customers. At the time of testing, these were the websites with eight or more steps: Sephora (83), Amazon (84), Peapod (85), Sony (86), Safeway (97), ShopNBC (98) and W.W. Grainger (99).

To recap: don’t focus too much on the number of steps in your checkout — instead spend your resources on what the customers have to do at each step, as that is what matters the most for the checkout experience. Three examples of this are the checkout processes of Apple, Walmart and Gap, which are all seven-step checkouts that perform approximately 50% higher than the average top 100 grossing checkouts (not to say that they are perfect, there are still room for further checkout improvements). While in theory it is possible, in practice none of the benchmarked websites with eight or more checkout steps had a checkout process that wasn’t greatly under-delivering in regards to the checkout user experience for a new customer.

81% Think Their Newsletter Is A “Must Have” (And Don’t Value Customer Privacy) Link

81% of the 100 largest e-commerce websites “assume” that their customers want their promotional emails by having a pre-checked newsletter checkbox (or worse) at some point during checkout.

Sehopra Pre-Checks The Newsletter Box10
Large view11.

One reason why customer hate being required to create an account to complete a purchase is because they have a mental model of account = newsletter. This became evident during the user testing, where we heard the same complaint over and over again: people hate creating an account when buying online. When we asked the test subjects why, 40% told us that they “didn’t want any newsletters”.

For years websites, including e-commerce websites, have tricked customers into “accidentally” signing up for newsletters that they didn’t want by visually downplaying a pre-checked “subscribe to newsletter” checkbox. So people have come to expect, that when they sign up for a new account, that they also sign up for a newsletter, or “spam” (as more than half of the test subjects had referred to such newsletters).

This mental model sadly isn’t just a misconception, but evidently something learned the hard way. Pre-checking the newsletter checkout is one thing, but of those 81% of the websites that think their newsletter is a “must have”, 32 of them proceed to do something even worse than pre-checking a checkbox:

Amazon Checkout Step 312
Amazon is just one of the 32% of the top 100 grossing e-commerce websites that automatically signs customers up for their newsletters, without clearly informing the customer (only via the privacy link), and without giving an opting-out option during checkout. Large view13.

These 32% automatically sign up their new customers for their newsletters with no way of opting out during the checkout process, and often burying this fact deep down in their privacy policy. Typically, the only way for customers to “opt-out” on these websites are either by a privacy tab in an account settings section (if they were forced to register for an account) or by an unsubscribe link in the newsletters that the customers will automatically start receiving.

So what the test subjects displayed of account = newsletter is something they learned from shopping at websites (such as these from the top 32%). Only 8% of the top 100 e-commerce websites value their customers inbox and ask them to opt-in if they want to receive newsletters (as does the last 11%, which don’t offer newsletter subscriptions at all during checkout.)

24% Require Account Registration Link

To put it differently: 24% don’t offer the customer a “guest checkout” option when placing an order, but force them to create accounts on their websites.

Sony Electronics Checkout Step 2 Account14
Sony (step 215) is just one of the 24% that require every new customer to register for an account when placing an order. Large view16.

During the checkout usability study, we (as have many others have before us) have identified multiple reasons why potential customers resent being forced to register for an account just for placing a simple order. We’ve already touched upon one of them, the mental model of account = newsletter. But let’s quickly list a handful more of them that we’ve found during the study:

  1. Signing up for an account means more steps and form fields to complete during checkout — essentially taking longer to complete.
  2. Most customers already have a myriad of logins and passwords to remember and don’t want more of them.
  3. When creating an account, customers are more likely to realize that you’re storing their information indefinitely.
  4. Many customers just don’t understand why they need an account to buy a product. As one test subject clearly expressed during testing: “I don’t need to sign up for anything when I’m buying a perfume in a regular [brick and mortar] store.”

Nordstrom's Checkout Process Step 317
Nordstorm (step 318) is one example of the 76% of the top 100 grossing e-commerce websites that offer new customers the much appreciated “guest checkout” option, but offering at the same time an easy optional account registration. Large view19.

