Why Tumblr Was a Massive Steal for Yahoo

tmblr$The blockbuster acquisition of Tumblr by Yahoo for $1.1 billion — all cash — has been portrayed as a gigantic gamble on the part of the acquirer. However, for those of us who believe there can be specific guidelines around the valuation of interest graphs, it looks more like it might turn into a game-changing steal. Let’s look at some of the factors which might lead a big company to value Tumblr more highly than Instagram.

The more specific the interest, the more valuable.

After reading my recent thoughts about why Tumblr is so valuable, Josh Elman of Greylock Partners made a most astute observation: Everyone’s Facebook feed is pretty much the same as everyone else’s of the same age. Twenty-year-olds pose in the club, 30-year-olds share wedding photos, by age 40 you’re looking at a lot of cute pictures of your friends’ kids. But with Tumblr, you never know what you’re going to get — even with people you know personally. That, in a nutshell, is the difference between a social graph and an interest graph.

I have noted previously that Twitter is potentially the most valuable interest graph of all, because any given user can pick and choose from the entire universe of interest-based feeds. However, Tumblr has reminded me yet again of one of the Web’s fundamental truths: A single resource is far more specific than a feed. Knowing that someone follows the NBA Twitter feed is nice, but knowing that they hearted an animated gif of Steph Curry hitting a three or saved a recipe for beehive cupcakes is specificity gold. More to the point, it’s obviously far more valuable to know that someone is interested in hotels on Maui or Braun dishwashers rather than generally in travel or home decorating.

Unfortunately, some otherwise-great interest graphs leave money on the table by underutilizing the Web advantage of a unique resource — in other words, URLs or pages. Endless-scrolling feeds implemented in DHTML are attractive and account for a lot of pageviews, but if users never click through to individual resources, it becomes much more difficult to specify the level of interest in each item. The heart or save gesture goes a long way toward fixing this problem, especially if it also includes a grouping (a.k.a. “pinboard”) mechanism, but sites that provide little value at the page level are going to miss out on some of the value creation mechanisms we describe later.

Readers are far more valuable than writers.

Way back in 2000, my PandaWhale co-founder Troutgirl was an engineer at a startup called Epinions. It later became famous for the wrong reasons, but its early iteration still stands as one of the purest interest graphs ever built.

When Epinions started out, it tried to force everyone to register for the site and write reviews. It took being bludgeoned over the head with overwhelming data to realize that readers and writers are completely different animals — and that in fact it’s generally the readers who are the foundation of any business. Basically, writers turned out to be people with a lot of time and not much money, and readers turned out to be the opposite.

Writers have time but no money. Certain groups are going to be overrepresented: Students, stay-at-home moms, the underemployed, retirees. Epinions, which paid for product reviews, especially ran into issues with writers whose relationship to reviewed products lay more in the realm of fantasy than reality. Writers are also going to have the time and emotional commitment to give your site a lot of feedback about their needs and desires … many of which will be counter to the best interests of the business.

Readers have money but no time. They don’t want to spend hours combing the Internet for photos of vintage jewelry. They want to see a picture of a watch they like, and buy it now. If readers don’t find your content valuable, they’re not going to send you a long email about what they don’t like. They’ll just silently hit the back button and get gone.

It’s easy to forget this hard readers/writers dichotomy, especially in the era of “scaling via social networking” a.k.a. “spamming your Facebook friends for fun and profit.” It totally seems to make sense that you’d want every reader to become a writer, and to recruit their friends to become readers in turn. Thus you see Twitter, Pinterest and Quora going through phases where they more or less force all readers to register for accounts just to look at a few pages of content.

However, this reasoning goes counter to the all-important social versus interest graph distinction. You’re going to be interested in your Facebook friends’ photos and updates just because you know them, but their strange hobbies are probably not going to inspire you to take up the same hobbies. On the other hand, you will be interested in people with your same strange hobbies even if you don’t know them. (The distinction is actually probably even less symmetrical than this, in that you will invariably respect your friends somewhat less when your nose is rubbed in the fact that they enjoy cake pops or boxing; but you enjoy your interest graph more if the writers occasionally betray a trace of humanity.)

Also, those who have only worked on social graphs might not fully grasp how much more usage you can build in the long run by allowing for anonymous, non-social browsing. If you have good content that is meaningful to readers who do not know the writers personally, you are leaving money on the table — and increasingly alienating your potential userbase — by forcing a social graph that doesn’t need to exist.

Tumblr gets this right by making it easy to browse anonymously but difficult to save or find anonymously. Thus you can take a quick tour through a basically one-joke album Tumblog like F*** Yeah! Ryan Gosling, or you can derive extra value by following a rich daily feed like SI Photos or America’s Great Outdoors.

Valuable content is king.

Recently, a rash of apps learned firsthand about the heartbreak of exponential growth followed by exponential decay on social graph platforms. Josh Elman opined that the fundamental strategic flaw of these startups was a failure to move users rapidly enough from the Facebook platform to the standalone Web experience, but that’s missing a gigantic step in the logic of value creation. The apps that shut down had failed to develop enough of a standalone experience for users to have any reason to move over. Fundamentally, their content was just not valuable outside a narrow social, temporal and platform context and therefore moving users to a standalone site would have been pointless.

