To potential buyers like Verizon (which already owns AOL and is buying Yahoo), Twitter is primarily a media distribution platform — or more bluntly, an audience to advertise to. To business software vendors like Microsoft and Salesforce, its main attraction is as a huge store of social data. For Google, it's both the above. A few, perhaps, still see it as a social media messaging platform, though ironically it has never made a business success of that core activity, which is how it comes to be up for sale in the first place.
Do any of those — or all of them put together — justify the $30 billion price tag that some say Twitter's board is angling for? Probably not. Twitter is a trophy acquisition. Its board can get away with asking a high price — despite its track record as a loss-making business — simply because it's such a resonant brand, both socially and politically. And with that status comes a lot of unwelcome attention for whoever buys it. Recode's Kara Swisher sums it up succinctly:
Twitter is the digital equivalent of a jewel wrapped in a very gnarly hairball.
It's the data, stupid
So what is the hidden jewel that makes Twitter so valuable now? Not the company's business acumen, that's for sure. Nor its reach, which though substantial is overshadowed by other, faster-growing messaging and sharing platforms. It's the data.
It's a dozen years since IT publisher and seminal conference organizer Tim O'Reilly began writing that "data is the next Intel Inside." His contention was that data was at the heart of every successful Internet business, from domain name registries to Google and Amazon. Therefore ownership of data would dictate market control in the Internet age, just as Intel's dominance of microprocessors allowed it to rule the PC era:
The race is on to own certain classes of core data: location, identity, calendaring of public events, product identifiers and namespaces. In many cases, where there is significant cost to create the data, there may be an opportunity for an Intel Inside style play, with a single source for the data. In others, the winner will be the company that first reaches critical mass via user aggregation, and turns that aggregated data into a system service.
That final sentence feels utterly pertinent to Twitter. But there was one thing missing from this analysis at the time. As I wrote back then, it isn't the data that is valuable so much as the insights that you can get from it: "owning data won't get you very far unless you can put it in a context that adds value." Collecting data per se is just meaningless hoarding (or more precisely, hoarding meaninglessness).
AI makes data meaningful
The difference today is that artificial intelligence now provides cost-effective tools that can automate the process of finding meaning in data — that can be trained to add valuable context. The past dozen years have seen massive advances in AI, especially with the advent of 'deep learning' techniques since 2006. The continuing acceleration in processing power that can be applied to AI is driving down costs, while the explosion in available data provides the raw material to feed these increasingly powerful learning machines.
AI is becoming a race that no one can afford to be locked out of. To succeed in that race, vendors need three ingredients — a team of smart data scientists, advanced infrastructure, and masses of data to mine. Collecting the third ingredient is no longer a meaningless activity if you can apply the first two to it.
In retrospect, therefore, Twitter hugely enhanced its value in 2014 when it bought Gnip for $134 million, the service that stored Twitter's historic data and made it available for analysis. What was merely a tactical move to boost revenues through selling high-priced access to that data store, now with the benefit of hindsight looks strategic.
As Salesforce demonstrated with the announcement of its new Einstein AI platform, the cloud business software giant has been making significant investments in AI expertise and infrastructure. During a media briefing to show off Einstein's capabilities earlier this month, the company showed one example of how it could be applied to Twitter data.
Einstein, meet Twitter
The demonstration showed how a user of its Social Studio marketing tool can simply drop in a set of pictures that include a specific object, such as a football helmet, and the Einstein AI learns how to identify images containing that object. This capability can be used, the company says, to identify pictures containing a company's product or logo (though this wasn't demonstrated), enabling what it calls 'visual listening' on social media — getting automated alerts and analysis when people tweet pictures of your company or product.
Imagine the power for a marketing team of being able to analyze the historical record of every image ever posted to Twitter, along with the location, the identities involved, and the sentiment expressed. And it's just one example.
As Business Insider's Eugene Kim observed on Friday, Twitter also plays a big role in sales, which is Salesforce's core market but one where it is keen to reinvigorate its growth:
Twitter is one of the most widely used prospecting tools used by salespeople these days.
In fact, Twitter is second only to LinkedIn as a favorite tool of salespeople, but Salesforce already lost out on that acquisition target when it was outbid earlier this year by Microsoft. All the more reason why it might covet Twitter.
Ron Miller at Techcrunch also drilled down yesterday on the value of Twitter's data to Salesforce, after speaking to several analysts including Brent Leary of CRM Essentials, who explains:
Salesforce’s competitors are snapping up [data sources] and will integrate them into their platforms to add additional perspective and intelligence. If this deal with Twitter happens, it’s to add a constant flow of information into their AI platform, to marry it with their transactional and customer information.
Where Quip fits in
While Salesforce already pays for access to that data, it would be able to do much more if it could unleash its data scientists and AI tools within the actual data store. As Miller notes, an acquisition would also guarantee continued access, which might be denied if Twitter ended up in a competitor's hands.
The prospect, of course, of a commercial organization constantly mining Twitter data for sentiment is one that is bound to raise data privacy and cultural hackles. Twitter's particular commitment to freedom of speech makes it popular among users in countries that take a less favorable view of commercial mining of personal data than does the US. Any acquirer will face questions regarding its intentions and will doubtless have to face down virulent opposition to its plans, however pure it may believe its intentions to be. That would be a headache Salesforce would have to take on.
But several observers have noted that Bret Taylor, CEO of productivity apps vendor Quip, which Salesforce acquired last month in a $582 million deal, is already on the board of Twitter, and so is close to the issues it faces. Quip of course, with its focus on cloud documents, ties into the whole collaboration space, which Salesforce already plays in with its Chatter social stream for in-company conversations. Having Twitter as the external counterpoint to Chatter and perhaps folding in some collaboration capabilities tied to Quip and other Salesforce assets may also be attractive.
What price rival bidders?
The trouble with all of this is that Salesforce is said to be up against deep-pocketed rival bidders. Microsoft is among them, although acquiring Twitter so soon after LinkedIn would probably be too much of a red rag to regulators to seriously contemplate.
The most likely acquirer is Google, where there are several synergies across both its consumer and business product lines. Given its current focus on building up its enterprise credentials, I was intrigued by the suggestion made by Madison Malone Kircher at New York Magazine that Twitter might complement Gmail in the identity space:
One direction for Twitter at Google would be to turn Twitter accounts into the universal online identities they always almost have been.
But Salesforce probably stands out as Twitter's most eager prospective acquirer. Whether it can succeed will depend on a combination of how much money it's willing to stump up (most of which it will have to borrow) along with how much CRM stock Twitter's board (and shareholders) are prepared to accept in lieu of cash.
On the eve of Dreamforce, this is an interesting spectacle playing out.