You're a spammer essentially. Using an automated bot to fake a human interaction (favorite) in order to get my attention and get me to follow you. It doesn't matter that it's more effective than "Twitter Ads". Twitter Ads are marked as Ads, your favorites are not. I really do hope Twitter shuts you down, as well as those follow bots that follow me 10 times a day and unfollow me the next day.
Sadly, Twitter doesn't do much to combat spam so they likely will not. But don't fool yourself into thinking you are going to partner or be acquired by Twitter.
Rather ingenious, if I do say so myself. The favorite interaction has become increasingly important over the past year or so... I know friends (non-technical friends) who joined twitter recently who would rather get a favorite than a retweet, as a favorite feels more personal. This opposed to a retweet, which seem to be given out much more frequently than a year ago. It would also be pretty cool to be favorited by a recognizable name. Now that I think about it... could be a pretty useful recruiting tool, if targeted properly.
It's also leveraging the fact that twitter (for now) emails the original user to tell them their tweet was favorited. So it's essentially spamming by proxy.
(Also, we've seen this before. It's just follower spam.)
I wasn't aware of the emails, I haven't received one and I've been favorited before.
The whole point here is that it's a mis-use of a messaging channel - they are leveraging the fact that users _don't_ expect favoriting to be a form of advertising, they expect it to be a message from an actual human being that has favorited their post because they like it. (Similar to the problems with follower spam.)
(I'd also point out that ads on web sites at least help subsidize user's free usage of those web sites, so there is an indirect benefit to the user that isn't there in the case of spam.)
You may have your Twitter notices already turned off - I get emails from them when someone favorites or replies to my tweets, as well as notifications from their app.
Second, the advertising on Gmail (or Yahoo mail etc) isn't an indirect benefit, it's an exceptionally direct benefit. I've gotten to use a truly great, fast, reliable email service - Gmail - for a decade, with gigabytes of storage, and all they've done is run ads on the right side that I've never once clicked on and they never get in my way. Talk about a helluva deal. You're completely correct in the fact that Gmail ads are nothing like spam.
Sometimes, if I really like a tweet, I'll favourite it because I have no other way to say +1 (other than replying to it). But I must like it enough to want to sacrifice the integrity of my bookmarks.
Based on what I'm reading here, it seems that people have wildly varying ways in which they use favouriting.
Twitter would look horrible by partnering with an automated service of any sort.
There are countless bots out there already abusing this idea with followers, and pushing it into favourites isn't making it any better. All it does is add pointless noise to the twitter network and unethically take advantage of the fact that they can't catch all of the spammers at once.
Of course you can make money when you violate the rules, but that doesn't mean doing things like that are an ingenious or stable business model.
Ultimately I want access to the Twitter Ads API, at which point I'll switch from favorites to serving promoted tweets/accounts. I think I've innovated on the Ads UX and the form which the impression takes is secondary. This is a surrogate impression. It has allowed me to build an Ads product and an Ads customer base in anticipation of Ads API access.
What you've done is move ads from the place where they're supposed to be (promoted tweets) to the place where they're not supposed to be (your interactions). Innovative spam is still spam.
There are versions of this approach all over Instagram too - I 'll wake up to a large number of favorites, only to find out that most of them were bots. (Although, with the OP's approach, the accounts that are favoriting this are real accounts, so I wouldn't be able to tell easily by going to their profile)
If I get a favorites notification, I want to go go their profile and find out if they share similar tastes. If it's just an automated system, they have ruined this for me by lowering the signal-to-noise ratio.
Regardless of how innovative this is, Twitter shouldn't get any flak if they decide to shut him down.
The distinction for me isn't between marketing/non-marketing but between humans and robots. If a human representing some business favorites my tweet because they feel that it's worth favoriting, that's great. In fact, it makes me feel good that some human at an organization is reading and caring about what I have to say. Wonderful, I write things so people can read them.
Robots just going out and marking things as favorites just to get attention and followers? That's the kind of spam actively undermines the value of the example above.
Yes, your customer list may be useful to Twitter - as a list of accounts to suspend for violating their ToS, devaluing favorites, and hurting any legitimacy Twitter had.
You shouldn't have existed (or been allowed to) in the fist place.
The meat of the story starts with this sentence:
> One of the peculiarities of my development environment was that I ran all my code against the production database....
Of course something is going to go wrong. IIRC this was already discussed on HN.
