The Twitter That Could Be

September 22, 2015
Henry Vasquez


The following is a letter I wrote to Sam Altman of Y Combinator, who had an inspiring rumination on Twitter back in February. He tweeted:
“tempted to write a YC RFS for a successor to Twitter. even with the network effect, this doesn’t feel like the last company in the space.”

I’m not entirely sure which direction he would like to see Twitter go, but it triggered a dormant vision for Twitter that I’ve had for a few years. So, thanks, Sam, for pulling this concept back out of my memory. Apologies to Dick, Jack, Ev, or Biz if you are already working on this.

— — — — — —
Sam —
You had a tweet on February 15th about a hypothetical YC Request for Startup for a successor to Twitter.
This concept is something I’ve spent hundreds of hours thinking about/discussing with others, and I’d like to share some thoughts. Perhaps you’ve done the same. Since I joined Twitter, I’ve always had this powerful belief in what Twitter could be, and I’ve seen the product and company evolve, but ultimately fall short of this vision.

Twitter is sitting on a gold mine much larger than they’ve actually realized, likely because they landed on this patch of Earth by accident and not intention. Twitter, in a way, created the new telegraph, or perhaps the telephone, and instead they fancied themselves a social media company or maybe a big data company, which they settled on by necessity, because they needed the capital of VC and ultimately the public markets to continue what they were doing. They could become a utility company of sorts, and create a wider moat and a formidable monopoly that is 10x larger than their current business.

TL;DR: Twitter has a constrained protocol of machine and human readable data that allows open communication across the internet and increases the velocity of information transfer, which means it has an opportunity to solidify itself as a utility of the internet instead of a social media platform. ;

The power of this lies in the 140 character limit and, in my vision, the marriage of machine and human-readable data creating a more semantic web. The result of humans of having limited characters is that we adapt and learn to communicate in more structured, semantic, machine-like ways. We use symbols (@, #) and abbreviations to pack more information into this medium, and in turn, we socialize ourselves and educate ourselves to communicate in more efficient ways. Twitter has ‘paved the cowpaths’ of these user habits by making some of these adaptive practices part of their platform, understood and object-oriented by both humans and machines.

But users have progressed far beyond where Twitter has taken them. They created short links to hyperlink more efficiently. They created image storage services to pack new content formats into the Twitter experience. They studied sentiment data and created news platforms, predictive analytics, and topical mediums outside of Twitter. They threw a “.” before the mention to hack the limitations of the original paved cowpath, the @ mention. They are constantly adapting, creating businesses and abstractions on top of Twitter, leaning into the 140-character constraint and enriching the ecosystem. But Twitter, the company, toes this interesting line in embracing, ignoring, and attacking these abstractions. They are inconsistent in their approach and don’t seem to lead, but react most of the time. They needed to generate revenue faster, so they embraced the paradigm of social media and advertising, because this is what the existing market leaders did before them. They tiptoed into the idea of data and the Twitter hose, because this is the second-order maturity of social media businesses. But they could have seen themselves as the communication protocol that brings humans and machines closer and proactively built their platform to encourage the abstractions to form and ultimately, to pay for the pipes that Twitter laid as a utility company of the internet.

What would this look like? This would look like a more aggressive paving of cowpaths to make it easier for developers to interact with these semantic objects. Twitter would study the ways in which users and ecosystem companies have built around their platform and provide a larger set of primitives to work with. These could include concepts like the @@ group mention, the location constraint, the stock ticker ($), the greater/better than (>) and its inverse, the continued thought (…), the answer prompt (____), the quotation (“ “), the statistic (%), the content format indicator ( [VIDEO]), the date (MM/DD). The counter case is that by making all of these semantic, we are constraining creativity of natural human language. This is valid, but I believe the trade-off is worth the value of having machines understand our semantics.

If Twitter saw itself this way, it could lay the foundation so that all written human communication online had a machine-readability to it, so that @sama on a comment feed on a random blog literally pointed at the person Sam Altman, that the Twitter protocol would exist throughout the entire web, capturing data and routing communication between humans and machines with ubiquity. And in doing this, Twitter would need to invent a pricing structure of connecting into this protocol that would look a lot more like a utility company. The value for all of the properties that leverage this platform would massively outweigh the incremental cost of plugging into Twitter, and the efficient market would pass these costs through to consumers in small ways by using this data to drive other spending, to increase the value of their products, and to enhance the customer’s relationship with those businesses that share this powerful platform in their unique ways.

This is what Twitter can be and may never be, as its path dependence makes such a radical paradigm shift unlikely and the opportunity for an entrant to fulfill this vision increasingly possible.


There are many Thought Leadership posts on Tribe’s Blog, here are links to other posts:

1 Comment. Leave new

Beautiful. I never thought of this and it is simply awesome. Twitter has made people learn how to use and interpret symbols and signs. It’s like an independent programming language.


Leave a Reply

Your email address will not be published. Required fields are marked *