Platform as the product

Today was a big day for platishing! Kidding, don’t use that word. But the vision outlined in 2014 by Jonathan Glick seems to be proving itself true, at least partly.

Unfortunately Medium’s endeavor in the space fizzled to much disappointment, and Gawker Media’s sale to the Univision Fusion Media Group ended the “let 1,000 flowers bloom” experiment that was Kinja. But two Washington-based companies have positioned themselves to capitalize on the still-present environmental factors that have created the space for this new business model. Vox Media snagged Bill Simmons’ Ringer in a simple rev-share agreement (no licensing fees), sparking this tweetstorm by Philly expat Sean Blanda:

I dig this idea, and it led to a great conversation with Sean as well as media pals Gabriel Stein and Noah Chestnut about the market opportunity and specific offering here.

I’m gonna try to synthesize the offering as we see it. Want us to build it? Hey maybe! Hit us up, @a16z.

  • Prioritize ad products that fill top of funnel for SMB’s. #
  • Ensure consistent ad quality and protect user trust at all cost. #
  • Support alternative revenue streams, from membership/paywall to products to events and beyond. #

These bulletpoints are abstracted best practices that, to varying degrees, media companies of all sizes are pursuing today. There are a couple hard points we discussed.

How could you fund this?

  • Gabe proposed having founding publications donate their engineering budgets and/or capacity, and also wondered if the Visa model could work (formed as a joint-venture between banks that could otherwise still compete).
  • I proposed tokens (half joking but read this).
  • VC would be an option, as would bootstrapping from within an existing media company (Vox did both).

How would outlets pay for it? Well that depends. The Vox model again shines here due to the simplicity of revenue sharing – the Ringer has no cash downside to this deal, just opportunity cost. But as Gabe notes, that deal works precisely because Vox can use its existing direct/programmatic salesforce to sell the Ringer without much marginal effort outside of normal deal flow. It’s not at all clear the same is true for local (or niche) publications, and in fact a prima facie analysis suggests that no, local doesn’t scale.

So there’s also a model where salesforces are no longer maintained by the outlet. The Washington Post’s Jarrod Dicker pushed this perspective, likely signaling groupthink within his R&D team working on the ARC project:

SSP’s, by the way, is an abbreviation of supply-side platform, the “buyers” of the programmatic market. This insight rings true to me based on my experience at, where programmatic ads were expected to pick up the slack for diminishing direct sales prospects. Thanks to Facebook and Snapchat, businesses of all sizes are more and more acclimated to purchasing a specific type of person or profile, rather than taking a run-of-site or section sponsorship. Yes, Facebook and Google can support this type of purchasing, but the more transactions we can get to occur outside the duopoly, the better off everyone who’s not named Larry, Sergey, or Zuck will be.

Publishers could also pay for such a platform through a license fee or some kind of volume based pricing, but rev share seems on paper at least to have a better chance of both parties rowing in the same direction.

Why couldn’t this happen today? Well, it does in some way, but collaboration occurs on components of the stack, not on how the components are assembled together. The integration layer is almost always proprietary (few if any major media outlets have open-sourced their entire codebase) and it’s quite frankly where user experience lives or dies. The integration layer can also be conceived as the product manager’s canvas; all the inputs to the product are interesting but ultimately the market will judge the output alone. Outlets don’t collaborate on their integration layers outside of whatever corporate parent they might happen to share.

One of the most difficult implementation details for this model would be identifying what product control publishers need to relinquish to the collective or platform management, and how pushback or disagreement on execution and strategy would be handled. Governance may become a complicated effort. This highlights the importance of a breakup agreement between the publisher and platform. Content portability would be required of course, but what other affordances would help publishers in the event of a separation?


Posted May. 30 2017, 12:41 pm by Davis