the frequency a kenny chung blog

I usually keep my public shaming of corporations confined to my Twitter, but sometimes I encounter an experience so egregiously bad that I have to blog about it. For your consideration today, I present what’s probably the worst UX I’ve ever witnessed on an online payment form. This embarrassment belongs to Con Edison.

Typically, when you’re asking customers for money, you want the payment process to be as seamless and frictionless as possible. But Con Ed decided to go in the extreme opposite direction and provide the most disjointed and unintuitive system imaginable. It’s like they don’t actually want their customers’ money.

Below are screenshots detailing each major step of the process. An important note is that in order to pay your bill by credit card, you need to complete the following steps every single time. So you can see how this might get annoying. Click the images below for additional commentary.

Step 1 of the online payment system of ConEd
Step 1: From the start, ConEd tries to limit what you can do. Here, the links are coded in Javascript, so you can’t open them in new windows/tabs via conventional methods. This becomes increasingly annoying later on when you realize that you needed some information from your main account homepage.

Step 1a of the online payment system of ConEd
Step 1a: Even after you click on the link to pay by credit card, you’re asked two more times if this is what you want to do. Why the process needs three clicks to confirm your payment method is beyond me. But what you’ll see later on is that this doesn’t actually matter.

Step 2 of the online payment system of ConEd
Step 2: You’re taken to a secondary domain for payment, and this site has absolutely none of your account information pre-populated.

Step 3 of the online payment system of ConEd
Step 3: In one of the more puzzling steps (and that’s saying a lot), you have to input your home address where you’re receiving Con Edison’s services. Mind you, this is after you’ve already provided them with your account number.

Step 4 of the online payment system of ConEd
Step 4: Once again, you need to specify that you’re paying with a card. This form isn’t even specific to credit cards, despite the fact that you’ve confirmed on three separate occasions that that was your payment preference. They could’ve asked you how you wanted to pay in this step.

Step 4a of the online payment system of ConEd
Step 4a: Now you have to input all of your credit card information in some weird self-refreshing form. Text boxes appear only when you’ve input previous data, which makes no sense. Then you have to input your billing address. There’s no way to save any of this information, so you have to do it every single time you want to pay by credit card.

Step 5 of the online payment system of ConEd
Step 5: As a final low blow, Con Ed charges you a “convenience fee” of $4.75 even though you had to go through 5 excruciating steps to give them your money. Talk about a con.

I hate you, Con Ed.

It may be a bromide at this point, but “Don’t be evil” has become a suggestion rather than a mantra for the Mountain View giant Google. At the root of it, the tech titan has been outpaced by companies in other online spaces in which Google wished to also lead. In some cases, Google has returned fire in spades and succeeded. But for every Gmail, there are a few Buzz’s and Wave’s. In the past few months and years, it seems that Google has also become increasingly comfortable with implementing opt-out models for its services, despite the consequences.

Let’s start with the big one- Google Buzz. Back in February, Google launched this poor attempt at a (US-based) social network. But what’s even worse was the way users were brought into the system:

Axiom #1: Google wanted all of their users to join Buzz.
Axiom #2: Google had already built up a ton of Gmail users.
Conclusion: Everyone who uses Gmail should automatically be opted into using Buzz.

The way this was done was a huge privacy oversight and a case study on how to alienate everyone. Not only did users see that they were opted in after Buzz launched, but they also soon found that the default settings made all of their mail contacts publicly accessible. You can imagine how this would be a problem for doctors corresponding with patients, men keeping in touch with ex-girlfriends, and people networking with recruiters to change jobs. In short, it was a privacy policy and a public relations nightmare.

Gmail Compose buttonNow, not every Google product launch arrives with such fanfare or contempt. But Google does have a long history of rolling out services, redesigns, algorithms, etc. in extremes and then inevitably scaling back after some much deserved negative feedback.

While this isn’t inherently evil, it can lead to bad User Experience. Just this past month, Google relaunched its Image search and redesigned the Gmail sidebar. I won’t delve into the plethora of reasons why I dislike the new Image search (it all comes back to Google Page Load time and bad UX), but it was a drastic change and people had no choice when it was initially rolled out. And now, almost all Google users have to deal with it. More recently, Gmail was redesigned with social/user connections at the forefront. Contacts and other Google services were emphasized over actual mail options and filters. The only update remotely mail-related was changing the “Compose mail” button into a clunky Web 1.0 grey box. While I understand the importance of maintaining a consistent brand experience, a lot of these changes don’t help most users. But does Google care?