When you do it right (as 76% of the e-commerce websites have done) and provide the much appreciated guest checkout option, you still have the possibility of asking for an optional account creation along (or after) the purchase. This can be done simply by creating a short section with a brief description and an optional password field. During the checkout usability study no test subjects were put-off by this approach, and just left the optional field(s) blank if they weren’t interested in creating an account with that particular website. But they generally liked the option on websites where they were interested in becoming repeat customers.

If we look into the type of websites that typically require account registration, there is a slight tendency towards them being the highest grossing websites:

Require Registration Compared To Size

Of the 23 websites that had more than $1 billion in online sales (Internet Retailer 2010 sales estimates), 35% of them required account registration, whereas for the rest of them grossing less than $1 billion (and down to $148 million) it was only 21% that required account registration during checkout.

41% Use Address Validators Link

Of these 41%, 12% (relative) don’t allow their customers to override the validation mechanism in case the address isn’t recognized (though the customer is absolutely sure the address is correct).

Amway is one examples of the 12% (relative) that doesn’t allow the customer to proceed in any way, in the event that the address validator is outright wrong, or the address validation database isn’t updated properly. Large view21.

An address validator can be a smart way to avoid common customer typos that might cause shipping problems, ones that otherwise would have resulted in undelivered or delayed orders. But street names, postal codes, etc. aren’t consistent, nor permanent. So the possibility still exists that it’s the address validation mechanism/database that is erroneous — not the customer’s input. Those subsets of websites that don’t allow the customer to force proceed through a potentially wrong address validator (at the time of testing: Office Depot, ShopNBC, Amway Global, FreshDirect, and CafePress) will leave the customers with no other option but to abandon their purchase as they are technically locked-out from completing the checkout process.

Overstock Adhered
A decent implementation by Overstock (step 322) that informs the customers that their typed address doesn’t match the address validation — and therefore, are likely to be wrong — while still giving the customers an option to force-proceed.

The advisable approach — implemented by the vast majority of the 41% of those websites utilizing address validators — informs the customer that the typed address doesn’t match, yet still allows them to force proceed if they are sure that the address is right.

50% Ask For The Same Information Twice Link

Instead of pre-filling the already typed-in information for the customer, 50% of the e-commerce websites add needless friction to their checkout experience by asking for the same information more than once. This is rarely at the same page (although that does happen) but is most often happening across multiple pages. Sometimes it’s the customer’s name that isn’t pre-filled from the address step to the billing step. Other times it’s the zip code that the customer provided at the cart step (e.g. for a shipping calculator) which isn’t pre-filled at the the shipping address step. Although it is only fair to assume that in most cases users calculate the shipping to a certain zip code, this would also be the zip code that they plan on shipping the order to.

Apple Step5 Crop23
Apple is one of the 50% of e-commerce websites that asks for the same information more than once. At their 524’th checkout step the billing email address isn’t prefilled — even when the customer clicks the “Same as shipping information”-link. Large view25.

Retyping information is a tedious task on a regular computer, but on a mobile device most users will find it outright annoying. Considering that all the benchmarked websites gross $148+ million per year in online sales, it seems rather sloppy that only half of them have dedicated the resources to removing needless checkout friction by ensuring that they don’t ask for the same information more than once (across multiple pages).

Hayneedle Step2 Cropped26
On the path to reducing needless checkout friction, only 10% of the websites helped their customers by pre-filling the state and/or city fields based on the zip code provided. Hayneedle (step 227) was one of them. The result: three less fields for the customer to fill + shipping dates and costs already updated at the page entry. Large view28.

On the same note for reducing needless checkout friction, only 10% of the websites helped their customers to fill-out even less form fields by pre-filling the state and/or city fields based on the zip code that the customer provides.

The Influence Of Revenue And Industry Link

The e-commerce websites grossing above the $1 billion mark scored 44% worse on checkout usability (for a first-time customer) than the e-commerce websites grossing below $1 billion.

