Facebook founder Mark Zuckerberg.

Now is the time for Facebook founder Mark Zuckerberg and his board to reinvent, not replicate. Photo: Bloomberg

The unrelenting negative reports about Facebook are enough to rattle even the true believers among us.

Since the initial public offering in May, the beleaguered chief executive officer and founder, Mark Zuckerberg, hasn't found much to "like". The company's stock is down by more than 50 per cent, an influential board member has unloaded shares, and Silicon Valley analysts are openly questioning the management.

Facebook needs to push into real advertising - which involves persuasion, emotion and intuition - with the added ingredient of deep technology-based customisation. 

Yet it would be premature to despair over the company's long-term viability and health. Facebook is still potentially the most valuable information-based business. It had 955 million monthly active users at the end of June, and benefits from a tremendous amount of attention.

More than any other company, Facebook is positioned to radically expand digital marketing, moving us from today's narrow, intent-based approach to a broader and more familiar persuasion-based model. For that to happen, it needs to step out of Google's shadow and embrace a strategy that plays to its strengths and unique position at the centre of today's internet. It must reinvent rather than replicate.

Google, as we know, came of age when our connection with the web was largely individual. We searched for information, read news, bought a few products and traded stocks. Today's internet is a vast digital space in which we live with our friends, family, colleagues and multitudes of slightly familiar strangers. Facebook controls the platform of this community; it mediates the digital link to our physical-space lives and owns the rich data trails that still might revolutionise our understanding of individual behaviour and social interaction.

So why does it seem trapped in the older world of sponsored search and product placements? One answer is that Google's tremendously successful business model has created a perception that digital marketing requires being narrow and precise, matching superficial intent (as expressed by a few words typed into a search box) with a targeted low-bandwidth message, such as a 35-word sponsored search advertisement, whose effectiveness is then measured by counting click-throughs.

But even a casual viewer of the television program Mad Men realises that successful advertising requires more than a carefully targeted message or generating leads or assessing intent. It requires persuasion, evoking a response, striking and maintaining a connection with the customer.

Facebook is ideally suited to expand into this terrain. It knows more about us than any company in history, information about what persuades each of us, what makes us react emotionally, what makes us sentimental and what turns us off. And persuasion doesn't end with the "what"; it also knows the "who". It can identify our friends and those whom we are likely to respond to, about what topics and in what contexts. It hosts user communities for most large brands, the starting point for digital customer retention. (After all, marketing involves more than just attracting customers: You also have to keep them.)

To capitalise on these assets, Facebook needs to push into real advertising - which involves persuasion, emotion and intuition - with the added ingredient of deep technology-based customisation. To succeed, it has to invent and execute a business model that expands digital marketing into a realm of truly attracting, persuading and retaining customers.

One building block will be the increasingly sophisticated science of automated sentiment analysis, which gleans the connection between words and phrases and the reactions and emotions they evoke. It can tell us, for example, whether specific users react more favourably to, and seem more persuaded by, Facebook posts that involve humour, sporting metaphors, religious allusions, logical arguments or references to history. Sentiment analysis can also help determine the best way to categorise these reactions and topics, the variations in the intensity of reaction across source and subject, and how these can be linked to emotional triggers embedded in marketing messages that will grab, engage and maintain a bond with a user.

Another piece of the puzzle lies in the new computational science of networks, which allows us to understand the structure of population-scale human social networks. Facebook has far better data on this than any competitor. This could allow us to determine whether users who occupy a particular position in their local friendship network or the broader social graph are particularly influential, or whether they are especially susceptible to messages of particular kinds, about particular products or topics.

The Facebook network generates data that can show how persuasiveness, or the tendency to be persuaded, varies across close friends, family members, colleagues, topic experts and casual acquaintances, across relationships of different strength and kind, and about how deep within the network this influence flows. This would enable the company to determine the mix of message content and source that will induce the right sentiment and the appropriate reaction.

As Facebook converts this heady blend of technology and sociology into a new business model, it will devise a way to personalise persuasion while simultaneously increasing its scale. This is well within the realm of possibility; the science already exists and Facebook has the users, the platform and the corporate relationships to make it happen.

The internet has moved beyond the Google-dominated formative era. Facebook's time has come. Rather than succumbing to the earnings pressure that will no doubt distract them in the coming year, Zuckerberg, his board and their investors need to summon the vision and courage to rise to this challenge, move past the fallout of their too-early IPO, and invent the future of digital marketing.

Arun Sundararajan is NEC faculty fellow and professor of information, operations and management sci-ences at New York University's Stern School of Busi-ness.

Bloomberg