Tumblr as one giant human relevancy filter and how e-commerce might save them

Back in 2008, David Karp and John Maloney visited me and my friend Bre Pettis at the Etsy offices to discuss an idea we had*. David shared with us this idea that Tumblr could be a new kind of search engine for the internet.. a search engine whose index only contained content a human had deemed valuable.

Anthony (soupsoup) has clearly understood that potential from the beginning, both in how he uses his Tumblr dashboard to curate news and how he manages Neighborhoodr.

Anyway, yesterday I finally got to meet Christina Cacioppo, Union Square Venture’s newest associate, and she told me that she’s been seeing a lot of pitches lately trying to tackle the relevancy problem through personalization (something I know a lot about having been the product manager for that team at Amazon.com). This is both relevancy as it relates to news, e-commerce, or otherwise (art, music, relationships). Our conversation reminded me of what David Karp said and what Anthony does… and given Tumblr’s latest round of funding I’ve been thinking a lot about what I think Tumblr could do to create $500M in value (which is what I think their exit would need to be given their latest round and valuation).

So where is the real value? I don’t have it all figured out yet. My intuition is applying the tumblr human filter network to long tail categories - fashion, news, humor, art, music - as a way to extract the diamonds from the rough. Is that monetizable? Maybe it is, in the context of e-commerce, where people pay to purchase those fashionable objects, artistic pieces, digital albums, or concert tickets.

Tumblr currently has 47.8M uniques a month, according to Quantcast. For the purposes of easy math, let’s say they get to 100M uniques in 12-18 months, and 3% of all users buy something via Tumblr post. Furthermore, let’s say the average item price is $15 (way less for music, way more for clothing / tickets / art) and Tumblr takes a 5% transaction fee. 100M * 3% * $15 * 5% = $2.25M / mo or $27M / year. Add on a couple other revenue streams (premium themes, sponsored posts in your dash, sponsored Radar items, sponsored Tumblrs you might want to follow) and I could see them getting to the $40-50M / year range. Not sure that gets you acquired for $500M, but it’s in the ballpark.

What do you think?


*Basically, give every Etsy user a tumblog, and populate it with all of the items they “favorited”. Then, you could follow the tumblogs of people who’s tastes you loved. The motivation was to solve the “I can’t find anything on Etsy” problem, plus increase user engagement. We both left Etsy before we could make it happen. Hopefully SVPPLY can fill the gap we saw.

  1. ssundar reblogged this from caterpillarcowboy and added:
    an apt description...some friends whats
  2. gsiener reblogged this from caterpillarcowboy
  3. houseoforange reblogged this from caterpillarcowboy and added:
    Our Hero’s inner cinema nerd read...Human Tumblr-pede.
  4. charleshuang reblogged this from caterpillarcowboy and added:
    thinking (particularly like...filter”, with nods
  5. hipsterdiet said: this is great, soup and i had a very similar chat in june about this. nice work.
  6. caterpillarcowboy posted this