Blog Post

The massive advertising shift that Twitter is trying to capitalize on with its API

Much has been written about the disruption taking place in the media industry as a result of the web and the atomization of content, but less has been said about how the advertising business — on which most of the media industry continues to rely — has been going through its own disruption. One of the seismic shifts that has been upending business models is the one that Twitter is trying to capitalize on with its API: namely, the increasing move towards “programmatic” or automated ad buying, as opposed to the traditional human-driven ad game.

This is about more than just some tinkering with the machinery that underlies the content we consume. As Felix Salmon notes in a post at Reuters, virtually every media entity — from behemoths like the New York Times(s nyt) to the smallest independent online player — is being forced to reinvent how they generate revenue because digital advertising is not paying the bills (Note: We will be discussing alternative monetization methods on a number of panels at our paidContent Live media conference on April 17 in New York).

Another battle between algorithms and humans


Like most of the other changes in the media industry, this shift didn’t just just happen overnight. Instead, it’s a wave or a series of waves that have been building steadily over the past several years. As with so many other elements of the disruption in digital content, Google arguably triggered this particular tsunami with the introduction of its search-keyword based auction process — an idea it borrowed from Bill Gross and Overture — but the ripple effects of that decision have continued to grow and expand in force.

In a nutshell, advertising has become yet another battleground for the fight between humans and algorithms: the human beings at ad-buying firms and ad networks who used to buy and sell banners and other traditional forms of advertising, and the algorithms that drive “programmatic” buying based on keywords, topics that are trending in social media and other factors.

As we described earlier, this is something the New York Times has been experimenting with via an in-house tool that tracks which stories are seeing the most activity on Twitter, and then offers advertisers the ability to insert their ads into those stories. It’s no longer about where that content appears in a physical product like a newspaper, or even what the story is actually about. Instead, it’s more about who is engaging with it, and where, and how.

Part of this shift is the transition from what some marketers call “outbound” marketing — which includes direct mail, banner ads and other methods that try to reach out and grab the attention of potential customers — to “inbound” marketing, which relies on search-engine optimization and other content-based strategies that make it easier for users to find a brand or advertiser without being bombarded by ads (Demand Media (s dmnd) and other “content farms” have tried to apply this approach to content, with mixed results).

Hubspot, which specializes in inbound marketing, says it is more or less taking over the online world, and the consequences for traditional media companies are fairly obvious:


Automated buying vs. human-created content

So what does all this have to do with the Twitter advertising API? In a nutshell, the API is a way for companies to automate more of their ad spending. At the moment, the list of official partners isn’t that long, but it will presumably grow — and the API combined with Twitter’s self-serve ad platform will theoretically allow advertisers to promote tweets based on what is trending and where the activity is. Twitter is also likely providing a host of information around users and their interests.

This is essentially the same game Facebook (s fb) is playing, and while Google doesn’t have an open API for Google+ yet, it is likely thinking along the same lines. For social networks, in which the content generated by users is almost indistinguishable from the advertising — and in the case of Facebook, actually becomes part of that advertising, through features like sponsored stories — offering tools that let advertisers automate their spending based on hard data is potentially far more lucrative than another generic banner ad.

The problem for many media companies is that this is a game they are ill-equipped to play: for the most part, they have little or no data about their users that can compete with the granularity that Facebook or Twitter can offer, and they have no APIs or other automated, self-serve features to offer even if they did have that kind of data. On top of that, brands are setting up newsrooms and becoming content publishers in their own right, and further disintermediating the media.

This is part of the reason why so many players like The Atlantic and even Gawker Media have been focusing on alternative methods such as sponsored content or “native” advertising or affiliate links. But as Salmon notes in his post, these kinds of approaches are often labor-intensive (if you want them to be effective), and therefore high cost. In some ways, the market seems to be bifurcating: on one side is a growing business driven by algorithms, and on the other is a human-driven business based on customized content. Is there room for both?

Post and thumbnail images courtesy of Shutterstock / Everett Collection and Borrell Associates

6 Responses to “The massive advertising shift that Twitter is trying to capitalize on with its API”

  1. Guy Cookson

    This is a really interesting piece, and we are looking at developing an API too in the future.

    But right now what we are focussed on is giving premium publishers a way to justify charging a premium for their advertising.

    The problem is IAB units have become just another commodity. And just as in any other market, if what you’re selling can be bought elsewhere for less it’s almost impossible to retain your margin.

    But we believe there is another way, and that’s for publishers to have their own native ad formats that are exclusively theirs and that deliver a better return for their advertisers and a better experience for their users.

    If of interest there’s a brief summary here:


    Guy Cookson

  2. Greg Satell

    I think you’re conflating a bit here. Without going into detail, different advertising channels have different tasks.

    For instance, paid online media today is often more about retargeting than targeting (i.e. identifying people who’ve already shown some interest), so it makes sense that they are automated.

    Owned and earned media are more about generating interest (again, I’m oversimplifying a bit), which requires more human driven imagination.

    They’re not competitive as much as they are synergistic.

  3. There will be a connection between algorithmic advertising and inbound marketing. I believe that it’ll be through paid distribution of content that is personalized to the user based on context, behavior and demographics. Outbrain is making the connection quite well proving that ‘delivering content in a personalized way’ gets much higher click through rates than advertising.

    Inbound marketers are creating lots of great content, but it doesn’t get read by enough people organically.

    Steve Hall wrote about this in “How Inbound Marketing Can Fuel Native Advertising” article:

    In an article I wrote recently called, “9 More Reasons to Invest in Inbound Marketing”, I talked about how “Inbound Marketing Creates the Foundation for Generating a Measurable ROI from Advertising.” #5 in the article:

    You’re spot on with this Matthew. Let me know if you want to talk about it. I think there’s a lot of disruption about to occur when inbound gets combined with the right algorithms.

  4. francine hardaway

    You might want to talk with Rodney Mayers at Proximic, a Palo Alto company that analyzes this kind of programmatic buying for “safety.” Is it appearing where you want it to? Are you next to something you wouldn’t want your ad next to? Ad collisions, sites with big hits but useless content for most brands (Manga, adult dating), and so on. I’ve been wildly re-educating myself as I teach Business and Future of Journalism:-)