Hacker News new | past | comments | ask | show | jobs | submit login
Don't send me a message to tell me I have a message (darrennix.com)
155 points by darrennix on April 18, 2013 | hide | past | favorite | 83 comments



Most companies doing this are not naive in their design. They want you to come back to the website so you can read it and hopefully be lured into doing something else on site while you are there. I agree for the PSD2HTML example it is silly though.

Most social networks and dating sites are the biggest offenders as they want you to continue browsing or in some cases, upgrade to pro to read the message.

It may not be 'desirable' ui/ux in the endusers view, but when its done like that I suspect the provider has found out "this makes me more money both now and in the long run"

I think we could just as easily see a post on HN tomorrow saying "One simple tip to increase your email click rates 2500%"


You hit the nail on the head. It's getting very annoying to read posts by people who get up on their high horse and proclaim "this is terrible design and I'm shaming companies who do this"... except that's not the case. It's just a mildly annoying thing for them but as far as the sites are concerned this is great design. Rather than go around saying "I proclaim this wrong" we should start saying "I don't like this" because in the end that's really what most people are saying and they come off like dicks when it turns out there's a pretty good reason for the "wrong" thing to be such.


Dark patterns suck. I'll shame companies that do this, regardless of how efficient it is. We should be focusing on selling products, not manipulating clients into buying them.

A business that needs this kind of practices is a sloppy one that will collapse before it gets big. It's a waste of time and money on both sides.


It's good design to provide a worse experience to your customers?

This is why I've always enjoyed Apple products. They don't perform these kind of ridiculous mental gymnastics to justify doing a worse job for their customers.


That sounds like a rather credulous statement in the face of regularly imposing restrictions in the ios ecosystem that punish or restrict users to damage a competitor. It's difficult to believe the ebook war was pro-customer for example.

They clearly usually make decisions in good faith, and generally fuck with the user less often than the competition. Unless it's a core strategic goal in which case they're arguably the worst.


"It's good design to provide a worse experience to your customers?"

Define "worse". People will complain about doing all sorts of things that are better for their long-term interests.

If you're building a product that provides greater value to users with greater user engagement, then it's in everyone's best interest to promote user engagement, even if that means upsetting the people who would rather gain the benefits of your product without actually using it. You can't make all of the people happy all of the time.


> Define "worse". People will complain about doing all sorts of things that are better for their long-term interests.

It's telling that you're comfortable deciding on your user's behalf what is better for their long-term interests, in a way that just so happens to benefit your perceived interests.

> If you're building a product that provides greater value to users with greater user engagement, then it's in everyone's best interest to promote user engagement, even if that means upsetting the people who would rather gain the benefits of your product without actually using it. You can't make all of the people happy all of the time.

If you're building a product that provides greater value with greater engagement, then it's in everyone's best interest to drive engagement by giving users value that makes them want to be engaged, rather than playing tricks like annoying them with e-mails that contain no content.


This is quite ironic, since I'm pretty sure Apple is the king of deciding what's in their users' long-term interests in a way that's actually better for their interests.

Some examples: removal of the floppy drive from the iMac, removal of the optical drive from the Air, removal of "Save As" from MacOS, removal of Google Maps data from the native iOS maps app, "you're holding it wrong" response to the iPhone's antenna issue.

Just because you agree with these decisions or the mental gymnastics required to justify them, doesn't mean they can be ignored.


Do you actually believe Apple made those choices to benefit themselves at the cost of their users?


There're plenty of self-interested design decisions in Apple products. Don't be naive.


Clearly I hit a nerve by praising Apple. You can argue that they're doing things contrary to what some user's might want, but I have a hard time seeing where they're doing it for self-interested reasons.

As far as I can tell, the people on the ground there genuinely believe what they're doing is what's best for their users. That's quite a bit different from knowingly sacrificing user experience for something you know isn't best for your users.