When taking a closer looking at the checkout experience of these 23 websites that gross over $1 billion, it’s likely that some of that gap exists because these websites are more focused on forcing as many customers into their account eco-system as possible. Furthermore, the top grossing e-commerce websites also tend to be the ones with the most complex marketplace systems. These marketplace systems often end up inducing a lot of derived complexity into the checkout process, due to shipping and legal constraints, for a deal where the website only acts as the middleman. In comparison, some of the “smaller” websites in the top 24 to 100 grossing range had one simplified goal for their checkouts: to let the customer move as swiftly as possible through the checkout process.

Usability Score Vs. Online Sales Scatterplot29
All the top 100 e-commerce websites plotted with checkout usability score moving up the y-axis and online sales moving out the x-axis (logarithmic scale). Notice that the far majority of checkouts that scored the highest on checkout usability are below the $1 billion sales mark. Large view30.

If we take a look at specific e-commerce industries, the Automotive Parts industry had much better checkout usability than the rest of the industries (scoring 110% higher) whereas the Office Supplies industry scored the lowest (38% lower than average). Food & Drugs followed right behind in providing the worst checkout experience.

It’s interesting to see the that in both the worst and the best scoring industries, all three have a very similar checkout process. In fact, their checkouts are almost identical; have a look at Staples’ checkout31, Office Depot’s checkout32, and OfficeMax’s checkout33. I’m not going to speculate on who “was inspired” by whom, nor does it really matter. But in the Office Supplies industry it’s unfortunate, because as a consequence they all suffer from a very sub-standard checkout experience (38% lower than the average). While it’s clear that some of the top 100 e-commerce websites are using the same system vendor (and thus, end up with similar features and sequences in their checkout flow), the tendency of similar checkouts between competitors weren’t noticed to nearly the same degree in some of the other e-commerce industries (e.g. in Electronics).

The General State Of E-Commerce Checkouts Link

If we have an overall look at the top 100 grossing e-commerce checkout processes, the average checkout violates 21 checkout usability guidelines. This is an indication that checkout improvements are still much needed if the average cart abandonment rate of 65,95%34 is to be lowered (“50% Ask for Same Information Twice” also points in this direction).

This overall lacking of checkout experience — even among the highest grossing e-commerce stores — is hardly rooted in an unwillingness to improve checkout experience, but is most likely due to a combination of factors, such as:

  1. Flows are much more difficult to improve than single pages.
  2. Checkouts often need deep, back-end integration, and thus require more IT capabilities to modify/test upon.
  3. Checkouts haven’t been on the agenda for top management (although, I believe this has changed a lot in recent years).
  4. Checkouts are for most designers much more dull to work on than product pages, home pages or new ad-campaigns.
  5. In a few cases, a poor user experience can still be good for business, at least in the short run (e.g. sneaking people into your newsletter).
  6. No Web convention for a checkout process exists.
  7. “Best practice” for checkout designs are scattered and scarce (only two to three research-based resources exist).
  8. Feedback from those who use the checkout process are only several degrees of separation from those who design and develop it.
  9. Improving most somewhat-optimized/decent checkouts aren’t 1 to 3 “big fixes”, but are most likely to be 10 to 30 smaller checkout changes.

If you want to further examine the checkout processes and flows of the 100 top grossing e-commerce websites for yourself — without filling out some 1,300 form fields, as we have done — you can do so in the free part of the 2012 E-Commerce Checkout Benchmark35, as we’ve decided to make that part of the database publicly available.


Footnotes Link

  1. 1
  2. 2
  3. 3
  4. 4
  5. 5
  6. 6
  7. 7
  8. 8
  9. 9
  10. 10
  11. 11
  12. 12
  13. 13
  14. 14
  15. 15
  16. 16
  17. 17
  18. 18
  19. 19
  20. 20
  21. 21
  22. 22
  23. 23
  24. 24
  25. 25
  26. 26
  27. 27
  28. 28
  29. 29
  30. 30
  31. 31
  32. 32
  33. 33
  34. 34
  35. 35

↑ Back to top Tweet itShare on Facebook

Christian Holst is the co-founder of Baymard Institute where he writes bi-weekly articles with their research findings on web usability and e-commerce optimization. He's also the author of the E-Commerce Checkout Usability and Mobile E-Commerce Usability research reports.