His point is probably that professionals don't do that. It's not the companies fault he was working directly on the production database.
If you follow someone without looking into their profile and deciding if you are interested in their content then you get what you deserve.
Note that you can substitute all sorts of adjectives for "dumb".
E.g., "those nerds are so unpopular they deserve to get beaten up". Or, "those people are poor enough that they deserve to live next to a toxic waste dump".
Basically, it's a way of taking whatever characteristic you pride yourself on and then making yourself feel even better about it by enjoying when shitty things happen to other people.
No, the "take responsibility for the outcome of your own actions" school of thought. Or do you think one should be able to wander through life, doing and saying anything one pleases, completely oblivious to your environment, and not expect bad things to happen sometimes?
A woman might choose to wear clothes that make her look sexy. But that doesn't justify a guy raping her.
I agree that people act dumb. But that doesn't justify predatory behaviour.
Nobody said it did, and that example is nothing like what we're talking about here.
This is also not something that is just happening to these people. They are taking an active part by deciding to follow someone.
I certainly don't belong to the school of thought that says you should never experience a negative outcome.
The key difference being that in the prior example the negative impact is you follow someone that may not be of interest to you, in which case you can always un-follow. While a nerd or the poor will find they can't literally click a button to undo the "damage" that was done.
I get what you are trying to say, I disagree that you can apply that to this specific example. It would be better aimed at a rich person who literally thinks everyone on food stamps is a lazy bum.
So, spam works. It still doesn't make it morally right to trick people into following you.
Favoriting a tweet has no connection with "Hey, I tweet about stuff that you are interested in so you should follow me." It's not even a request for someone to follow you.
I think this is incorrect. The intended purpose of "Favorites" is to communicate that a living, breathing person (as opposed to an "account," as you describe) saw my tweet, enjoyed it, and liked it enough to Favorite it.
The OP's product fraudulently claims to Twitter users that the customers of this product are engaging in that action.
edit: The main point being that the intended purpose is to communicate that an actual person actively read the tweet, and for whatever their reasons, actively decided to favorite the tweet.
Please don't use that rationale. Lots of people never get what they deserve, and pretty much nobody deserves the shit some people have to go through.
Either way, there are some people who don't just follow back, so they don't deserve the spam even by your definition. Most people here can automate follows, favorites and similar interactions, but don't. Some of us would like to keep those meaningful.
If a service simply pulled a list of tweets from users that you might be interested in there's nothing wrong with that, correct? It's just curating a list of tweets for you. There's also nothing stopping you from favoriting them all to get the most attention from those users.
But when the human act is entirely faked it's spam. And I agree. But it's also a pretty slight distinction we're making since humans can easily choose to behave like bots.
1. shut you down
2. offer you a job
3. give you nothing
4. see no benefit in letting you use their API. But depends on whether your idea for it adds value.
1. is innovative
2. probably didn't expect it to work so well
3. realizes as soon as they notice, he is toast
4. is trying to offer an olive branch
5. is likely doing some good, despite the deception
My advice to OP:
1. Use your innovation to subvert something that harms the public good, thereby making the world a better place. If you are really clever, you will figure out how to do this and make money too.
2. Keep thinking about this problem, and how you can do it in a honest way, which won't lead to abuse or dilute the value of the platform (any platform, not just Twitter).
"I publicly acknowledge I'm wrong. I'm letting Twitter know they should shut me down, but I'm just going to keep riding while the money is hot. Everyone will think I'm a great guy because of this post and my reputation will be salvaged after I'm shut down."
Spammers are the lowest. If the OP has any shred of dignity left he would shut down before Twitter can do that for him.
Given these facts, I'm not sure how many Twitter users actually care if the favorites they receive are genuine or not. The reality is that few users on Twitter are real anyway. We once paid thousands of dollars for a tweet to go out to millions of followers of a celebrity. The cost wound up being about $207 per collected email address for a free newsletter that was highly targeted to that celebrity's followers. We got tons of hits to the link, but approximately 94% of the traffic to the link posted was bot traffic. This was a celebrity that the ad network we used to purchase the tweet had rated as having extremely high follower engagement metrics - apparently they didn't bother checking user agents when determining this. If you want actual users listening to you, don't go to Twitter.