It's a shame that the leading comment on hacker news is that smarmy and user-detrimental business behavior is a good thing. Then again, the audience is as much "startup hustlers" as it is hackers, if not more so.


The reality isn't quite so simple.

Take the email case for example - while it's mildly annoying to the user to have to click on the link to read the message, it may be proven by data that by doing so they're improving their experience with the product in the long run.

Take a dating site as an example - you click on the link to read the message, which causes you to browse more profiles, which causes you to message a few more users - improving your chances of success overall and creating more interactions for other users. For a service whose goal is to get as many people interacting with potential partners as possible, the slight bit of anti-user design serves a greater purpose and improves the end experience not only for you but for the userbase as a whole.


> Take a dating site as an example - you click on the link to read the message, which causes you to browse more profiles, which causes you to message a few more users - improving your chances of success overall and creating more interactions for other users.

So you don't think users are capable of making that decision on their own? There's no other way to provide a fair exchange of value so that users want to visit the site, other than holding their message(s) hostage?


They're perfectly capable of making that decision on their own.

But they won't, because the amount of knowledge transfer required for them to make that decision is beyond what they're willing to invest.

Take the dating site example - for the user to act optimally (view many profiles and message many people), they'd have to understand not only the full mechanism of the website (pretty simple to do), but also usage trends, psychology and attitudes towards online dating, social attitudes towards dating in general, and a slew of other topics.

All in all, the "here's what you should do and here's how we know" conversation is several hours long. If there was a Matrix-like way to jack your users' brains in and explain the full scope of why they should be behaving a particular way in an understandable, absorbable way, I'm all for it.

But alas, we can't do brain-downloads. Yet.

But what about simply informing them without the messy explanations? Well, it turns out this is what we did in the infancy of consumer websites - and it never worked. "You should message more users" falls on deaf ears - even after many, many attempts to reframe the message in more compelling ways.

This occurs for many reasons, and differs depending on the exact message being conveyed. For dating sites one persistent trouble is getting your users to upload good photos of themselves. We know from an insane amount of data that people make online dating decisions in a highly visual way, and that anyone without a picture is basically shit out of luck - but yet you wouldn't believe how many users think that rule applies to everyone but themselves. Even explicitly throwing up big warnings about not uploading a photo only convinces a small chunk of users to do it - and in the end you'd resort to tactics like holding their search results hostage until a photo is uploaded. And it works. And it improves their experience dramatically because no one replies to someone who doesn't have a profile picture.

I can keep rambling and naming examples like these. There is no shortage here.

I don't think you'll find anyone who thinks that hiding your users' data behind a mouse click is the best thing in the world, but you will find many people who have actually been in the trenches, who have done it the "right" way, and in the end found that the "wrong" way is a necessary (and very minor) evil.

You've also made the assertion elsewhere in this thread that people who work on these products don't care about user satisfaction. This is pretty far off the mark - user satisfaction is one of those things that's pretty easy to measure... and the unfortunate truth is that, in most circumstances, the minor loss in satisfaction from having your message hidden behind a link is more than made up for by the overall improvement in experience.

It's a somewhat common stance from people who haven't worked in consumer web before to assume that these strategies are the result of arrogant people looking down and disrespecting their user base. The reality is that no one I know assigns any value judgment to any of these observations - but we will follow where the data leads us.


> It's a somewhat common stance from people who haven't worked in consumer web before to assume that these strategies are the result of arrogant people looking down and disrespecting their user base. The reality is that no one I know assigns any value judgment to any of these observations - but we will follow where the data leads us.

You can either target the high road, and the customer base (and margins) that come with that, or target just the data while aiming only for mass appeal. I've always worked for companies that target the high, but to each their own.


All of what you've said boils down to one thing: sites and companies doing this put their interests and concerns ahead of yours, utility, and functionality.

Which is another reason I make very little use of such services.


Is it not in the interest of a service's users for that service to operate at a healthy clip of profitability? Email alerts like this can make a very big difference in clickthrough rates, implying both that users find them useful and that the company would take a significant viewership loss if they stopped the practice.