  1. 1

    I typically don’t like using guest checkout because if I am buying something from a site then odds are I will be returning to that site in the future and I prefer not having to re-enter all of my information every time I visit the site. That’s not to say that I have never used guest checkout.

    I use Amazon’s 8 step checkout processes fairly often and even though it’s a lot of steps they make it pretty easy to go through and it doesn’t take very long. I think that is a good example of a site that uses a lot of steps effectively.

    • 2

      I disagree. Amazon uses a long process but it is not optimal. Amazon cannot be compared to other sites that don’t have any brand value or recognition. Apple could ask you to go through a twenty step process and you’ll do it. The Amazon user flow is not optimal. I’ve seen may flaws in plenty of the checkout flows out there. Many steps can be incorporated into one, two , or three.

      • 3

        I’m incredibly impatient when it comes to buying things online and simply shop somewhere else if I have to sign up for something (unless it’s unique to the store- in which case I may begrudgingly continue). I think it’s mostly because I am AWARE of better alternatives; so perceive their lack of implementation to be laziness/carelessness on the part of the store I’m (not) buying from.

  2. 4

    Recently faced the biggest challenge to redesign the checkout process for one of the client, after designing 5 – 6 different checkout process, still the client wants all the details every time user finishes the earlier one, its so frustrating for the user as well as for the UX designers to ask the user to enter same thing every time, this information provided above is really helpful for me when and what to consider while I design the next checkout process for the next client, I can give them a reference material for them to understand, what exactly user wants.

    Thanks Christian for providing such vast information in such a short article. I will be using it for my future checkout designs.

  3. 5

    Solid article! It’s interesting to see there’s actually data indicating that longer checkout processes aren’t actually causing more user agony if they’re done right.

    I want to add that it is easier to get the customer to create an account (and thus, sign up for your newsletter) if doing so adds value to your product or their experience somehow. You have to tell them about this up front. Standard features are order tracking/modification, easier customer service, easier purchasing next time around. In some cases, you get member/repeat customer discounts, some additional features of the product (especially software), or access to members-only parts of the site.

    • 6

      Christian, Baymard Institute

      September 4, 2012 11:46 pm

      Yes, explaining the benefits of an account is essential. Just as essential as actually providing some benefits that the customer would want ;)

      One site I once encountered e-mailed their account customers 14 days before the warranty for any of the customers prior purchased products was to expire. Telling the customer that “the warranty for product X purchased 1 year and 11 months ago is about to expire”, and he should therefor check the product thoroughly. Great service and utilization of their account data.

    • 7

      Michel Hendriks

      September 7, 2012 1:58 am

      Why “thus”? Account and newsletter should always have been separated from each other. In the minds of consumers apparently this is not the case, due to years of bad behavior by webshops probably, but just as an account really offers limited advantages, the two just shouldn’t be linked at all.

      Order tracking and easier customer service are of course nonsense. There is no reason why order tracking can’t be done by asking a customer for their order number and the email address used to place their order. Easier customer service; same thing. Faster checkout and perhaps order history are about the only advantages of an account, apart from any discounts that the store decides to give the customer (but again, why relate that to an account, unless you know that the fast checkout will improve conversions for return customers so much that it’s worth it).

      My advice to webshop owners is always: move account creation and newsletter signup to the ‘thanks for your order’ page. Customers have just completed the checkout process (which is by nature a very frustrating process) and so they’re happy that their purchase is on the way. Psychologically, this is the best time to get them to say yes! Also, all the customer needs to enter to create an account at this point is a password (or use Facebook login to make it easier still). And finally, the thanks for your order pages are so boring most of the times. “End of the line, you can go away now!”

  4. 8

    Interesting article, the numbers are fascinating. As indicated this area of production has become most looked over.

    The most astounding statistic the article gave was “24% require account registration”. Therefore, extending their checkout process nearly double the average!

