So, it’s official – while many of us (especially on Twitter) waited and watched over the weekend after All Things D broke the story, Yahoo has agreed to buy content sharing platform Tumblr for $1.1 billion in cash. The analysis has been varied and thoughtful: why Tumblr’s exit is about 30% of what it might have been (they failed to demonstrate that they could monetize on a timeline that would suit their investors/were running out of cash), why the value isn’t in their technology, but rather the networks’ function as a “vector for viral sharing“, and, finally, how Tumblr can “make money without pissing us off” (leverage tag pages, keep your hands off my content).
I had the chance to hear Tumblr CEO David Karp speak at the GigaOm paidContent Live conference in New York just a few weeks ago. In a 1:1 interview with Mathew Ingram, Karp referred repeatedly to Tumblr’s value in “Helping get people to the stuff that they’re actually going to love,” which struck me as both interesting and incomplete. If Tumblr’s value lies in not just your content, but ultimately in its ability to filter and anticipate and deliver that content in a way that adds incremental value, that’s promising. It’s also a model that requires significant technology investment (robust search and a brilliant, best-in-breed recommendation engine that uses highly sophisticated collaborative filtering). Content is like ore, and those technical filters are required to help users refine it in order to mine and generate that additional value. To do this your tech had better be killer; from what I have seen so far, I’m not sure that’s the case today.
Nevertheless, if that’s the proposed unique value prop – rather that just a publisher of cool stuff, building a better filter in order to be a content force multiplier (which is really the “building a better mousetrap” of the Information Age, when you think about it) – how do you monetize it? “Display advertising” and “paid search” are incorrect answers. These revenue models are so deeply commoditized that to begin to rely on them for revenue serves to commoditize your business as well, regardless of how unique or robust (“commoditization by association”). This is short-term gain for long-term pain, and as we’ve seen, deadly poisonous to future innovation in generating revenue from attention. If that’s the best Yahoo and Tumblr can do together, Tumblr will die a slow death like so many other past acquisitions, in this case because of what you might call the “MySpace death spiral”: reliance on eyeball- rather than engagement-based advertising erodes the user experience. Users leave (especially a risk here since micro/blogging technology is pretty ubiquitous). Eyeball-based numbers start to drop, revenue targets are missed, anxiety ensues and the short-term answer is to further junk up the user experience in order to deliver more impressions. Rinse and repeat, and within a very short period of time you have a virtual ghost town.
I would argue that this will be a hard trap not to fall into: platforms like Facebook have come to rely on the easy source of display revenue to the detriment of figuring something out that will actually allow advertisers to add value to the user experience at scale (sponsored posts are just another form of “spray and pray”). Of course, advertisers, and particularly their agencies, are the enablers here – they want three basic things: round pegs for round holes, scale, and “set it and forget it”. In my experience, anything that doesn’t fit into a CPM or CPC model is a bespoke option that is lovely to pilot, generates great results and shows the CMO some cutting-edge thinking. It then quietly gets discarded in favor of more efficient and reliable agency revenue streams.
There are three things on the Tumblr/Yahoo to-do list for 2013:
1. Push the edges on intelligent recommendation technology (just think of the dataset they must have!)
2. Innovate on engagement-based ad format design
3. Bypass agencies: try to build partnerships directly with brands to help bring scale and meaning to a revenue model that is based on adding value rather than the very comfortable “interrupt and repeat.”
If they don’t get two out of three right, we’ll be talking about Tumblr in the past tense faster than you can say “Rich Kids of Instagram”