Look at what Google did that other portals weren't doing.

Yahoo, AltaVista, Excite, Jeeves, etc., kept trying to get stickier and sticker, and reduce bounce rate.

Google sought to get increasingly relevant, and increase bounce rate. A user hitting the site and not clicking through to another page was viewed as a success, not a failure. Sure, it's a lost ad impression, but it's a satisfied user.

By putting the user experience first, Google completely trounced a "sewn-up" market.


Google won by trouncing the others in search, but Google has expanded into all the same kinds of services that Yahoo, AltaVista, et al were attempting to provide; there is Google Mail, Google News, Google Finance, etc., for the same reasons that Yahoo and others provided this -- because they live and die by page views.

Google simply attacked the problem from a more intelligent angle; first, they greatly focused on providing hugely, undoubtedly better search results, which, despite the "niceties" of Yahoo et al, was a game-changing deal. Secondly, they outsourced their ad network much more effectively than everyone else, so that most of the time when you followed a link outside of Google, you'd end up on a page that showed ads from Google anyway, so Google still made money. Thirdly, they started wrapping up as many services as possible under their umbrella, both to increase page views and to increase the intelligence and targeting of their ad delivery.

So Google really did end up following the same approach, they just did it in a cleaner, more subtle fashion, and I don't even think that's the main thing that allowed them to excel; the primary reason they killed Yahoo and friends is because their search results were ridiculously superior, so there was no way Yahoo could compete despite its "personalizations". If Yahoo was smart, they would have licensed the tech from Google early on and Google could've been an IP vendor to the extant search engines, licensing out its algorithm. Yahoo's dominance would have remained that way, because once a certain critical mass is reached, the inertia is difficult to overcome unless you have a very serious improvement over your competitor's product.

The secondary reason Google killed Yahoo is that they aggressively expanded AdSense so that when you followed a link off Google's site, they were still making money, and the tertiary reason is that they bundled as many services as they could under the Google umbrella, increasing data to mine for ad targeting and increasing ad views, expanding even to your cell phone (that is, they did the "portal" thing).

Yahoo et al neglected points 1 and 2 because they thought they had it locked up. Google came out of left field with a far superior search system, and that was really the killer app.


This is correct. I've observed dramatic differences in click-through rates when including even a tiny bit of message text in the email notification you send to users. There's not one right answer.

What (some) users want isn't always the same as what's good for your business, so you have to be judicious.


I can't help but wonder what the long-term business effects are of intentionally damaging your relationships with your customers by wasting their time and choosing to prioritize arbitrary business priorities (eg, e-mail click-through rates) over actual user satisfaction.


"I can't help but wonder what the long-term business effects are of intentionally damaging your relationships with your customers by wasting their time and choosing to prioritize arbitrary business priorities (eg, e-mail click-through rates) over actual user satisfaction."

Exaggerate much?

As I said to your other comment, sometimes it's in everyone's best interest to promote engagement. If you're running a product with a strong social component, it isn't an "arbitrary business priority" to want to get people back to the site.

As in all things, good judgment is key.


> As I said to your other comment, sometimes it's in everyone's best interest to promote engagement. If you're running a product with a strong social component, it isn't an "arbitrary business priority" to want to get people back to the site.

If you can't get people back to the site by providing value, then it's worth spending time figuring out how to provide that genuine value, rather than trying to force your users into it.

As you said, good judgement is key.


In all honesty, though, email is not exactly a secure way to send a message.

Now, I'm not disagreeing with you, but there are times when it makes sense.


If someone has access to your email, it is almost certainly trivial for them to access your account for the service holding the message.


I'm not arguing with you, but that doesn't dispute it's insecurity though. You don't have to use email to reset a password; there are other methods.


Pretty awesome that the company the OP is complaining about has already posted a comment in reply to the blog post (on the actual blog)...

