Posts from Search Engines


The first great investment we made at USV was Indeed in the summer of 2005. Brad had been looking for a search engine for jobs and I saw this post on John Battelle's blog in late 2004. I forwarded it to Brad and he reached out to Paul and Rony. It took two tries before we could convince them to take our money. They had bootstrapped the company, launched the service, and were well on their way. They didn't need our money. But eventually we convinced them to take it, along with the New York Times Company and our friends at Allen & Company.

Indeed has always been the quiet one. Nobody really talks about them. But as I have said a number of times on this blog, they are the most complete company in our portfolio. They have it all. Two world class entrepreneurs as founders. A solid management team all up and down the company. A product that is beloved and services more than 80mm people worldwide every month. An engineering team that has kept the service up with literally no down time that I can ever remember. A business model that, like Google's, is the best on the Internet. Revenues, profits, customer satisfaction, shareholder value. They built a fortress and I am just so happy to have had a front row seat watching them build it.

The quiet one is the one that can do a big M&A transaction over the summer without anyone finding out. The quiet one is the one that puts out the news on their blog and goes back to serving customers. The quiet one is the first great investment we made at USV and one that will always have a special place in my heart. Congrats to Paul, Rony, and the team. We will miss working with you.

#VC & Technology#Web/Tech

Copyright and The Internet

Google has announced they have made a change to their algorithms to include valid DMCA takedown requests as a negative signal in their search results. I believe this is a step in the right direction and I am pleased to see Google do this.

It's hard to argue that the Internet can regulate itself when content owners point to Google search results that show blatant copyright violators on the first page and fully licensed services buried four pages down. So Google is addressing that. That's a good thing.

But I think we can do more. In the world of spam and malware, there are third party services that provide whitelists and blacklists that are used by various services and platforms to help them keep their services clean. I would like to see this approach extended to copyright.

It would be best if a competitive marketplace developed for copyright whitelists and blacklists. The list providers would use signals like number of valid takedown notices and a host of other data points, ideally provided by the marketplace of services and platforms, to produce real time lists of what services are fully compliant (whitelist), what services are blatantly violating copyright (blacklist), and everyone else (greylist).

Being on the greylist would not hamper a new service from entering the market. All new services built by three engineers in a loft would start out on the greylists. But over time they could get onto the whitelists by properly respecting copyright in their service. Whitelisted content services would benefit in the same way that whitelisted services benefit in the word of email.

Google effectively runs their own whitelists and blacklists in their email business and in their search business. But if there were truly independent and competitive providers of whitelists and blacklists for copyright, the entire market could do this sort of thing. And ideally Google would participate by providing data to these third parties and using the third party lists in their algorithms.

When I think about solving thorny problems like copyright, I like to look at how the Internet has solved similar problems and think about how these approaches can be extended to solve new problems. Whitelists and blacklists have been very effective in areas like spam and malware and I think they would be a great approach to address the copyright issues as long as they are built and managed by independent competitive third parties.


Two New Ways To Find A Better Place To Stay On The Road

My partner Albert said this a week or two ago on his blog:

When people ask me whether investing in web services is a long-term opportunity, I often say that “we ain’t seen nothing yet.”

I was reminded of that when I went to the web to find a hotel to stay in Miami when I am down there for Future Of Web Apps in February.

In the past, I'd have tried TripAdvisor or maybe even Google. But just in the past few months, there are two really great new ways to do this.

The first is Oyster, a web service dedicated to honest and excellent hotel reviews. They literally send out a person to stay in the hotel for a few days, take their own undoctored photos, and present in a clear and concise format. Here's their page on Miami hotels, and here's a page on Tides in South Beach, which they say has awesome rooms.

Oyster's strength, personally crafted reviews, is also its weakness because they only have reviews right now for certain destinations. I suspect it will take them a few years to be totally comprehensive. However, because so much hotel seeking traffic comes from Google, they can be a great service for people who start their hotel searches at Google. If you see an Oyster result in Google, you should absolutely click on it.

The other way to do it is Hunch. Instead of hiring "experts" to go out and review the hotels, Hunch takes an approach that is more like wikipedia. They crowdsource questions and answers from thousands of "contributors" and then package them up in easy question and answer sessions that lead you to the "right answer".

Try this hotels in miami page and see how it works. You'll notice that Hunch doesn't give you just one answer. It gives you a "best" answer and three other suggestions. I've found that not only do I get good advice from Hunch, it also helps me quickly understand the tradeoffs I am making in hotel decisions.

Of course, there's a natural partnership between Hunch and Oyster around hotels. Hunch is good at framing the decision and giving you a short list to consider. Oyster is great for digging into the specifics of the hotel to get to the final choice.

Since both companies are in NYC and are part of the great startup culture we've got going here, I bet that's not too hard to make happen.

Reblog this post [with Zemanta]
#Blogging On The Road#Web/Tech

"Audio Preview" Is A Bad User Experience

I read the Google blog post announcing their enhancements to music searches this morning. I think it's terrific that Google has made these enhancements but there's one thing I don't like.

Google says:

Now, when you enter a music-related query — like the name of a song,
artist or album — your search results will include links to an audio
preview of those songs provided by our music search partners MySpace (which just acquired iLike) or Lala.
When you click the result you'll be able to listen to an audio preview
of the song directly from one of those partners. For example, if I
search for [21st century breakdown], the first results provide links to
songs from Green Day's new album. MySpace and Lala also provide links
to purchase the full song.

I really dislike the "audio preview" experience. It's been available for years in the iTunes store and I never use it. A 30 second sample of a song is an awful experience in my opinion.

When I want to search for music, I'll do an mp3 search on the Internet or go to the hype machine and do the search there. I almost always get the result I want with one of those two approaches and I can listen to the entire song.

Of course, it is not Google's fault that they are being limited to an "audio preview", it is the fault of the rights holders who won't let Google offer a full song sample. But as we've seen again and again, this only drives users to the "gray market" where they can get a full song sample which is often just a right click away from an illegal download.

A smarter approach would be to allow Google to offer a full song sample (one play per person based on cookies or some other approach) and then a link to purchase. That would allow Google and the music rights holders to take share back from mp3 search engines and encourage music purchases instead of illegal music downloads.

Reblog this post [with Zemanta]
#My Music#Web/Tech