    It appears these companies are paying strong attention to their CRM model. But at what point does this form of data mining become detrimental to the UX. It just seems as if there is never a clear cut road but, for those companies who are grossing 1 billion and are within that 24% they seemed to have found that “Yellow brick road”!

    Thank you for the reading Christian Holst!

  5. 9

    Mr. Holst ‘dosen’t’ use spellcheck. :]

    • 10

      Christian, Baymard Institute

      September 4, 2012 11:38 pm

      Hi deborah,

      Thanks for pointing it out.

      I try to do as much proofreading as possible, but I’m not a native english speaker so it will never be perfect.

      • 11

        Davin Bergstrom

        September 5, 2012 4:15 am

        Haven’t found a native speaker that does it perfectly either :D.

        • 12

          Mr. Bergstrom makes a good point! Very thorough and well-written article for a non-native or a native speaker, Mr. Holst.

  6. 13

    These are top 100 grossing sites.
    Many of the site that are in ‘violation of usability findings’ gross $1 billion +.

    Who’s to tell them that they are doing things wrong?
    If my site is grossing over $1 billion, I must be dong something right.

    • 14

      I see your point and I always wonder how do you say to someone who’s earning zillions dollars more than you that he’s doing it wrong.

      But actually, that’s not the point. It’s not that he’s doing it wrong, it’s that he could do it better!

      Everyone who’s grossing $1bi probably must be smart enough to see when he can gain even more!

      • 15

        Based on the success of many of the aforementioned sites, there obviously is some discrepancy between what people express during usability tests and how they behave in real life.

        Most people who are already signed up and are regular users of these sites are conditioned to how the site works and are comfortable – again, judging by the sales figures – people seem to have no problem shopping again and again from these sites.

        Factor of familiarity has to play a part somewhere in the usability analysis. Users always will be quick to express their discomfort and frustration toward anything to which they’re not accustomed. Therefore, applying user test results literally has its risks.

        We as mankind will forever come up with an excuse for infinite improvements for anything we can name. But sometimes we need that familiarity.

        A pencil isn’t perfect (doesn’t stay sharp, needs sharpening every few minutes, can get messy, can break, wears out – what a drag..!) But the civilization as we know it seems to be comfortable with it despite its imperfections.

    • 16

      Christian, Baymard Institute

      September 4, 2012 11:25 pm

      Yes a 1B+ company most certainly are doing something right; your market timing, brand positioning, product design, pricing, logistics, SEO, page navigation, product images, community engagement, etc. And their checkout experience might be ok. But that doesn’t mean that their checkout experience is “perfect” and that they can’t improve it further.

  7. 17

    I don’t like guest checkouts.

    I prefer to design so they end up in the checkout regardless of where they come from and can autofill using their “account” (email/password) or they can simply fill everything in.

    This way they do not go cart (or product page) -> login/guest checkout page -> checkout.

    They go cart (or product page) -> checkout

    • 18

      Christian, Baymard Institute

      September 4, 2012 11:33 pm

      In our testing we found serval of such checkout design approaches, an in it self it didn’t proved to be a problem for the test subjects. But be careful how you integrate the “autofill using their account” fields. See the video in this article on how’s approach confused some test subjects:

      But my main concern with this approach would be in regards to returning customers with an account. To what degree do they overlook the account option and create a new account?

      That being said, the approach can certainly work well, it just need a much more careful implementation than the more widespread “cart -> login/guest -> checkout”

      • 19

        That video doesn’t seem to be about auto filling at all? At least not the section. And the section is not relevant because we do not have the login/guest checkout step. Maybe I was not clear.

        Click on a checkout button -> checkout page

        We have increased conversions by doing this and offering an “invisible” guest checkout. If users do not “‘log in” but use an existing email the order does not end up on their account. This caused some confusion initially, however it was easily out weighted by the spike in conversions, new buyers and extra revenue. Plus since we run around 250 orders a day the few that do not end up on the right account receive polite phone calls from our customer service team :D

        Tip: Prefill early and people don’t notice. We prefill stuff before they even scroll down to it. We get their email first and prefill something 1500px(roughly) below.

