數字化時代的定製廣告很可怕

While I flew to Barcelona last week to speak at a conference, my iPad was at breakfast at a restaurant in Cambridge. That, at least, is what I deduced from the device’s location, transmitted to me after I activated the Find My iPhone app on my mobile phone.

最近，當我飛往巴塞羅那出席一個會議併發言的時候，我的iPad正在劍橋的一家餐廳裏用早餐。至少，這是我根據手機上“查找我的iPhone”應用傳給我的設備位置信息推斷出來的。

I was relieved: the tablet was neither lost nor stolen; it had been accidentally picked up by the organisers of a meeting I had attended the previous day. If, however, another app had found me at the airport and started to badger me with offers, based on my movements, prior purchases and reputation as a loyal or fickle customer, I might have felt a little uneasy.

令我寬慰的是：我的平板電腦既沒丟也沒被偷；我前一天參加的會議的主辦方偶然撿到了它。但是，如果另一款手機應用發現我在機場，並根據我的移動路線、之前的消費記錄以及客戶忠誠度高低等信息頻頻向我推銷，我可能會覺得有些不舒服。

Here is a question companies increasingly need to answer: what is the creepiness quotient of your product, or marketing campaign, and how would you know? The problem is no secret. Public examples abound. They include embarrassing personalised marketing gaffes — encapsulated in the popular, but possibly apocryphal, tale of the retailer Target, which outed a pregnant teenager to her parents by pitching certain products to her — and the more recent suspension of sales of Google Glass, amid queasiness about the device’s potential misuse. “Problem” may even be a misnomer. While Julia Angwin’s recent book Dragnet Nation describes the dark side of surveillance by companies and governments, a new book by Michael Fertik, founder of , which offers ways of enhancing online reputations, sees it as a simple fact of modern life, which we can exploit for advantage.

如今有個問題越來越需要企業來回答：你們的產品或者營銷有多令人恐懼不安？你們又如何知道？這個問題已經不再是祕密。公開的例子比比皆是，包括令人難堪的個性化營銷失誤——一個廣爲流傳但可能不足爲信的故事把這一點體現得淋漓盡致：零售商Target由於給未成年少女推銷相關產品，向其父母暴露了她懷孕的事實；還有最近暫停銷售的谷歌眼鏡(Google Glass)，人們擔憂該設備可能遭到濫用。用“問題”這個詞甚至都可能不恰當。朱莉婭•安格溫(Julia Angwin)的新書《天羅地網》(Dragnet Nation)描寫了企業和政府的監視活動的黑暗面。告訴客戶如何提高網絡信譽的，其創始人邁克爾•費蒂克(Michael Fertik)的新書更提出，監視是現代生活中的一個基本事實，我們可以加以利用。

In The Reputation Economy, he and co-author David Thompson lay out plenty of examples that I find creepy. They include Facedeals, which aimed to combine facial recognition and your Facebook profile to push special offers to you when you arrive at a shop. Another is Moven, a mobile payment app, which originally set out to score customers’ social media credibility alongside traditional credit measures.

在《信譽經濟學》(The Reputation Economy)一書中，費蒂克與合著者戴維•湯普森(David Thompson)闡述了許多在我看來令人毛骨悚然的案例。比如Facedeals，旨在將面部識別和Facebook上的個人資料結合起來，這樣當你去商店的時候，可以向你推送特別優惠活動。再如手機支付應用Moven，該應用最初打算給用戶的社交媒體可信度評分，與傳統信用衡量標準一同作爲參考。

“Future legal cases will have to decide at what point digital stalking gets just too creepy,” Mr Fertik and Mr Thompson write. They recommend, instead, that you publicise recent job promotions on social networks, tweet about your forthcoming purchases (“Looking for new SUV, considering @BMWUSA or @MBUSA, any experiences?”), and reconcile with bitter ex-partners who have badmouthed you online — all in the interests of making algorithms think you are a successful, luxury-car-loving, perfect date.