Yes, but that's exactly the reason why OP's behavior can be labeled as spamming. You "theoretically" adds to the user's credibility, but by automated means, so the added credibility is actually artificial, which then diminish the value of "credibility by the number of favorites" for everyone.
Um, no they won't. Things you favorite don't show up to others unless they specifically go and look at things you have favorited.
Their actually doing a pretty good job for the amount of spam they actually have to deal with. Once they figure out that people have started to use favorites to spam, they'll start killing those kind of spammers too.
Agreed, and speaking personally, I would love to pay Twitter $XXX/month, especially if it included some form of "we'll reply to X emails within an hour" support.
The fact that they're not even charging $20/month for minimal API access is baffling to me. I'd expect it'd only help reduce spam as well!
The devs who experiment and make something cool are often worth more than the minimal charge.
However I would agree that rather than having arbitrary limits using devs as a revenue stream makes sense (give everyone a portion of what you currently allow and allow payed upgrades)
Exactly - there are plenty of options available.
I'm just wondering how many such developers would be put off by such a relatively small charge?
Obviously, the situation now is very different to when Twitter was still building momentum.
I'm just confused that they don't come out and say "From the 1st of May, we will be charging all new API accounts for access."
It's like (snail) mailing someone a bill for $0.25.
No one in their right mind would want to support the Twitter or Facebook API's...
This could be done sensibly though: limit the number of emails per month; or only include support in higher-priced plans.
eBay's API can be pretty tricky at times, and if you want any help that doesn't involve using their forums, it costs $75/hour:
I think employees would be happy to support an API, if it is a revenue producing product.
There are people making millions using these APIs now with no support. Why would charging them $25/month suddenly require support?
Any product or service can be used for evil by evil people.
I've used Myles' service. It works amazingly well. So well, in fact, we looked at using his API to integrate the functionality into Socialyzer. We ended up not coming to terms and we built some scripts to do this for ourselves internally.
We use this functionality to build a very targeted following of people who may find our service (http://www.rewardrkt.com) useful . Our conversation rates from favorite, to follow, to registration on our site are extremely high. The people we auto-favorite are finding value in what we offer.
Favoriting their tweets is an efficient way to "introduce" yourself to someone without being super pushy.
After Twitter shutting down Flattr I reached out to Myles again to see if he'd seen the news. Of course he had, and I see his blog post as a bold move. Why wait around for doomsday if you can either (1) make it go away or (2) bring it about faster and move on to the next thing afterward.
In fact, they are gaining value out of it.
Who's spamming and who is not?
I would argue that the way we're doing this is not spam, as it is clear the people we favorite are getting value by virtue of the fact that they are following us, registering for our service, etc.
What percentage of people who you auto-favorite follow you back and/or register for your service? What threshold do you consider appropriate.
Do you have any visibility into the people who are annoyed by seeing your robot interaction in a place that they feel should be reserved for human interactions?
How would you feel if every single service provider interested in finding you were to favorite all of your tweets to get your attention?
It is greater than 50%. As I said, we do this in an extremely targeted way. We may only favorite 10 tweets in a day because the targeting we do is so specific. To be honest, I haven't thought about any particular threshold being appropriate/not.
> Do you have any visibility into the people who are annoyed by seeing your robot interaction in a place that they feel should be reserved for human interactions?
No, but I think you already knew that. This question sounds like your projecting your personal feelings quite a bit.
> How would you feel if every single service provider interested in finding you were to favorite all of your tweets to get your attention?
Personally, I wouldn't even notice it. I have all notifications turned off for that type of stuff. If anything, they'd be giving me more exposure to the people who may look at their account/favorites.
"How would you feel if every single service provider you were interested in finding were to favorite all of your tweets to get your attention."
I'm not suggesting that the technology is there yet, but the future is a big place.
Anyone I'm interested in finding I'll go searching for. They don't need to inject themselves into my interactions list just because they might trick me into getting my attention.
Do you want to get my attention? Great. There are plenty of well understood and well established ways of doing that. Place an ad in my favorite magazine, sponsor a podcast, create something so great that people can't resist telling me about it. Just don't spam me and pretend you're doing me a favor.
We analyze your social audience (location, activity, tweet content, etc) and use that information to predict when and where you should engage with them.
We don't take it to any automated action other than publishing your content, but it's possible to do.
Can you give an indication how high? Greater than 1%? 10%?
Or you think it's just a super clever (and cheap) way to get around actually paying Twitter to promote your posts?
