The New Social: Brands Are Embracing Private Sharing
Source: Joanna Goodman
“Just because you’re paranoid doesn’t mean they aren’t after you,” wrote Joseph Heller in the novel Catch 22. But it is doubtful he envisaged that algorithms used by search engines, social media platforms and websites would be after you to ‘personalize your experience’. Consequently, perhaps, personalization and privacy have taken a new direction with the shift to messaging as the go-to communication platform.
Search and social media platforms derive their income from advertising, so it makes sense for them to tailor what users see. Facebook’s presentation at the AI Summit in San Francisco explained how artificial intelligence (AI) is used to personalize the pages of each of its 1.7 billion users.
How accurate -- or intrusive -- this is will be determined by our online activity. Connected devices and location data add context; Google’s DoubleClick (which develops internet ad serving systems) take this further by combining personally identifiable information from Gmail with browsing data.
Slaves to the Algorithm
We have little or no control over the multiplicity of algorithms using our data, particularly as more devices become connected. Gartner predicts that by 2021, 20% of all activities an individual engages in will involve at least one of the top seven digital giants -- defined by revenue and market capitalization as Google, Apple, Facebook, Amazon, Baidu, Alibaba and Tencent.
Algorithms are skewing our perceptions and narrowing our perspective. Tailoring content to user preferences potentially compromises online research. Search engines and ad networks are fingerprinting devices in real time, so it is impossible to know how neutral your search results are. And because search results are based on browsing history, we are less likely to see something new. Added to this is the platforms’ own bias. The past few weeks have seen reports of Facebook’s selective news curation and Google’s perceived political bias towards Hillary Clinton’s presidential candidacy.
Josh Sutton, global head of data and AI at Publicis Sapient, believes that transparency is the key as it enables people to make a well-informed judgement call. “Trust is enabled by transparency” he explains. “There has been an implicit bias in our communications since the invention of the printing press, and we accepted this because we could assess the people who were creating and curating content. I believe that a similar opportunity must be provided in the age of machine learning by enabling people to understand the sources of the algorithms and the training data that powers them.”
But we cannot blame the personalization phenomenon entirely on algorithms. It is also a conscious choice. As social media has matured -- and perhaps peaked -- we have become adept at blocking what we don’t want to see -- or don’t agree with. Elections and referendums have seen social media users divide into tribes rather than engage in political debate.
A New Kind of Privacy
While voice is the obvious interface for human-machine conversation, with chatbots on many websites and virtual assistants on multiple devices -- human-to-human conversation has shifted to messaging platforms. Snapchat, WhatsApp, Facebook Messenger etc. allow us to have rapid, private interactions with selected parties.
This is a new kind of privacy because we don’t know the extent to which the platform can read or understand our messages. WhatsApp, a secure messaging system with end-to-end encryption, still includes in its terms the option of sharing our account information with Facebook to improve ads and product experiences.
In 2015 mobile messaging platforms overtook social networking platforms in terms of monthly active users (MAU). This has not been lost on investors -- or advertisers. eMarketer forecasts that Snapchat will generate $367m (£300m) in advertising revenue this year, rising 155% to $936m in 2017.
A large part of the appeal of messaging is that it’s “friction free” -- WhatsApp’s brand promise is “no ads, no games and no gimmicks”. So how do brands join the conversation?
“The responsibility of brands is changing,” says Sutton. “Historically, the goal of advertising was to make people aware of a product and its merits. Today that is shifting to providing people with hyper-personalised information about products that might be useful to them at a specific time.”
Brands Join the Conversation
The iMessage App Store is a case in point, offering apps and extensions which perform actions within a conversation.
Jason Alan Snyder, CTO at Momentum Worldwide, underlines the experiential nature of bringing brands to messaging. “Machine intelligence makes conversational commerce possible even in a ‘private’ conversation,” he says. “It is all about timing, so it makes sense to use data around location, weather and sentiment to determine the right opportunity to engage.” Crucially, personalization must be subtle and appropriate. “For example, it is not awkward to use an iMessage app to call an Uber or book a restaurant within the context of a conversation.” And messages increasingly include emojis and stickers, which represent a massive opportunity for brands.
“Messaging is peer to peer,” observes Vivian Rosenthal, founder of Snaps, which creates chatbots, emojis, GIFs and stickers for brands. “There’s an immediacy that just isn’t there in social.”
| }
|