Recently, there has been a lot of talk about companies being “too big to fail.” With entire industries contingent upon its search engine, Google definitely falls under this umbrella. Barring regulatory review, I don’t foresee any changes in Google’s opt out model practices unless the changes affect litigable issues like Terms of Service or Privacy violations. In general, with Google as powerful as it is in its specialized sectors, there’s little anybody can do about it besides send angry e-mails or blog about it.

In his book “Predictably Irrational,” Dan Ariely recounts one of his most famous observations regarding the prevalence of organ donors across different countries. You can read more about the organ donation phenomenon on his blog, but the basic gist is that countries with an opt-out model have significantly more citizens registered as organ donors. Compare this to places like the United States, which has an opt-in model (you have to check a box on the back of your license to become a donor). The proof is in the pudding- people are lazy. That’s oversimplifying it a lot, but the bottom line is that the path of least resistance is usually the most appealing to the majority of people. Making that extra check mark or reading every line on a form is not worth those precious seconds of mental processing.

Why do I bring up this study? Because the propensity to gravitate toward inaction cuts both ways. By changing the options, you can influence how users opt in or opt out of actions/services at drastically different rates. But to quote Spider-Man’s fictional uncle, “with great power, comes great responsibility.” As I’ll explain later, the ability to affect actions doesn’t always help in the way you might have hoped.

Let’s start with a simple example: consider your typical eCommerce shopping cart system. Buyers reach a registration screen, where they have the option of signing up for a free, optional newsletter. There will be a percentage of people who completely glaze over the option. There will also be a percentage who will actively check or uncheck the box to indicate whether or not they want it. If you only care about more subscriptions, a pre-checked box is the way to go.

Where can a pre-checked box go wrong? Just ask all the people who, at one point or another, ended up with a Yahoo or Ask toolbar on their computers by accident simply because they wanted to install a program or browser plugin. While it is perplexing to novice users who can’t figure out why they’re losing web browser real estate, it is downright infuriating for computer savvy individuals who impatiently click the “Next” button waiting for installation to begin. This is where you can hurt your brand. Your service can be labeled spammy or adware. Some will view it as an underhanded (and downright desperate) tactic to have people use your product. It also violates people’s privacy (as users did not explicitly give permission to install, which is, in a way, kind of like the opposite of organ donation).

So back to newsletters. Everybody wants to believe that they have the best, most useful, awesome newsletter; but this simply is not the case. For example, I used to subscribe to Flavorpill, but realized it consisted mostly of thinly veiled advertisements and self promotions. On the other hand, I know many people who swear by Gilt Groupe’s newsletters on daily deals and spend inordinate amounts of money as a result. If you actually do have a good newsletter, then you’re fine. You might get some accidental subscribers who recognize the quality of your work and its utility, and you’ll also have subscribers who receive exactly what they were expecting.

But take for example concert ticket e-mails. I subscribe to Ticketmaster and LiveNation for the off-chance that I’m unaware that a band I like is playing in my area. Both do DMA-style targeted newsletters which are also genre-specific. But I also end up randomly getting single e-mails about a Nickelback or Jonas Brothers concert. Every time I get one of those, I want to unsubscribe.

There are many ways to create a bad newsletter: be too text-heavy; be too generic in your content offerings; don’t include any accompanying visual aids/pictures; don’t make your links readily available and clickable; don’t include a call to action; include important information below the fold, etc. Those are mostly design aspects, but there’s also an important marketing piece- who do you actually want subscribing to your services?

There are so many ways to demo target: retarget users and a serve popup ads; run SEM campaigns on specific keyword phrases that indicate who a person is likely to be; utilize social media channels for awareness campaigns to reach certain types of readers, etc. With all of these other, much better, tools at one’s disposal, is it really prudent to automatically opt everybody in every time? Are you even using the best metrics? That is, do subscription events matter when 75% unsubscribe after their first email? What if the people who stay subscribed never click through to the site? Retention rate would be a lot better KPI, but still not the most efficient. If you’re not sure who wants your content delivered to their inbox, then let users opt in, and also do a better job of targeting them across different interactive media.

Also, stop trying to make us install your resource-hogging toolbars!

My next blog post will single out a particular company that follows the opt-out model (and alienates a lot of users in the process). Are they too big to fail? Are they doing evil? Stay tuned to The Frequency to find out!

Creative Commons License
licensed under a Creative Commons Attribution-No Derivative Works 3.0 United States License.