When Websites Won’t Take No for an Answer
nytimes.com
nytimes.com
On the LinkedIn issues, this tweet surely is the funniest: https://twitter.com/darylginn/status/590664399041519617
On dark patterns, or at least it felt dark... I was booking an overnight ferry trip across Bass Strait in Australia. After selecting the fare, entering my details about my car make and model, name and address, cabin choice etc, and about to enter credit card details, the website proclaims "our real-time fare system has adjusted the price" and my fare went up $40 right there and then on the final step of the booking process. Not happy.
RyanAir in Europe is notorious for this. Cheap flights but expensive, unadvertised baggage fees.
I was booking well in advance - more than a month, and I was using the site after midnight.
There should be regulation against changing a price that is advertised, at least lock it in for a minimum of 5 or 10 minutes of selecting the fare. Bumping the fare up at the last step only a couple of minutes later is highly suspicious.
So in this case, mailchimp is getting marked as spam, even though they are an honest player.
If the unsub requires me to log in first, I mark it spam instead. If the link looks fishy and/or I can't readily discern it's of any quality it's spam. If I have no clue why I could've gotten the message it gets flagged for sure.
Luckily for me I get very few emails per month that I don't know why I would've gotten them, so it isn't a painful exercise to give them a chance to not get flagged.
I use Unroll.me[0] to manage the messages I do want to get, but want to check on my own time and I don't want them sitting in my inbox. I'm sure there are other types of services, but I'm generally happy with what they do. They also can manage a lot of the opt-out features as well, not sure if it 'protects' me any more than doing it myself though.
#6: https://www.ftc.gov/tips-advice/business-center/guidance/can...
Instead, I just blacklist the sender's IP address. It's even nicer when the sender (e.g. LinkedIn) publishes TXT RRs for SPF in their DNS zone as it allows me to easily blacklist all of them.
Whether you consider this good or bad may vary: it means MSPs have a valid claim that senders exercise Good Citizen practices on all mailings, not just those circulating via the MSP.
Then there are the major players. Yahoo's been a freaking PITA for years on outbound spam (both spam and joe-job bounces), with no effective reporting mechanism. Gmail's been largely better IME, other providers vary.
1. If I start receiving spam I know who leaked my address
2. If I can't switch off the spam hose, I just delete the alias, which severs any way this website can contact or bug me.
It should be generalised. All it requires is a sort of index to remember who was given what alias.
I was buying a ticket for a football game, but just to make sure there wasn't a better seat I clicked the back button on my browser to go back to looking at seats. Suddenly I saw that the same seat had gone up by 10% in price.
Anyway, I pressed the forward button to see if the new price would apply, but it didn't. So I somehow avoided paying an extra 10% because their confirmation page was cached, but not the seat selection page.
That's exactly what they do. If you're browsing for plane tickets and you're a bit undecisive (like me), don't be surprised if the prices go up by 200% for that specific day you wanted to book. Going incognito and in some cases returning after a hour will bring the prices back down.
What's really happening is that prices move around all the time.
I've done this with 2 browsers looking for a car on the same site and they will have different prices. I generally never log in on chrome anywhere and only use it for twitch so it was a lot cheaper than my firefox that's logged in everywhere. Like $150 on a $300 ticket.
While I don't have any other evidence than my pretty specific anecdotal experience to prove it's true (in addition to what friends in the airline industry told me), it is common knowledge that price discrimination is a pretty fundamental aspect of airline pricing so this "myth" is not that inconceivable.
That being said, a lot of people have done testing of this and found no evidence of repeated cookies specifically leading to higher prices. Here are some of the links I've found:
* http://www.usatoday.com/story/travel/columnist/seaney/2013/0...
* http://blog.getinvisiblehand.com/2011/04/08/ryanair-manipula...
* http://www.lifehacker.com.au/2014/08/ask-lh-will-deleting-co...
I'm not disputing that price discrimination happens. Just that by far the biggest determinant of it is time, as airlines constantly publish and unpublish different fare buckets.
In fact, what I suspect is happening when people see lower fares is just caching. All OTAs heavily cache prices in search results and then only fetch the updated price once you drill into results. I could easily imagine a scenario where the prices has actually gone up, but users in incognito (who haven't drilled down yet) see the old, outdated price.
The notion of specific price discrimination on specific users by cookies doesn't make sense though. For one thing, I don't see how it could possibly work technically. Information on airline prices flows in direction—airlines are constantly telling aggregators the latest prices/fare buckets, which the aggregators then display to users. Aggregators don't make a request to the airline for pricing every time you search, so how would they even know to update your price based on previous activity?