        *Scroll down* oh my delivery details are there, nice. I can change them or leave them.

  8. 20

    Torkil Sinkaberg Johnsen

    September 4, 2012 10:44 pm

    DIBS, the largest PSP in the Nordic countries, has published a 2011 e-commerce survey, which has been conducted in Northern Europe mostly. They had a list of the top reasons for users to abandon the checkout process, a list which I think is highly relevant both for online stores and UX designers:

    1) Uncertain conditions
    2) Not preferred payment
    3) Lack of confidence
    4) Technical problems
    5) Insufficient information
    6) Registration process
    7) Payment process

    Notice how the registration process is only number 6 on this list.

    The survey:

    • 21

      Christian, Baymard Institute

      September 5, 2012 5:12 am

      Hi Torkil, thanks for the link, it’s a very useful survey indeed.

      In regards to “registration process is only number 6 on this list” then it is very unclear how DIBS define registration process, and how they asked the survey respondents. E.g. if you combine “registration process” (19%) and “payment process” (17%) it would probably be no. 1 in their list.

      And what does “5) Insufficient information” cover: insufficient info for a product the customer would like to buy, or insufficient info for a password form field during registration?

  9. 22

    Excellent article/study!

    Is there a similar benchmark study coming for mobile/handheld devices as well? Would love to see one and would most likely pay a few bucks to read it as well.


    • 23

      Christian, Baymard Institute

      September 5, 2012 4:56 am

      Yep, we’ve finished the actual usability test sessions for a M-Commerce Usability study and are currently analyzing the results. Hopefully it will be out Q1 2013.

  10. 24

    I have always been one for giving customers the option, if I’m going to be a returning customer I’m happy to create an account, but for things that I’m likely to only use once every couple of years (eg Hotel bookings in the US), then I want the option to simply just pay for the product and get my receipt.

    A good thing in the UK for customers is now you can’t have the newsletter auto-filled, it’s opt-in now rather than having to opt-out.

    Being a UI designer I really see the issues with check-out processes, yes Amazon is easy to use, but its long, annoying and could be done so much better. These companies can be making billions of dollars, but imagine how much money they could be making with a smoother checkout experience.

  11. 25

    You left out one of my least favorite things that sites do, which is to lure me in with one price, make me complete 5-9 steps, and then spring an egregiousness fee of some kind on me at the end. I want to know how much something costs right up front.

    • 26

      Yes this is one of my biggest pet peeves. They obviously figure that you’ve invested that time filling in details and you’re less likely to back out getting notice of a fee late in the process. Personally I’m the opposite, if sites pull that on me then I will not buy from them. Ever.

  12. 27

    Kick-ass article :)

  13. 28

    Christian, Thanks for this article!
    It is worth taking a look at or where we implemented a lot of the ideas you bring out in your book and in this article

  14. 29

    Re: large companies’ compliances vs small companies. I have and still am been working with billion dollar companies and tiny startups. I am sure that everyone who has had the same experience would agree with me: making changes in large businesses is way harder than in small ones. Simply the approval process, mixed with opinions of numerous people who are feeding their egos on their position titles is a wall of resistance. However the changes (even drastic ones) often do happen, however much much more painfully as there are usually many people working on the same task, increasing the livelihood of shit going wrong.

    You also may want to consider that the users of a large company might have their own system for doing things and just because something is not up to code, it does not mean that bringing it to it will make their lives easier. They will often resist as well.

    On the other hand, squeezing every penny is the lifeline for small enterprises. When there is not enough user base to make reliable experiments, no customer culture – there is nothing but the arbitrary guidelines. That is why it is so common that small businesses have better looking, more usable web pages. But it is not a rule for everyone, it is not a religion and it will never be complied to by everyone.

    My 2¢ :)

  15. 30

    This is a cool UX artcile, and I’mma let you finish, but it’s hard to take you seriously when your graphs are incomprehensible to people that are red/green colorblind. Just saying.

  16. 31

    Great article – thanks. Interesting comments too.