How effective it is doesn't enter into the equation really.
If people are upset/annoyed that is was automated.. fine.. do it manually.. the fact remains that it is a very effective method to build your audience in a targeted fashion.
Maybe Myles should just "pivot" to compete with Little Bird and focus on just surfacing the data on who you SHOULD be engaging with.
The purpose of this blog post is to save what little of his reputation is left by making it seem like he's redeeming himself. He isn't. He's waiting for Twitter to shut him down while he keeps making money as a spammer.
I say a slow clap is in order here, no matter what his true intentions are, only if for being so focused on what he wants from Twitter Inc.
There goes $50/month/customer, for someone with supposedly hundreds of customers. He probably could have gotten at least a couple months more without being shut down and some free publicity on Hacker News afterwards complaining about being shut down, but he traded that for a maybe 1% chance at some kind of introduction.
And, I guess, a HN post in a couple hours complaining about being ingloriously shuttered.
I'm making a big presumption about his intelligence though.
On the flip side, maybe instead of shutting you down, they should acquire you and then shut you down. That's a win-win, right?
To clarify: I don't mean to disparage him, but acquisitions have to bring some sort of value to the acquirer.
These aren't ads. These are "person abc favorited your tweet" notifications that the mobile twitter client pops up, and twitter emails you. If that gets overrun by spam (a) people will turn them off and (b) favorites will be useless as a mechanism for displaying "top tweets" like they do.
Even in the case of this all being true, the real problem is the ability of in-house people to ignore reality and refuse to favor a not-made-here product. Always difficult to tell which way the wind will blow on that. Companies make really stupid decisions all the time.
Make it $1 for every follower, straight-up no ads or nothing. For $100,000 I can get 100,000 more real human followers on my account. Advertisers would love that better than promoted tweets or bot favorites.
But is that good for the users?
Presumably Twitter achieve this based on stats like who I follow (including who i've followed recently) and i'd guess search activity etc. but hey, works for me.
My account has been hosed for two weeks, I log in and just get an error message -- once I log in I can't even view the help page. And of course, there's no option on their forms for the problem I'm actually having. I've been thinking about sending a fax to their number for law enforcement queries since somebody might at least see it before they throw it away.
I'm certainly not getting a message saying that I am, but I could be hellbanned in some special way.
It definitely makes me question if Twitter is going to be part of my life and the web sites I make in the future though.
One of my friends who applied, was told he had to email the recruiter daily in order to expect a response. And then ended up being called in the middle of class, because the recruiter never bothered to tell him he had scheduled an interview for him.
1. starting the post with 5 paragraphs and a list of all the virtues of Twitter's engineering
2. spin the fact that, after a simple DB screw-up in production it took them about 6 hours to decide to just rollback from a backup
3. end the "postmortem" (which is, by the way, completely devoid of apology) by trying to hire
all the while talking about how working at Twitter is like building a rocket mid-flight (so, just like every other company)?
It would however cause users to start ignoring Twitter messages/notifications. People are already ignoring responses from people they don't know, and it can get even worse if we think most people interacting with us are just bots.
On the contrary - it does. Because the point of "favoriting" is that when my tweet is favorited, I expect that the other party liked it. Now that's not true anymore, since it could have just been a bot who performed that action based on keyword-targeting me, not by judging that the tweet has a value to be favorited.
For me, it just sounds like this project is merely a throwaway project and the aim is to get banned, at least then to get onto the Ads API where they may have another concept, but couldn't get on it any other way.
Might be me, might just be extra dubious, but it doesn't look right to me.
My guess is he's really looking for a job. Or testing the hack-ability of ycombinator (and non-researched news aggregators in general).
Is there anything we can do about this, or is it just a product of building on someone else's platform? Obviously, this has happened throughout history, but on a much smaller scale. I can't recall something as prevalent and simply integral to most people's everyday lives as Twitter/Facebook/the App Store where this sort of heavy-handedness could happen. Even when Windows was in its prime, people could pretty much build anything they wanted for it (for better or for worse).
If anything, we might be less susceptible to it now. It would make a fascinating study, anyway.
My opinion, is that ads suck, most of them do (except the ones that make you laugh and remain classics which in that case fall into another category, and as far as I'm concerned that only happened with TV commercials and a very few "internet" ones) ... and being spammed even though it's becoming a habbit with all the interwebs bullshit we've been and have to digest everyday is more than annoying.