For just two examples, look at SocialCam (which let users easily share videos from their phones to Facebook) and BranchOut (which mined your Facebook graph to see who worked where). Both enjoyed what looked like explosive growth, but neither one had managed to encourage the development of content that was meaningful outside the Facebook social graph. SocialCam nimbly sold itself right before the implosion, while BranchOut reportedly went from 40 million to 40 thousand active users pretty much overnight when Facebook cut off their viral distribution. Exponential growth can lead to exponential decay in the absence of good content.

The most valuable content does not require a specific social relationship between the writer and the reader. It does not lose value quickly over time. It gains value through aggregation with similar content. And it should not change value radically based on existing on one platform versus another.

The uncomfortable point to all this is that super-rapid viral growth and long-term value of content are often at odds with each other due to the social versus interest graph dichotomy. A careful reading of the Tumblr and Pinterest stories shows that they had to carefully nurture a corpus of valuable content for quite a while before being able to achieve hypergrowth through the judicious use of social networks. At that point, viral spread added fuel to the fire, but a large body of valuable content was the real value proposition.

The closer to the point of sale, the more valuable.

Beginning entrepreneurs often ask what the business model of my consumer Internet company is, and I always say the same thing: Every consumer Internet company basically can do one of three things: sell stuff (commerce), point to other places that sell stuff (advertising) or sell add-on services to its own users (freemium). There’s really no other consumer business model.

It should be obvious that over the long run, the closer to the point of sale your user is, the better. This is why everyone is so excited about Pinterest. The entire point of Pinterest is ideally to function as a gigantic shopping catalog, with every viewer just one click away from the purchase of every item. It’s Amazon masquerading as entertainment. In fact, Amazon is the interest graph that has realized its value the most effectively.

But this is why social graphs — and some interest graphs — are difficult to value: So much of the content is not directly monetizable. A great example would be 9Gag: Not only is the content generally image-based and jokey, but a lot of the user base appears to be young and in China. That is three strikes against the commercial relevance of this site’s content ever being realized.

An even better example of high-usage low-value content, however, would be messaging. We’ve entered into yet another round of hot messaging (or photo-messaging) startups, this one smartphone-based. I have no doubt that they will all learn the hard way what generations of failed startups before them have learned: People who want free messaging are very far from wanting to buy anything except possibly same-day delivery of condoms (paging Google Shopping Express!) — and, in fact, any sort of commercial interest makes them disproportionately angry. Also, cross-platform messaging startups almost always end up being worthless in an acquisition context.

In the context of Tumblr, so far the commercial potential has been largely unrealized — but that doesn’t mean it doesn’t exist. A lot of Tumblr to date does trend in the 9Gag direction: Humorous rather than informative. However, the repeated republishing of attractive content is very attractive to brands that have strong social media strategies. Forward-thinking brands such as Nike, Target and the San Francisco Giants have demonstrated that they can give away images to get buzz among trend-following consumers.

Search is the best entry point.

Partly as a consequence of my previous point, search — whether from an external search engine or from your site’s own search — is still the very best entry point for any visit to any public resource. Search gives you intent, which is even stronger than interest. It also gives you more usage, in that any resource that can be found by one person through search will be desired and found by others.

Therefore, it never fails to surprise me that searchability seems to be so ignored by startups today. SEO is maybe out of fashion and not well understood by the youngs, but it is the driver of more value creation than any other similar mechanism on the Internet. Search-initiated transactions of all kinds are far more monetizable than their non-search counterparts, such as social network feeds or serendipitous browsing.

This may, in fact, be the area in which Yahoo can add the most value to Tumblr’s content by making it more searchable. Tumblr’s search functions to date haven’t just been weak — I have argued that they were actually broken by design. Not only will stronger search drive more traffic to Tumblogs, but it will almost certainly decrease the proportion of objectionable content, making it easier to restart growth after the merger on a healthier basis. Then the task of monetization employing targeting should be child’s play for any organization run by Marissa Mayer.


Tumblr is one of the most fully realized interest graphs built to date, and full of commercial potential that can be best unlocked by Yahoo. Interest graphs are more valuable insofar as they:

  • Allow for the expression of extremely specific interests
  • Cater to more readers than writers
  • Host content that is valuable without any specific social, temporal, or platform relationship between readers and writers
  • Capture intent, particularly commercial intent.

Tumblr is in an excellent position to fulfill all of these strategic goals, and as such must be considered among the most valuable interest graphs. If properly managed, Yahoo just made a game-changing acquisition which was well worth the $1.1 billion in cash.

Adam Rifkin is co-founder and CEO of PandaWhale. His interests include Grumpy Cat, Mark Sanchez buttfumble and 15 foods you can regrow from scraps.

Must-Reads from other Websites

Panos Mourdoukoutas

Why Apple Should Buy China’s Xiaomi

Paul Graham

What I Didn’t Say

Benjamin Bratton

We Need to Talk About TED

Mat Honan

I, Glasshole: My Year With Google Glass

Chris Ware

All Together Now

Corey S. Powell and Laurie Gwen Shapiro

The Sculpture on the Moon

About Voices

Along with original content and posts from across the Dow Jones network, this section of AllThingsD includes Must-Reads From Other Websites — pieces we’ve read, discussions we’ve followed, stuff we like. Six posts from external sites are included here each weekday, but we only run the headlines. We link to the original sites for the rest. These posts are explicitly labeled, so it’s clear that the content comes from other websites, and for clarity’s sake, all outside posts run against a pink background.

We also solicit original full-length posts and accept some unsolicited submissions.

Read more »