GiftRocket combines the thoughtfulness of a gift card with the flexibility of cash.

Does Paul Graham Get What He Asks For? [infographic]

Posted 31 August 2011 by

I first came across Paul Graham’s article, Startup Ideas We’d Like To Fund, in a Google search in mid 2010.

Though published two years earlier, I thought the article was the most thorough compilation of problem spaces for startups to tackle anywhere on the web. Before coming up with the idea for GiftRocket, we (the founders) treated it like a problem set. We’d methodically work through the list and have late night discussions about what kind of startup would be most likely to win the space PG was talking about.

Earlier this summer, I found myself explaining how nearly every idea PG outlined in the essay was being tackled by a startup today. Many of those startups were funded by YC, as recently as this S11 batch. And, YC-funded or not, many of those companies are considered to be “hot” in Silicon Valley these days. So yeah, YC has largely gotten what they’ve asked for.

I think that “Startup Ideas We’d Like To Fund” is one of the most prescient posts written in the past three years about the web 2.0 space. The infographic below maps recent YC startups and popular non-YC products / companies to the 30 categories outlined in the article. Here’s a larger version with years and comments. Please be sure to look past the image for some embed code and further notes.

Embed in your blog:

400px width version

600px width version

900px width version

If you do share or repost, please provide an attribution link to


  • There’s an obvious selection bias happening here. A lot of companies have tried to tackle these spaces and haven’t been successful. The ones that have been successful have been much easier to remember. Moreover, there are tons of companies (for example, GiftRocket) that YC has funded that don’t fall anywhere within the 30 listed.

  • Some of the companies don’t exactly fall into the space where I’ve bucketed them. For example, Disqus isn’t a news website. It is, however, related to the rise of blogging in that the company made a turnkey comments solution that ended up being used all over the blogosphere.

Greplin isn’t exactly a form of search that depends on design. But they are competing with Google, very directly at times. Conversely, while Hipmunk is a form of search that depends on design, they aren’t taking on Google so much as shitty flight search UIs (though Google is entering the space).

Perhaps my most egregious categorizations are the “buffer against bad customer service” companies. The intent of the category seemed to be finding a wrapper around particularly terrible experiences, like paying a parking ticket over the phone or booking a DMV appointment online (not great examples since they’re still somewhat terrible). But rather, I included two YC startups that are making it easy to communicate with people higher up in websites and physical stores.

Many listed companies fall into multiple buckets. For example, both Fitbit and Wakemate are companies that measure things that were formerly difficult to measure. But they are also examples of hardware / software combos.

There were several companies that matched the categories far better, but they were still off the record as of Alumni Demo Day.

  • You’ll notice that there are several spaces where I just didn’t fill in any companies. They were:

Enterprise Software 2.0 (#5): I’m just not familiar enough with that field to know any startups there. If you have thoughts, I’d love to hear them in the comments.

The webOS (#18): I think it is too early to say for any company, even including Google and Facebook. While there are startups pursuing this goal, right now they’re still building early versions of their product where it is too hard to tell.

A web-based Excel/database hybrid (#22): As far as I know, there isn’t anything like this out there. Google Docs does a decent job of scraping data and importing forms directly into spreadsheets, but Google Spreadsheets resembles Excel so closely that I don’t think this is what PG had in mind.

  • Thanks to the Crunchbase for providing easy access to company logos and founded-at dates.