    My thought on registrations is if customers can see no point to it other than facilitating unwanted newsletters then they will resent the effort required however if there is an explained benefit then it is less onerous; for example it will save time on future purchases (once registered Amazon shopping is a very simple login and click process) or access to additional tools such as parcel tracking then there is a real user benefit to offset against the extra typing.

    Another angle to this is the naming of the process. “Registration” is suggestive of an additional activity (for the company benefit) to the intended purchase process. Naming it as first part of the purchase process/Checkout and then pre-filling those details thereafter into payment and delivery will make sense to most purchasers. Also state that they will only need to do this once.

    Once the purchase is completed offer the function to create username and password for a repeat visit (but offer incentives to do so as above) . If you cannot provide of a good reason why they should do so then the customer will also conclude it is in the company interest and not their own creating resentment or avoidance.

    I agree with previous comment add the newsletter option on thanks page where you can also direct to social media channels.

  17. 32

    I was interested in the address validation section and the number of e-tailers that don’t allow you to proceed unless you give an exact match to their address. A sure fire way of stopping a transaction dead in its tracks if ever there was one.

    In the UK, we typically validate our address details by entering the postcode/ zip code and choosing from one of the 20 odd validated addresses that appear. Whilst you can’t do this with a zipcode, there are other methods which exist to automate and validate the address capture process based on auto suggest methods found in search engines.

    In all cases you should be able to correct what’s provided as the data is never perfect.

  18. 33

    Nordstrom asks for the email address during the checkout. The fact that they don’t require a password is meaningless. In the back they may create an account with a random password and people might be able to “retrieve the password”. They still have the email address and they might send you promotional messages.

    I’m sorry to see that UX is actually about catering to the lowest common denominator. Dear web developers, you are contributing to dumbing down the population for making things too easy for them. In the last 6-7 years the checkout abandonment rate has increased by 15% although you jumped through hoops to make it easier for users. Why? Because you are helping users become more impatient and less accepting of small “defects”. You taught them how to give up if they can’t figure something out in less than 3 seconds and now you’re harvesting what you planted.

    Don’t get me wrong, you’ve done that for the same reasons athletes use drugs: everybody else around you is doing it and if you’re not doing it, you won’t “perform”. I’m just saying that next time you’re glad that because of you the conversion rates are up think that the general population IQ just lowered a bit. Think about it.

    • 34

      The money the lowest common denominator has is worth just as much as the money everyone else has.

    • 35

      Jep, thought about it. Rejected it. Your argument basically is:
      “tools should be at least challenging to use to increase the mental capacity of the user” – or in other words “Hammers and Screwdrivers should never be sold assembled – rather provide the parts and a construction manual to intellectually challenge the wielder”.

      Sounds ridiculous? That’s because it is.

      The purpose of design in general is to facilitate the ease of use of any object. That’s true everywhere. To turn around your argument: just because UX designers have come up with good ideas to facilitate checkout flows other checkout approaches lose – the users rather use a tool that works as intended than to fiddle with something (intentionally) incomplete.

  19. 36

    I redeveloped a checkout for the company I am Creative Director for last year. We do ecommerce for the majority of Australian sporting clients. We took a 6 step checkout back to 3 steps and modified most of how the checkout actually worked. The upshot – improving the checkout funnel from roughly 35% up to close to 50%. Those figures are averaged over hundreds of thousands of transactions. With a few companies turning over more than a $Billion per year with the worst usability you’d have to wonder if they got their checkout right how much more income they’d be generating.

  20. 37

    Manzoor Kottayil

    September 10, 2012 11:35 pm

    Good stuff. You have done a good job Holst.
    “Account registration = Spam mail”
    As a normal user several times I have thought like this. When I worked as a UX designer I am aware of that. Even I have scared about the spam mails (Newsletters :) )when I put comments in this smashing magazine comment box.

  21. 38

    Vincent van Scherpenseel

    September 11, 2012 2:27 am

    You state: “The average top 100 checkouts violate 33% of the checkout usability guidelines.”. I assume that you refer to the guidelines as stated in the Baymard usability report. Although I understand that you point to that source (as resources on this are scarce), such a statement serves no purpose if the guidelines themselves are not available without paying for the full report.