-----------------------

Darren,

You're right, this sucks. Looking into this now and see how this can not happen again.

Michael La Rosa CMO at PSD2HTML.com


Awesome? Really? Not in the slightest.

The CMO just sent him yet another useless, contentless message.

If he had any sincerity he would have first looked into it, then fixed it and only then replied to say that he had fixed it already.


Not in the slightest? So the CMO gets no credit for acknowledging the mistake?

I would be willing to bet that the CMO is not in any sort of position to make the necessary changes to ensure that this doesn't happen again, and so your suggestion is beyond his ability. By acknowledging the customer's frustration he lets the OP know that the company understands that this is an issue, and is now going to make an effort to correct it. In other words, "he has been heard".

Obviously, the comment alone is not enough, but it's a vital first step towards high quality customer support, and probably an instructive lesson for anybody in a service business.


This is needlessly hostile. Being responsive is good. It's not contentless; he admits their current process sucks.


He also posted a longer explanation in the comments which to me seems pretty reasonable. Sounds like in this instance there's a switching cost they are trying to overcome. We've all been there. I think the article's point should definitely resonate for anyone building a new app, and/or anyone with an app that has a lower cost of updating.


"Lookin' to make a blog post on this bro. Keep checking the site, should be updated soon!"


There are some instances where this is necessary though. For example, my company has an internal messaging system and for HIPAA compliance we never send out the message contents or subject line. But I agree with xoail, most companies are just trying to get you back to website.


Dear customer, we have a Hippa/PCI/confidential-in-some-other-way message for you in our in-app message center. Please log in to view it. Yours, The Management.

There, that was easy. Otherwise, just send me the message.


Agreed. Insurance providers, banks and other people required to send secure messages probably already have you as a customer. They're not trying to drive traffic to their site for an extra ad impression.


My bank does that with bills. Honestly it is one of the reasons why I hate my bank, but the worst part is I can't turn it off.

Just send it to me, my email is better protected than your site anyway.


This is often a security feature.

If my bank has an important message for me, telling me to log in to my account to look at it is a good idea.


The sad part is that the majority of messages (at least in the case of Chase bank) are duplicates of what are in the email, or at least able to be figured out from the email (e.g. a large deposit has been made). I'd be concerned that with the frequency of emails received, it might prevent people from logging in when there actually is a real important message.


I would honestly be applaud if my bank or hosting provider sent any sensitive information via e-mail.

"Switch to email. Trusted since 1971."

Sure - Email may be 'trusted' but that doesn't make unencrypted plain-text messaging over the Internet secure.


If my bank sent me an email with a link, I'd be yelling to my account manager in a few minutes.


Telling me (or my wife) to log into my account for sensitive information may be a good idea, but including a link in the email for the purpose does not seem to be. It opens possibilities of phishing attacks.


Agreed, I would prefer that my account information be accessible at a secure web site rather than being sent out in cleartext email.


Sometimes that's the best way to do it. HIPAA requirements are a pain, but they exist for a reason.


Agreed, I automatically exempt anything dealing with healthcare or finances from stuff like this. I definitely don't want my bank emailing me customer service messages which could contain who knows what kind of sensitive information.


Bank simple does this, it is a security measure, imagine a random person just opens your email in your phone (most likely note secured) and reads a message that reveals things like account balance.


I'd like to think that if a random person got ahold of my phone and was in my email, the balance of my account is the last thing I'd worry about.


Perhaps, but I'd like to worry about N-1 things


Though annoying for the user, it's the best way to engage the user. Bring back to the website and boost up hits, opportunities to cross-sell, upgrades, or drive retention. Besides sending the message itself in the notification adds a bit of technical complexity.


I initially felt this way about forum posts, but after reading Joel Spolsky's take on it in Building Communities with Software,[1] I changed my mind. Basically the reasoning was: do this and you'll never get your forum to critical mass -- new users will just read replies to their questions in their inbox and never come back to the forum and contribute.

