SimpleGeo

» see simplegeo.com as it lived, on the wayback machine

~ the obit ~

SimpleGeo was founded in 2009 in San Francisco by Matt Galligan and Joe Stump, the latter having been lead architect at Digg, and the pitch was clean enough to repeat: Amazon Web Services, but for location. In 2009 a developer who wanted to answer "what is near this point" had to stand up PostGIS, find a places dataset with a license they could live with, and maintain both. SimpleGeo sold that work as hosted APIs instead. Storage held your own geodata and Places answered queries against a business and landmark index, while the one they pushed hardest was Context, which returned what was simply true at a coordinate: the weather, the census figures for the block you were standing on.

It raised about $10 million across 2009 and 2010 from a list of names that tells you what the era valued, with Ron Conway and Shawn Fanning among the early investors. The bet was that location was a utility layer and utility layers get bought by everyone. What happened instead is that the companies who most needed the layer built it themselves, Foursquare included, while Google gave a competing one away. Developers who only wanted a places index found that Factual would sell them one directly. From June 2011 the Places data behind SimpleGeo's own API was coming from Factual, which is a strange position for a data company to be in.

Galligan stepped down in August 2011. On 31 October 2011 Urban Airship, the Portland push-notification company, announced it had acquired SimpleGeo in an all-stock deal reported at roughly $3.5 million, against the $10 million SimpleGeo had taken in. Urban Airship went from 38 employees to 51. CEO Scott Kveton described the fit as a natural extension, and the honest reading is that push notifications wanted to know where the handset was, so buying the location team was cheaper than hiring one.

On 12 January 2012 Urban Airship said it would wind down the current versions of all three APIs, with the lights off on 31 March 2012. Existing customers got six free months to the door and a migration path to Factual, which was already supplying half the answers. Ten weeks between the acquisition announcement and the shutdown announcement is not a change of plan. It was the plan.

~ the rap sheet ~

~ the pitch that aged badly ~

"AWS for X" was a good pitch in 2009 because AWS itself was still being explained to people. The flaw in applying it to location is that S3 stores bytes you already own, while a places API sells you somebody else's dataset. A storage layer gets stickier as your data grows in it. A data layer gets weaker every time a bigger holder of the same data decides to compete on price.

SimpleGeo built the developer experience well and never solved the supply side, which is why the Factual arrangement in June 2011 reads as the real end date. Four months later the company was sold.

~ the six-month wind-down ~

Giving departing customers six months free is generous by the standards of acquired API shutdowns, and it was also the cheapest available way to make an awkward number of angry developers into somebody else's customers. Factual took the migration. Urban Airship kept the context work and folded it into push targeting, which is the part it had actually paid for.

Joe Stump stayed on as an advisor. Urban Airship later renamed itself Airship and is still trading, so the acquiring company outlived the thing it acquired by well over a decade.

~ what is at the address now ~

simplegeo.com still resolves and still answers with a 200. The response is 114 bytes. No farewell page, no parking lander, not even a redirect to Airship. A hundred and fourteen bytes of nothing, served by a host that has no reason to stop.

This whole category sat outside my reach in a way that has nothing to do with Turkey. I was working as an engineer through these years, and location APIs were exactly the kind of thing I would have read the docs for, but the products built on top of SimpleGeo were American check-in apps, and the deals its investors were making were the kind you hear about later in a post-mortem. The distance here is the distance between a developer who reads the changelog and one who gets the acquisition email.

@}-->--

              )
             (_)
            .-'-.
            |   |
            |   |
            |   |
            |   |
          __|   |__   .-.
       .-'  |   |  `-:   :
      :     `---'     :-'
       `-._       _.-'
           '""""""

« previous  ·  random  ·  next »