“將來的法律訴訟案將不得不斷定，數字化追蹤在什麼程度上會變得過於可怕，”費蒂克和湯普森在書中寫道。然而，他倆仍推薦人們在社交網絡上公開自己最近的工作晉升，爲你想購買的東西發條tweet（“想買輛新SUV，正猶豫是買@BMWUSA還是@MBUSA，有什麼建議麼？”），與在網上說你壞話、與你怨恨頗深的前任和解——這一切都是爲了讓算法認定你是一個事業有成、熱愛豪車的完美約會對象。

Research used to show personalised marketing was persuasive and well received. But Lisa Barnard, who once worked in advertising and is now assistant professor at Ithaca College, ran some experiments aimed at identifying the creepiness quotient (she calls it the “creepiness factor”) in ad campaigns. Tailoring online advertising to individual behaviour still works, she found, but “perceived creepiness” makes customers 5 per cent less likely to make the purchase. That is 5 per cent of the budget that could be spent elsewhere, if a campaign’s CQ could be cut to zero.

過去的研究常常認爲，個性化營銷既有說服力又容易被接受。但是曾在廣告界任職，現任伊薩卡學院(Ithaca College)副教授的莉薩•巴納德(Lisa Barnard)進行了幾項試驗，旨在確定廣告宣傳的可怕程度（她稱之爲“可怕因子”）。她發現根據個人行爲進行定製的網絡廣告依然有效，但是“覺得可怕”會讓顧客購買的機率降低5%。這意味着，如果廣告的可怕程度可以降低到零，就可以將消費者可能花到別處的5%消費預算賺回來。

Even pioneers recognise personalisation has its limits. Facedeals has become Taonii, an app which still offers tailored deals, without face recognition. “Consumers were just not quite ready,” a spokeswoman said via email. “They wanted the benefits but in a slightly friendlier [way].”

甚至那些最先試水的商家也意識到個性化營銷的侷限性。Facedeals現在變成了Taonii，該應用依然提供定製推送服務，但是去除了面部識別功能。“消費者還沒怎麼準備好，”一位女發言人在電子郵件中說，“他們想要優惠，但是要以一種更友好的（方式）。”

Keith Weed, chief marketing officer of Unilever, the consumer products company, says digital personalised marketing is “a bit like when you to go to your local shop and they know you and perhaps even have what you want waiting for you”. But cosy as that sounds, he concedes that getting the online and mobile version right is “a fine balance”. For now, giving customers an easy opt-out and ensuring they know what will be shared, where and with whom, are the keys to not creeping them out, he says.

消費品企業聯合利華(Unilever)的首席營銷官基思•威德(Keith Weed)說，數字化個性營銷“有點像你到當地的商店買東西，店家認識你並且可能還準備好了你要買的東西”。雖然那聽起來很暖心，但他也承認，在網絡和移動端做好這一點，是很難把握的“微妙平衡”。他說，就目前而言，讓顧客可以方便地退出，並確保他們瞭解哪些信息將被共享、將在何處共享以及和誰共享，纔是不把顧客嚇跑的關鍵。

Going back to William Lever, Unilever’s founder and early adopter of persuasive advertising, marketing has a history of constant experimentation, in which you and I are the guinea-pigs. Rapid evolution is inevitable, because the line between creepy and friendly is always shifting. A user may willingly give up information for one purpose, only to react with disgust when it is used for another. But companies owe it to their customers to come up with a better way of defining their creepiness quotient. Otherwise, deciding where “cool” becomes “eeugh” will continue to be a matter of trial and uncomfortable error.

回溯到威廉•利弗(William Lever)——聯合利華的創始人和勸說性廣告的早期採用者，營銷的歷史就是持續不斷的試驗，而你我就是其中的小白鼠。迅猛的轉變是不可避免的，因爲可怕和可愛之間的界線總是搖擺不定。用戶可能情願爲了某個目的共享信息，而在信息被挪作他用時卻滿心厭惡。但是，企業要想找出一種更好的辦法來確定廣告的“可怕程度”，還需要求助於它們的客戶。否則，確定“爽”何時變成“不爽”依然將是一個不斷試驗和令人難堪的錯誤反覆出現的過程。