He never says it's not okay to enable this functionality after you have a solid userbase, though. =)

[1]: http://www.joelonsoftware.com/articles/BuildingCommunitieswi...


Worse: Sending me an email after I unsubscribe because I don't want your emails, to let me know that I've unsubscribed.


Just because it's a pet peeve of yours doesn't mean it's bad design. It's standard practice and should happen all the time.

If a user likes me enough to forward my newsletter to 60 of their friends and one of them clicks unsubscribe because they don't take 10 seconds to read the subject, I'm out an extremely loyal reader/customer if I don't send that notification email.


Let's rephrase, using two scenarios -

One:

"If a user likes me enough to forward my newsletter to 60 of their friends and one of them -who also happens to be subscribed to the same newsletter- clicks unsubscribe because they don't take 10 seconds to read the subject -and then scroll through the email and still fail to recognize that this is a newsletter they had also subscribed to-, I'm out an extremely loyal reader/customer -who can't even recognize my newsletter- if I don't send that notification email."

and Two:

add to the end, "when both problems could have been solved at once by requiring the user to confirm the email address they wanted unsubscribed when they clicked on the link, thus providing both verification, and ensuring that a friend does not mistakenly do the same."


I think you misunderstood my point.

1. You receive a newsletter from [Retailer] and forward it to me.

2. I click unsubscribe

3. Ideally I'd enter my email. No damage done. But what if the unsub link is like a lot of unsub links, and it's something like: http://email.retailer.com/unsub.php?e=firebeyond@ycombinator...? Then you get unsubscribed, I've done nothing to make sure I don't get the emails anymore, and if [Retailer] doesn't send you a confirmation, you never get another email and you probably forget about them.


I like those "you have unsubscribed" emails for another reason beyond what you layed out.

I collect them so that I can keep a record of what I've unsubscribed from, so if I ever get spam from them again I don't have any doubt as to whether or not the sender broke their promise to unsubscribe me. It might just be me, but it is doubly aggravating to get spam and feel like I had previously unsubscribed but not be able to prove it. With proof I can be righteous in my anger and blacklist all future email from them. Without proof I am left to doubt my self and wonder if unsubscribing (again?) would be futile.


"If a user likes me enough to forward my newsletter to 60 of their friends and one of them -who also happens to be subscribed to the same newsletter-"

No, the people the user forwarded the email to are not subscribed. You have misunderstood the scenario.


No, it's still a bad design feature. Having just gone through this less than 60 seconds ago, it really feels like a slap in the face from a company I otherwise respect.


Do you actually have a solution to the problem that pc86 laid out? Otherwise, we're talking religious differences.


I will agree with you if you have to type your email in first.

The ideal situation, I think, is to have the user type in their email. If it's an existing subscriber, unsubscribe. If it's not, add them to a secondary list just to keep track of who definitely does not want the emails.


Pissing off your users is always a bad design.

In this case it is why I don't usually click unsubscribe, but either set up a filter or report it as spam.


I got one of these today. I thought "well that's annoying" but then I thought "nah I guess it makes sense. I wanna know they know they aren't going to email me any more."


I strongly disagree.

Every change to user account should send automated email stating what changed (and if email changed send it to both old and new address).

This is so you know that changes took effect, and in case somebody else tried changing info for you.

This includes cancellation/termination of service.

Think of it as a commercial transaction, you request an action, and after action is completed you receive receipt.


That's not so bad though


My favorite is nelnet.com (manages my student loans) who emails me to tell me I have a message, after which point I log in to the website, visit the message center, and click on the message which starts my browser downloading a PDF usually containing several pages of terms and branding bullshit and a sentence or two of actual text which most of the time is nothing I needed to know anyway. And of course the link text of the message that I click on is something useless like "Message regarding your account".


I think I've only ever seen this from banks or medical organizations, where the annoyance is a little more acceptable.

I'd much rather they send me an (encrypted) e-mail, but I understand why they can't.