You seem to be smart and you should definetly put your talent in something else.
Under US law, the 9th circuit held in a criminal case that an employee who violates employer computer access policies with intent to defraud and by that action furthers the intended fraud and obtains something of value. http://law.justia.com/cases/federal/appellate-courts/ca9/10-...
The 1st Circuit held in a civil case found grounds to uphold an injunction against someone redistributing data obtained from a scraper on the grounds that scraping public data in the circumstances exceeded authorised access, and that the data was something of value that would harm the target website. http://law.justia.com/cases/federal/appellate-courts/F3/318/...
I don't think these laws are a good thing, but the reality is that US courts have found that accessing websites for purposes beyond the terms of services with 'intent to defraud' violates the Computer Fraud and Abuse Act.
Defrauding could potentially mean misleading people into following you on Twitter if having Twitter followers is valuable for marketing.
The OP's post could be introduced as evidence by Twitter that he intended to exceed authorised access, that he had intent to defraud, and that he knew his actions were harming Twitter. I think the OP would be well advised to take legal advice on whether to shut down the site now and / or retract the blog post (although it might be too late to suppress it now, given how many people have seen it and possibly archived it).
I'm not sure if I picked bad targets, or if more technical users are less likely to "follow-back". But, those were my results, do with them what you will!
edit: I'll add there were quite a few "misfires" that led me to favorite some rather strange tweets - particularly #ruby which I favorited several tweets about people's young children or pets named Ruby.
You knew Twitter's TOS & you broke them. That's okay if you're willing to put up with the business risk that--if you gain scale--Twitter may be forced to buy you (eBay/Paypal, Photobucket/Myspace) or at least work with you.
That said, your service as others acknowledge below is akin to spamming. This is not a service that is for the betterment of Twitter users AND advertisers. It merely betters your company's coiffeurs and advertisers who are getting "a deal."
You want to win Twitter over? Make it painfully obvious to customers what is occurring. Disclose your business metrics. Put out case studies that show how both users AND advertisers benefit.
The appeal for public support/disclosure is old hat.
- Why does the application have to be web based? Why don't you build a desktop application where the user has to use their own API key? This will also help bypass the IP limit.
- Aren't you already breaking their terms? 
"Using [third party applications to “Get More Followers Fast!”] is not allowed according to the Twitter Rules."
- How many favorites can an account do every day? Is it unlimited?
If you don't, then please continue spamming. I think that would make you a pretty big a-hole, but it's your life. shrugs
One of the ways to start thwarting this spammy behavior is counter-network of twitter accounts, randomly inter-followed for further legitimacy.
Once a follower is identified as a spam, it's being mass followed. It would skew conversion rates and perhaps force clients to pay for fake followers, which they won't appreciate.
The cost is lower, but can someone do studies on the lifetime REVENUE of a real, targeted follower? It might all be meaningless if that number is 0.01 cents.
This isn't exclusive to Twitter. Not every person is worth the same value to advertisers. I advertise on platforms ranging between $0.20/click to $1.50/click, and different platforms offer different value per click.
This part won't make them any happier:
>> I started renting a large cluster of virtual machines to scale the service while staying within Twitter's 'per IP address' rate limits.
Quickly I got sick of the number of favourites it did - it was really spammy. I am not a fan of this model.
And the result is http://imgur.com/pXVWaue
Hope the author finds this and fixes it.
An option that is probably more likely to be API TOS friendly is to show the user relevant tweets and suggest that they favorite, rather than auto-favoriting them.
If anyone deserves to succeed, it's you man. I hope twitter comes around for you.
There are good reasons for this, but it it's still a generalization. Like all generalizations, it isn't always true. It might be the case that this service actually provides value to both parties: the person with the good or service initiating the activity and the person who is made aware of it. This is the definition of a win-win interaction. The only real way to determine the quality of this service is to try to measure whether value is being created on both sides.
^^^^ this COULD MAYBE happen. I still don't want people to stab me with a sword.
I could MAYBE get spam that has something I want to buy. I still don't want to get spam.
When someone fav's one of my tweets, I get a notification. I like this because it tells me that someone liked something I said. It makes me happy. This reduces the value of these notifications and, if it becomes widespread, will probably make me turn off notifications, which will make me sad. Don't you see how this is producing negative value?