Signed up for pickydomains a while back, they send me emails every time there's a new listing. I decide I don't want to get the email anymore. Click the unsubscribe link. They want me to sign in. I sign in, and am unable to do anything without agreeing to their changed Terms of Service.
In other words, I have no way to get them to stop emailing me without agreeing to new terms.
Similarly, I once signed up for a trial of IBD newspaper. I still get emails from them occasionally (although those at least go to spam), and the unsubscribe link requires a sign in, which I don't remember. I could probably reset my password if I cared enough, though.
(These are just the two I remember, I'm sure there's been others that were similar. I had an issue with PayPal spam at one point as well but it was more complicated.)
As far as I can tell, both of those are in violation of the CANSPAM act. Companies are required to offer an opt out link with no additional requirements to opt out. But there's no consumer right of action, so the only one who can enforce it is the government, and they focus on worse spammers, so real businesses that are just aggressive mostly get off free.
(I don't think spam flags are personalized, so even to me future emails wouldn't go to spam. I could set up a filter, though.)
EDIT: To clarify: in Germany you can sue companies for spam and making the opt-out unnecessarily difficult. Of course the damages are so trivial you'll likely overpay but there are enough bored lawyers that these things tend to happen every now and then.
I looked up the federal law at one point and private individuals had no rights beyond reporting to the FTC. ISPs may have standing if they can show harm.
See https://www.law.cornell.edu/wex/inbox/can-spam_standing and http://www.dwt.com/advisories/9th_Circuit_Deals_Blow_to_Prof...
I see now that the majority of this document could have expired. Not sure where to find these "notes" that the document talks about. It also says federal law would preempt. So yeah... talk to a lawyer I guess!
As I'd just noted in another comment, many senders use a mail service provider, and a single bad actor can have very bad knock-on consequences, even for mail which doesn't transit the MSP (SPF records listing MSP IP space).
For many major brands / companies, unsubscribes work reasonably well. I've seen some success with this, don't offer email addresses generally, and flag the remainder as spam, with blackhole rules in my various filtering tools.
The larger problem with email is I simply don't trust it any more.
https://www.facebook.com/help/224562897555674
(If you can't get in, reset password, etc?)
I find it surprising that the researcher apparently makes the distinction whether a certain UI is "dark" or not depending on what it is used for.
As a customer, unless I signed up with the explicit goal of habit-change (e.g. for a fitness service) I would very much not like to be manipulated no matter if the cause is "good" or not.
That is stunningly poor reasoning. It doesn't matter whether a company benefits more than you do as long as you are not harmed. Not all games are zero-sum. Thinking you should be getting more or equal to the other party is not harm, it's envy.
I don't think that's true, because a company could have the means to do something profitably that an individual couldn't. For example, I just changed the toner cartridges on my printer, and HP provided UPS stickers so I can send the old ones back for "recycling", which is obviously just them filling them back up and selling them to someone else. Let's assume they just made $20 on the deal. That doesn't mean I'd have $20 if I'd kept them, because to me, lacking a factory to refurbish them, they're just garbage. Same goes for all of that personal information Google and Facebook make money on -- you can't go down to the pawn shop and sell your own information, because it only matters in bulk.
As for the personal information, there is an entertaining talk[1] by Pinboard's Maciej Cegłowski about how we can't even know if collected personal information will at some point come back to hurt you or not.
Nothing like transparency to keep me as a loyal customer.
The problem is the definition of "manipulated" is not black and white. Consider the relatively recent change in StubHub's fee model. Consumers will always tell you they hate having a service fee tacked on at the last minute when checking out. So, for a while, StubHub tried all-inclusive pricing on their search results, and consumers said they liked the change. However, it ended up tanking their conversion rates. When they A/B tested putting the fee on last, they immediately saw an increase in shoppers willing to buy.
Is this manipulation? The fee was still clearly stated on checkout, but I imagine the subconscious thought process of shoppers was "well, crap, I've already made it this far, I just want these tickets now."
I agree some of the outright "tricking" examples from the article are over the line, but most forms of marketing are manipulation in some respect. When does "persuasion" become "manipulation"?
Yes.
They presented a meaningless price when in situations where potential customers would compare them with competitors. Only when they can be sure that a customer would likely not check the price anymore (because they think they already know it or because of psychological reasons), they show the actual amount.
That's a very clear case of improving their own situation (better conversion rates) by worsening their customers' situation (higher likelihood of overspending, reduced ability to compare prices) and making them act against their own interests. This would be manipulation by the all of article's criteria.