The corollary to this is "Don't leave me a message to call you back, instead leave me a message why you called." That way I can call you with the answer, or handle it via a better means (email, text, tweet, etc.), or delegate if appropriate. It wastes both of our time leaving a message to call you back. This is in my voicemail greeting. Seems to work.


I wish there was some mechanism to allow a recipient to upload a PGP public key, that way future messages could be sent securely directly to the account-holder in question. Added bonus - it would bring more widespread adoption to PGP and related PKI/web-of-trust email systems.


The problem here is that the most reliable delivery channel (email) is not sufficiently secure enough for a lot of situations. Worse, some users share email accounts, so you can't even depend on the auth of the account. If an app or service has to achieve a certain level of security, it can't rely on email as a direct message channel - it can only really use it for notifications that a message is available.

Contrast that with Android apps, where the app can show a notification at the system level but then handle (or retain) the service-specific login (and/or respect its rules about timing out the login.)


I like how easy it is to follow, comment and close the tasks on Asana: - They send an email with the full comment/update made on the task - Reply to the email to send a comment and it will update Asana - There is a direct link in the email to “View and edit” the task or to unfollow it. - You can reply with “complete” to mark the task completed - You can assign the task to a teammate by adding him to the email “to” field. All this without leaving Gmail.


The only place sending an email to notify someone of a message is a Good Thing(TM) is in highly sensitive or security conscious environments like medicine, law, finance and the like. Where the sender must be certain that the message is received in a secure manner, ie. over https.


Ha, I thought I was the only one to get annoyed by redundant messages: http://devel.io/2011/05/08/electronic-communication-etiquett...


It's especially irritating on sites that delete old messages after a certain amount of time. I'd rather opt-in to an archival strategy than have one forced on me by a site.


You have a new compliment on Yelp

Someone on Yelp has sent you a compliment!

See your new compliment

See you on Yelp,

— The Yelp Team


oDesk does something similar, but something I like about their approach, is that they include the message in the email.

There is a "Click to reply" button, but you can just reply to the email and it will reply to the message. Only benefit I noticed to actually using the oDesk interface to reply, is that any blank lines in your email response is omitted, while doing it in the oDesk interface keeps your formatting.


Rackspace support does this, drives me up a wall!


If you are discussing something private that relates to your account/server/network, the rule of thumb is to avoiding sending anything over the e-mail for security reasons.


I get these from airbnb. SO ANNOYING since there mobile site doesnt work on my windows phone and they have no native app.


Don't ask me "Can I ask you a question?"


favorite comment:

There has been a new comment posted, click the following to view the comment: http://www.example.com/?comment=324243&rerfer=http%3A%2F....


Also, don't send me an email to confirm that you won't send me any more emails.


Ah, the "indignant post-it note on the fridge". Works every time.


The worst are companies with both sites and apps and they send me both an email notification and an in-app notification for every fucking thing. And so now I've got multiple notifications going off on my phone for exactly the same actual event just in different formats (one notification letting me know I got an email about a change on the site, one notification from the app reporting the same thing as the email is). And then throw a digest email on top every day or so and the notifications quickly start spiraling out of control. I'm looking at you Facebook, LinkedIn, etc.

Granted, I can usually disable one or the other or both but it is a pain that I have to and I wouldn't even mind the notifications if I didn't feel like I was being assailed by them at times.


Sending me an email to tell me that I have a message with a link to actually read the message is unforgivably bad design.

I don't want to:

Why not? It's "unforgivably bad design" only if enough people don't click, browse and sign up for services. Who cares if they annoy me or you? It's a numbers game, thy don't need to please everyone. As the bad pitch goes"...if we only get 0.1% of users to..."


It's "unforgivably bad design" only if enough people don't click...

I understand your sarcasm, but 'good design' is an aesthetic, not a metric.

[edit: on second thoughts, I think I may have over-simplified things. Is 'good design' purely quantifiable?]




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: