June 2010 Archives
ad.ly is a relatively new, well-funded startup that puts ads "in-stream" on twitter, and is supposedly expanding to Facebook, MySpace, etc. Such expansion may be wise in the wake of Twitter's TOS changes, but that's another story.
This story is about what happened when I tried to spend money on ad.ly. In a nutshell:



- Downloading the full list of top influential twitter users from trst.me (~22K users).
- Hacking ad.ly's URLs and then downloading a big list of people you can advertise with.
- Cross checking with the trst.me list to only keep top influencers.
- Cross checking that output with the twitter API to only keep people recently retweeted.
- Taking that subset and downloading ad.ly info including price, followers, & avg tweets.
- Filter out people > 10 tweets/day and then sort by price/follower count.
- Fun: try this new search engine for a week http://duckduckgo.com/ PLZ RT! (Ad)
- New search engine http://duckduckgo.com/ PLZ RT! (Ad)
- Check out Duck Duck Go, a cool new search engine http://duckduckgo.com/ PLZ RT (Ad).
- Duck Duck Go is the new Google http://duckduckgo.com/ (Ad)
- Duck Duck Go is a new search engine http://duckduckgo.com/ (Ad)
- New search engine Duck Duck Go http://duckduckgo.com/ (Ad)
To get them off my desk and into your hands, simply send me your name and address and I'll put some in the mail!
Recently I spoke at Dreamit Ventures on SEO. I used Manu Kumar's #20tweets presentation format where I opened twenty tweets in twenty tabs and clicked them off as I went. The talk was recorded by Vijay Kailas from Numote.
The point of the talk was two-fold. First, I wanted to convince the Dreamit startups (and you!) that SEO is not only worthwhile, but is a traction vertical that all startups should consider seriously in the pursuit of traction. Second, I wanted to lay out a an actionable startup SEO strategy that all startups can follow.
- Startups should have an SEO strategy and spend significant time tracking, evaluating & tweaking it. 5 reasons to follow.
- Reason #1 to do SEO: HQ traffic. 88% of Google clicks are organic search (via @randfish). They're looking for stuff!
- Reason #2 to do SEO: edge. 2009 US spend on search ads, $18B; on SEO, $1.5B (via @randfish).
- Reason #3 to do SEO: precedent. It's worked for @yelp@aboutdotcom @scribd @slideshare @docstoc @nytimes@linkedin
- Reason #4 to do SEO: intl users. Just 22% of search traffic is US (via @comscore). Intl can make you an acquisition target.
- Reason #5 to do SEO: you can do it. It's not black magic. It's really just generating content and building links.
- SEO fact #1: ranking/click % drop-off is huge, ~40% for #1, 12% #2, 8% #3, >=10 (second page) just 10% (via @aol).
- SEO fact #2: top 10K keywords get ~18% (via @experian). top 1M ~(very rough)~50% (via @duckduckgo). 50% "long-tail"
- SEO fact #3: "20 to 25% of the queries we see today, we have never seen before" (via @google)
- SEO fact #4: conversion rates are much higher in the long-tail bc intent is less ambiguous (via @oneupweb).
- SEO how-to #1: you really need two SEO strategies, one for the fat-head and one for the long-tail.
- SEO how-to #2: what unique data can your business generate? That's the content basis for your long-tail SEO strategy.
- SEO how-to #3: make pages people want to link to naturally, e.g. for reputation, posts, resources, answers, stats.
- SEO how-to #4: help/nudge people link to you, e.g. URL shorteners, widgets, copy/paste, JS, customer emails, etc.
- SEO how-to #5: continually triple check your site for basic issues, e.g. robots.txt, URLs, redirects, duplicate content.
- SEO tip #1: anchor text still matters a lot, which can work wonders within your fat-head strategy.
- SEO tip #2: for your long-tail strategy, you want as flat a hierarchy as possible, and/or deep links.
- SEO tip #3: focus on pages that convert for your business. Traffic is a means to an end.
- SEO tip #4: avoid anything black hat, e.g. buying links. It's not worth the risk of getting blacklisted.
- SEO resources: @seomoz @seobook @adwords @compete Google Trends/Insights/SEO Starter Guide
The other day I had the pleasure of being on Mixergy with Andrew Warner, which is a popular video interview series about startups and business tips. There is also an audio version.
The interview is mainly about my last (second) company, Opobox, but it also includes (in the latter half) some q/a about DuckDuckGo and my blog.
It's about an hour, and yet I just really scratched the surface of a number of topics, and realize I left a number of other topics not completely explained. So if you have any follow-up or other questions, feel free to ask them on HN, in the comments below, by email or on formspring.

Let's put some #s on it. Suppose you want to end up with $5M (gross). Let's say for the moment, after all is said and done, the initial founders end up with 30%. If you have two co-founders (3 people) you'd need a $50M sale to end up with $5M, as you'd get 10% of it personally. With one co-founder, you'd need a $33M sale and with no co-founders, a $17M sale.Everyone obsesses with dilution from investors. The biggest dilution comes from co-founders. If you have 2 co-founders, you've diluted 66% before doing any of the hard work. Start by yourself and bring in co-founders for smaller stakes once you've got initial momentum. Unconventional wisdom, but the most economically practical advice you'll ever get.
After my super-slow Gmail post was picked up on HN and on NYT, Google reached out to me. I gave them my username and 39hr later my account is back to normal.- Lots of companies won't need VC. Chris Dixon recently wrote about how old VC firms should get ready to get disrupted. That was painfully obvious at both these events.
At Techstars, every single company was raising a small angel round. Not one company pitched for a Series A and yet a decent portion of the room seemed like VCs. This could have been due to a lot of reasons (advice, selection by TS, etc.), but bottom-line everyone felt it was in the interest of the companies to pitch this way.
Technology and customer acquisition are now cheap. $200K can get you a lot of validation. But interestingly, because validation is cheap, if it works, the companies may not need to raise more because they'll be instantly profitable. And if they do need to raise more, the valuations will be good.
I felt validated in my strategy to look for companies that, if everything goes as planned, may not need VC money at all.
- I need to do more deals. I was already planning on doing more deals, but this need was reinforced. There are two main reasons. First, the simulations show you need to do about twenty deals to guarantee some return given a reasonable distribution of hits. Second, deals create deal flow, and deal flow creates returns. By doing deals your network grows, and in turn, your deal flow.
- I should invest in hackers. I'm a hacker and I believe I understand the hacker mentality pretty well. I want to invest in people I trust, understand, and most importantly that I personally want to spend a lot of time working with. Basically, I want to invest in people like me :). It's become clear that doing so will both increase returns and increase satisfaction in my angel investing.
- I need to get plugged in more to the community. I suck at "networking." Like most people, I'm known in some circles, and completely unknown in others. I'd like to be more known in the angel investing community, and should find ways to make that happen.
- Philly needs work. Boston kept comparing itself to Silicon Valley. Philly is well behind Boston. It's clear that some local super angels would help. Boston seems to have a few.
