(单词翻译:单击)
This is an ad for the 2018 Camry that Toyota published on Twitter.
这是丰田在推特上发布的凯美瑞2018款的广告。
They also published this one. And this one.
他们还发布了这个版本，以及这个版本。
For this one campaign, Toyota released 83 different versions of the same ad,
就这一支广告，丰田就发布了83个不同的版本，
and every version targeted different users --
而且，每个版本锁定的都是不同的用户——
not based on their gender or their age, their political affiliation or their location.
这种不同不是性别或年龄的不同，也不是政治派别或所在地区的不同。
The ads targeted users' emotional states through their emojis.
这些广告针对的是表情符号透露出来的用户的情感状态。
A targeted ad is where a company shows their ads to only certain kinds of people,
精准投放的广告指的是企业仅对特定类型的用户，
certain people who are more likely to buy their products or like their message.
也即最有可能掏钱或者最有可能喜欢广告传递的信息的那些人群，展示他们的广告。
That's why as someone who creates videos just like this one, I see ads for Adobe's video-making products in my Facebook feed.
这就是制作了当前这个视频的我，为什么会在我的Facebook feed里看到Adobe的视频制作产品的广告的原因。
But in 2016, Twitter began giving advertisers access to emoji data like who is posting what and when and which emojis are the most popular.
然而，2016年，推特开始为广告商们提供用户的表情符号数据了，比如谁在用什么表情，什么时候用的以及哪些表情最受欢迎等。
That is totally unique compared to advertising before this.
这种广告思路和之前的广告思路完全不一样。
Emojis have an emotional context paired with them
表情符号都有与之相匹配的情感背景，
and that lets advertisers better gauge the feelings expressed in people's tweets.
这便让广告商们能够更好地把握人们在推特上表达的情感。
With emoji targeting, every highly tailored ad would be triggered by the emojis a user would post, in real time.
通过“表情锁定”，每支具有高度针对性的广告都会在用户实时发出表情的那一刻定位到用户。
Tweet a pizza emoji and Domino's would reply with a coupon.
用户在推特上发一个披萨表情，达美乐就会回复用户一张优惠券。
Tweet any emoji at Google and get a handy link for the top search results on their platform.
在推特上发任何表情的同时@谷歌，就可以方便地获取该表情在谷歌上的最佳搜索结果。
Tweeted a heart eye emoji today?
今天在推特上发了个“心心眼”的表情？
Well, Toyota might determine that you're feeling positive and serve you this ad while you're in that feel good mood.
丰田就可能认定你今天心情很好，然后趁着你心情还不错把这个版本的广告发给你。
Some emojis are pretty obvious right.
有一部分的表情符号的含义是很明显的对吧。
Smiley face, I'm happy.
笑脸的表情说明我很高兴。
Frowny face, I'm sad.
愁眉苦脸的表情说明我很难过。
But you know there's a bunch of emoji which are much more --
但还有一些更加……
the line's much more fine between what that person is actually feeling or thinking at the time.
就是你用的这个表情和你当下的心情或者你当下的所思所想之间的关系更加微妙的表情。
For those emojis that express more ambiguous emotions,
对于这类表达的情绪更模糊的表情符号，
advertisers can use artificial intelligence to predict if an emoji is used in a positive negative or neutral context.
广告商们可以使用人工智能来预测表情的上下文是积极的，消极的还是佛系的。
Let's look at Toyota again.
我们还是以丰田为例。
In January of 2017, Donald Trump tweeted a major criticism of the company for planning to build a plant in Mexico.
2017年1月，唐纳德·特朗普在推特上对该公司计划在墨西哥建厂提出了严厉的批评。
After that tweet was posted, the number of social media posts about the automaker spiked.
在这条推文发布后，社交媒体上关于这家汽车制造商的帖子数量暴涨。
But if you look at this chart, you can see how people felt about Toyota, not just how much they talked about it.
但通过这张图表，除了能看出人们对丰田讨论得有多热烈，我们还可以看出人们对丰田的态度。
Right after Trump's comments, the percent of negative posts spiked when compared to positive posts about the company.
就在特朗普发表评论后不久，相比正面评价，对这家公司的负面评价的比例就立刻出现了飙升。
For an advertiser, knowing how people are feeling is immensely valuable,
对广告商来说，了解人们的感受是无比有价值的，
and they can target consumers with positive feelings and avoid those with more negative ones.
这样他们就能在锁定那些对品牌持正面态度的消费者的同时，避开那些态度更为消极的消费者。
Emojis are just one more tool for advertisers to assess people's emotions.
表情符号只是广告商们评估人们的情绪的又一个工具。
The idea is that if the advertisers are using it effectively, we are gonna see more relevant ads.
这种广告思路的理念是广告商们要是能够将这一工具充分地利用起来，我们就能看到更多与我们的关联性更强的广告。
But regardless of how relevant those ads are, the process is never going to be fully transparent.
但无论这些广告跟用户的相关性有多大，整个广告的制作过程是绝对不会完全透明的。
As a consumer it's difficult, if not almost impossible, to tell what information a marketer is using to target you.
作为消费者，你很难，甚至几乎不可能，知道一个营销人员是用什么信息来锁定你的。
Most advertisers argue that tracking the emojis you use is no different than tracking the keywords you use on Google,
大多数广告商都声称，追踪用户使用的表情符号和追踪用户谷歌上的搜索关键词并没有什么区别。
because you volunteered that information publicly.
因为你是自愿公开这些信息的。
You shared that data freely with a free website that is ad supported,
你既然在一个由广告撑起来的免费网站上免费分享了你的那些数据，
you should be able to understand that the same type of thing is going to happen on a social media platform.
你就应该能够理解在社交媒体平台上也会发生同样的事情。
But consumer advocacy groups disagree.
但是消费者维权组织并不赞同这种看法。
They argue that advertising to people based on a psychological profile of their emotions is intrusive.
他们认为，照着通过剖析用户情绪得出的心理画像向人们推销广告侵犯了用户的隐私。
About half of Americans share a similar skepticism, many of whom aren't confident that social media sites actually protect their data.
大约有一半的美国人都持类似的怀疑态度，他们中的许多人对社交媒体网站是否真的能保护他们的数据都没有信心。
And emojis are part of that data.
表情符号就属于这样的数据。
For all the privacy concerns, emoji advertising is still in its infancy
尽管有上述这些隐私方面的考虑，表情符号型广告目前还处于起步阶段，
and though it only exists right now on Twitter,
尽管这种类型的广告目前还只有推特上有，
it wouldn't be a far leap to see multiple platforms offer a similar service in the future.
但我们距离多个平台提供类似服务的时间并不会太远。
And in the best case, we may get ads that give you immediate and valuable information.
而且，最好的结果是，我们或许能够获取到既及时，又有价值的信息。
Maybe you'll post an eggplant emoji and Durex will send you a condom emoji with 10 percent off your next purchase.
说不定那时候你发一个茄子的表情，杜蕾斯就回复你一个TT的表情，还在你下次购买他们的TT的时候给你个九折的优惠价噢。
Or if you don't like your emoji data being used, maybe you just don't use emojis.
或者……你要是不喜欢自己的表情数据被他人利用的话，或许你可以直接不用。
Yeah, right.
嗯，没错。