Of course as a company you can decide do use that strategy - it's not your job to think for your your customers. But don't be surprised if consumer advocacy groups flag you and at some point you find people demanding a law against that practice.
More generally, I'd draw the line to "marketing" if you supply information to the customer that's either meaningless or misleads the customer, but could not possibly help them reach their own goals. (As in this case presenting an amount in the search results that has no practical relevance.)
I'd draw the line to "manipulation" even earlier, if you make a person act in a way that's against their own conscious interests in the moment. As I said, there can be rare cases in which that manipulation could even be wanted by the person, but I don't see that changing the fact that it's manipulation. It's different from "persuasion" where you make the person change their interests.
You're starting with the prior that buying stuff is bad and should be discouraged in general. This is quite arguable.
Nope. Otherwise people should absolutely love my approach of forcing them to buy at gunpoint - perfect conversion rate!
Even if you assume that the data implies customers prefer the fee at the end, you'd also have to research why they do so - and solve the mystery of why your statistic suggests they prefer the fee at the end, yet direct user feedback suggests the opposite.
>It is just as irrational to not buy the product as to buy it in either scenario, since the price is the same.
That kind of rationality requires perfect information. But it is likely that the price appears lower in the fee-at-end-case, because that is the only effect that the change has - so perfect information is not given. As the price is in fact the same as you say, that would make fee-at-end the case with irrational behavior.
>You're starting with the prior that buying stuff is bad and should be discouraged in general. This is quite arguable.
No, my premise is that buying stuff while you don't know the price is bad, which I think is a commonly accepted sentiment.
That instance seems very clearly manipulation. When the cost is known people don't buy, only when they've sunk time on the process do they buy - probably because they have to assume other sites in the field are equally manipulative.
Unless the full cost is available without going through the sales funnel I'd say it's fraud.
Say people come to the store after having seen the advertisement, but when they arrive they decide that after coming here all the way, they don't care to save $100 and they buy the car.
It is conceivable (even probable) here that if I had honestly advertised it as $20000, few would have come. Therefore, making this change would increase my conversion rates on my advertisements. The argument stated by you would mean that my behaviour is defensible, and that I should advertise prices I do not intend to deliver, allowing advertisement conversion rates to be my guide.
However, that's a classic bait-and-switch scam, a kind that is illegal in the US, and I doubt you'd find anyone who'd support it.
In stubhub's case, they are being super-sketchy, and they offer a fig leaf of opting in to showing fees:
http://time.com/money/4018864/stubhub-fees-all-in-ticket-pri...
Trying to do business honestly is hard when your customers are too ignorant and irrational to appreciate honest business.
I'veb been long aware of such trends. Something as simple as a multi-step transaction process, across multiple screens. I literally cannot see what the end-stage requests will be before committing. If I don't bug out immediately, I'll just throw random inputs at the page to see what the hell it does.
But it's another reason why I'll strongly prefer in-store, brick-and-morter purchases, or anything that doesn't have a service contract (the "agree to our new ToS to cancel your account" story below really puts the shine and hurt on this), etc.
Much as the whole microlawsuits thing sounds like an absolute PITA, something which reduces frictions for consumers to sue would be interesting. Respond to failed opt-out tools with a claim for damages and costs....
Sigh.
Amazon UK has an horrid check-out process that omits any clickable bail-out links, should you want to back-up to change something you have to edit the URL bar.
They also only provide a total for all postage, on one page, with no break-out of what applies to which item ( from multiple vendors ).
I assume it's meant to be slick and non-distracting but it feels like a greased laundry chute through which you are sliding...
well, yes, non-distracting from completing checkout using the most-expensive options.
> Whereas Ryanair once placed the insurance opt-out option in a drop-down menu of passengers’ home countries — between Denmark and Finland — it now simply gives passengers the option of adding insurance.
If you use Amazon but don't want Amazon Prime, you get deceptive dialogs shoved in your face at least twice per purchase.
If you try Amazon Prime for FREE, you also don't have to pay for shipping.
To me the entire ad-supported "business model" is the darkest pattern of them all. Please read this comment before a knee-jerk rejection of my claim: https://news.ycombinator.com/item?id=7485773
-
[1] https://www.eff.org/deeplinks/2015/10/internet-companies-con...
In return I've been given free access to a huge amount of free content and free functionality. Plus - very occasionally - the advertising is beneficial rather than merely irrelevant.
Really, her last name is Swindler? That's just funny.
See: http://www.90percentofeverything.com/2014/08/26/some-dark-pa...
Note that this article is from two years ago and at least one of the sites called out is still doing the same thing.
A dark pattern if I ever saw one.