Facebook Goes Nuclear On The Messaging War With Its M Assistant

Next Story

Breaking: Vegemite May Power The Electronics Of The Future

All messaging apps are the same no more. A few years ago, there were few differences between SMS and a slew of chat apps like WhatsApp, WeChat, Kik, Line, KakaoTalk and Facebook Messenger. But each have slowly defined themselves differently, whether through simplicity, connections to businesses, media, stickers, games, location, and other features.

11095239_10156070655280195_5860295573424755388_nBut today, Facebook escalated the battle for chat with the official announcement of M, its new personal assistant built into Messenger. It can actually complete tasks for you, such as buy things, deliver gifts, make reservations, arrange travel, or just about anything else you ask of it.

What’s truly unique, though, is what’s behind M. Specifically, a lot of live bodies. M isn’t just artificial intelligence. Facebook has contracted real people to help M answer people’s requests while teaching the technology how to handle them automatically in the future.

M is still in early beta testing, so it’s hard to assess just how well it accomplishes this grand ambition. But if it works, Facebook may have developed a product so useful yet so complex and resource-intensive that it could differentiate Facebook Messenger in a way its competitors can’t or won’t follow.

  • There are digital assistants like Google Now and Siri. But those are so mechanized that they can only provide rote answers and reminders.
  • There are personal assistants like Magic and Operator that use humans to answer complex requests, but they’re independent apps without massive scale.
  • And there are messaging apps fighting to grow their already-huge audiences, but that still look similar despite their attempts to differentiate.

M combines the power of a world-leading artificial intelligence lab with the dexterity of humans Facebook can afford to hire and the scale of its 700 million user Messenger app.

As I noted earlier, while Google and Apple were dicking around with the pure science of artificial intelligence, Facebook used human helpers to brute force a full-featured assistant. The closest thing to M might be the third-party app WeSecretary built atop WeChat.

11890988_10156070655185195_4945843629380291283_nMaking M work for all of Messenger’s users might be slow or expensive, but it’ll probably be both.

Eventually, M would ideally work with minimal human assistance. To get there, Facebook needs time for its M contractors to teach it the best way to solve problems. With a small workforce and small beta, that could take a while. Growing M’s test base and the legions of helpers behind it will cost a ton.

But that’s why Facebook is so distinctly well-equipped. It has money. Not quite Google or Apple money, but with $4 billion in revenue and around $700 million in profit last quarter, Facebook has resources to throw at M.

It also has time. Facebook’s ownership of both Messenger and the 800 million-user WhatsApp gives it a decisive lead in messaging. It doesn’t have to worry about falling further behind while it concentrates on R&D. I’m looking at you, Google Hangouts/Messenger.

And finally, with both Messenger and WhatsApp in its possession, Facebook has arguably the best ways to leverage a chat-based personal assistant.

11898862_10156070655505195_4986153176672234416_nHow will Facebook recoup this massive investment? There are plenty of ways M could make money.

For example, Facebook could establish relationships with certain product or service vendors, earning a cut for making them M’s go-to provider for certain requests. Similar to how Shyp earns money on the difference between the bulk shipping discounts it gets and what it charges users, Facebook could surely find some margin to absorb it if it can power frictionless personal assistant shopping and travel booking.

Then there’s the massive opportunity to run M-triggering ads. Imagine a Facebook ad that prompts you to message M and set up an appointment at a local barber or restaurant. Facebook already has “Missed Call” ads in India that get you to call a business and immediately hang up so you aren’t charged, then get called back by the business with more info.

And there’s always the potential it could just charge users directly for the service, though that’s really not Facebook’s style.

But the beauty of Facebook’s strategy is that it doesn’t have to monetize M, or Messenger, directly. While competitors scramble for ways to earn cash, Messenger lets its parent app handle the finances.

All Facebook has to do with M is make Messenger more useful, and thereby used more. The chat app locks people into Facebook’s social network, and that’s where it keeps the real money-maker: the ad-filled News Feed.