    Apart from that, great article!

  22. 39

    Thanks for this article – interesting research. A couple of points…

    “If we have an overall look at the top 100 grossing e-commerce checkout processes, the average checkout violates 21 checkout usability guidelines. This is an indication that checkout improvements are still much needed”

    Not sure I understand that conclusion. Sure all checkout processes can be “improved” but these are the top 100 grossing e-commerce checkout processes. As businesses are in business to make money, is it not a case of “job well done” to a large extent (and perhaps some of the guidelines are thus shown to be less valuable than they might be)? I’m talking from a commercial perspective.

    Also, can you actually “violate” a guideline? Assuming the guideline is intended to be advisory, rather than an absolute rule of course ;)

    • 40

      Great Article, solid numbers and very helpful.

      UX experts always want designers and developers to take UX into consideration during their work. Should we also think for other departments too, such as marketing. There must be a reason or even some solid numbers behind that “must-have-newsletter.” From a user perspective (a female consumer who is constantly looking for coupons online), newsletter sometimes really bring good news, believe it or not. A lot of times, users are not aware of the fact they will come back in the future. Just like TV commercials, annoying but working. We all have to admit that a lot of newsletters are tailored for the target audience. I agree with Chris that we should give users more controls over it, but at the same time, we should also use our design to influence and lead UX.

  23. 41

    Very interesting research here. I am surprised there is nothing having to do with the single sign on through social media or integrations with social media into registration or non-registration.

  24. 42

    Interesting read, although I’m surprised there seems to be so little focus on form field validation. This is a really developing field in terms of inline validation.. That i’ve not seen anyone crack.. although saying that.. and are pretty close..

    Looking through the top 100 grossing e-commerce websites i’m baffled.. Are USA sites really that far behind?! Lots of the checkouts in top 10 look awful. I can see
    – lots size 11-10pt font,
    – iffy qty fields and update boxes.
    – Lots of pages with tons of extra text
    – random mixture of colours and massive lack of consistency
    – Forms that have fields all over the place which are hard to scan
    – Payment options that ask for entire months and not as listed on card.

    Would be interested to see how the top 100 fares against UK based e-commerce sites.

  25. 43

    Guest checkout may not create a formal account but the only difference between the two in terms of data entry is that you don’t enter a password. All other information is the same.

    As you note, generally people don’t create an account if they think this is a one-off purchase. What we fail to do in the industry is allow them to later change their mind and let them create an account in a way that would link past guest orders to their account to allow for easy re-ordering. As in “make an account later” once you decide to be a repeat customer.

    So people create an account for insurance. If they are confident they won’t need to track the order and/or dispute the order then guest checkout is an option. If people are not 100% sure they will get what they ordered, they create an account assuming their order and all history of it will be available to them saving them the need to hang onto emails. But too bad if you need an order placed several months ago since a lot of the top sites clear out their order history and you find it is gone!

  26. 44

    The final “step” of checkout processes I’ve designed in the past worked like this: after the user has completed the entire process (cart to payment), the final screen displays:

    “Thank you for your business. Would you us to save your information and create a personalized account?”

    Somewhere on the UX are short bullet statements underneath that outline why it’s a good idea:

    [checkmark] we don’t SPAM you
    [checkmark] faster checkout upon your next visit
    [checkmark] easy order tracking

    Then, finally, there are Yes/No buttons. If you click yes, a random password is generated and emailed to the user with a nice welcome message and additional details concerning how to access their account. If “no” is selected, the user is either taken to the home page or a message is displayed, e.g. “Okay then!”

  27. 45

    Hey this is super interesting stuff. @Christian, could you expand on the experience from the Automotive Parts ecommerce experiences?

    I’d love to see any other resources / studies that exist or that you’ve put together.


  28. 46

    I’m not sure if my subscription info came via to you nor am I clear how to pay the $5.00.Thanks this is actually a terrific idea if I can connect! Judy


↑ Back to top