Friday, June 29, 2007

Marc Andreesen's Blog: Startup advice

I just discovered Marc Andreesen's blog. There are a number of interesting posts here about doing your own startup. One post was about what VCs look for when they invest in a company. The second was about how to get to know VCs. And the third was about the most important thing for a startup.

What do VCs look for in a company?
  • Founder risk - right founding team for the opportunity?
  • Market risk - is there really a market for this?
  • Competition risk - are there too many startups/incumbents in this space? Is this startup sufficiently differentiated?
  • Timing risk - too early for this? Too late?
  • Financing risk - how much total capital will be required? How certain are we?
  • Marketing risk - will this startup be able to cut through the noise? How much does it cost to acquire a customer?
  • Distribution risk - are there certain distribution partners required? How will it get them?
  • Technology risk - Can the product be built? Does it require fancy, new, unproven technology? Is it dependent on some other tech breakthrough?
  • Product risk - can this team build the product?
  • Hiring risk - what positions does the startup need to fill into order to execute this plan?
  • Location risk - is the team not located in Silicon Valley?
How to get to know VCs?
  • VCs work through referrals
  • Work at a VC-backed startup, kick butt, get promoted, and network all the way
What is the only thing that matters for a startup?
  • Out of team, product, and market--it's market.
  • A great market (with lots of potential customers) pulls product out of the startup
  • There's a concept called Product-Market Fit (PMF)--being in a good market with a product that can satisfy that market
  • If you're Before PMF (BPMF), focus obsessively on getting to PMF. Most startups fail because they can't get to PMF before they run out of cash.

Wednesday, June 27, 2007

Notes from Reid Hoffman talk at Stanford

Just listened to an archive of Reid Hoffman's talk at Stanford a few months ago. The link to the talk is here.

Key take-aways:
  • Think of entrepreneurship like investing. In investing, there are 2 strategies: (1) portfolio diversification/risk mitigation, and (2) accurate contrarian theories. Entrepreneurship is like investment approach #2. If it's conventional wisdom, probably not a good start-up idea.
  • Test your hypotheses as quickly as you can. Work as if you're going to succeed, but get to your failure points as quickly as possible. Sequence it so that you test the biggest risks as early as possible.
  • 2 reasons for this: (1) Don't want to waste your time. If you could have found something out in 1 year that would have led you to abandon your idea, but you waited until year 3 to test it, then you've wasted 2 years. (2) There's a market readiness and luck factor to all startups. Sometimes you're going to be lucky, sometimes you're going to be wrong. If your time cycle for testing out new ideas is long, you won't have as many chances over a career to be lucky.
  • Success in entrepreneurship is solving the easiest, simplest problem that's valuable--not necessarily by coming up with the most elegant solution to a very difficult problem (especially if it's not valuable)
  • On competition: Best scenario, do something where there are no competitors. Second best scenario, do something where there are slothful competitors (like banking). Don't go after an aggressive, smart, incumbent competitor head on.

Sunday, June 24, 2007

Founders at Work - Excite

Excite founder Joe Kraus

  • Excite started as a group of friends from Stanford who wanted to start a company--but didn't know what they wanted to start
  • The idea went through some iterations--CD-ROM search to web search
  • "All of the pivotal things [in the company] were unintentional... The intentional things were rarely pivotal in those early days, but the being persistent, following-your-nose thing made a big difference."
  • Getting the Netscape "search" button was a major tipping point for Excite. They bid $3M for the contract (which they didn't have--that only had $1M--but Vinod Khosla said they could raise the rest). And they lost. But Vinod told the team not to quit.
  • Re Netscape: " We did all this stuff; we called them constantly; we just basically acted like the bidding wasn't over. And made a total pain in the ass of ourselves... Then luck struck: MCI couldn't deliver its service to Netscape on time... I can tell you that, had we given up, we never would have gotten this deal back. And without that deal, I don't think Excite would have had its run at all."
  • "The persistence part is the part I like. It's actually not fun when it's happening, but you know it makes a difference because 99.9 percent of the people give up. And Vinod gave me that lesson in spades. I think I would have given up with Netscape. I wouldn't have known what to do. I wouldn't have had the chutzpah to say, 'No, we haven't lost, we're still negotiating, aren't we?' And treating it as if I didn't hear their 'no.'"
  • Key take-away: Persistence pays.
  • Excite was #17 in a 17-horse race when they launched in oct. '95. But by end of '96, they were #2 in the portal game. They did it because they set that as a goal--and the company rallied around it.

Founders at work - PayPal and Hotmail

Started reading "Founders at Work." Here are some of the key take-aways from the 1st 2 chapters.

Paypal founder Max Levchin

  • Paypal iterated through 6 different ideas/business plans. (1) They started with enterprise security on Palm Pilots. No traction with customers. (2) Then they went to consumer security on Palm Pilots. No traction with customers. (3) Then storing wallet information (e.g. credit card number) on Palm Pilots. No traction with customers. (4) Then storing all of your passwords (protected by master password) on your Palm Pilot. Some small interest, but no real traction. But getting more useful. (5) Then, store money on Palm Pilots. A lot more interest, lot more useful. But mostly still the geek crowd. Created a companion website to the Palm Pilot application. Realized that a lot of people were trying to use the website for transactions, growth of that was much more than Palm Pilots. People from eBay were contacting them and asking to put Paypal logo on their site. (6) Optimized website experience, shut down the Palm Pilot app.
  • Key lesson: be willing to iterate your idea a lot. Keep testing with real customers. Each iteration should be more interesting and useful to customers.
  • Core competency became fraud detection--that's what let them beat the competition over time.
  • "Try to have a good cofounder. I think it's all about people, and, if you are doing it completely alone, it's really hard."
  • "The hallmark of a really good entrepreneur is that you're not really going to build one specific company. The goal... is you realize one day that you can't really work for anyone else. You have to start your own thing. It almost doesn't matter what that thing is. We had 6 different business plan changes, and then the last one was PayPal."
Hotmail founder Sabeer Bhatia
  • Built Hotmail initially on nights/weekends
  • Started with the idea for personal database of information, morphed into web-based email
  • Killer idea emerged because they were trying to solve a personal email exchange problem for themselves--they could access email only from 2 places (home, work). And at work, they couldn't access personal email accounts.
  • Pitched the personal database idea to VCs, because they were afraid of someone stealing the idea for web-based email
  • Didn't listen when VCs told them, "You're too young, you don't have enough experience, you're hardware guys not software guys."
  • "You have got to own the customer... OK if you don't monetize them right up front. Eventually you will be able to. But having that customer base and being able to tap into that customer base and upsell them on services, or advertise--you can always make money off them." (Note: see Google!)
  • "Don't try to change user behavior dramatically. If you are expecting people to dramatically change the way they do things, it's not going to happen. Try to make it such that it's a small change, yet an important one."
  • Genesis of the idea: "That's when it occurred to us. 'Wait a minute, we can access any website in the world through a web browser. If we made email available through the web browser, that would solve our problem.' And then it occurred to us, 'If that would solve our problem, it would solve the problems of many others.' We didn't know how many others, but email was something that everyone used."

Thursday, June 21, 2007

Made to Stick: Unexpected (2)

You've got to make your message unexpected. In addition to getting engagement (surprise), how do you create interest?
  • Mystery--start with a thought-provoking question, build suspense
  • Curiosity in Hollywood screenplays--"Curiosity is the intellectual need to answer questions and close open patterns. Story plays to this universal desire by doing the opposite, posing questions and opening situations." The audience wonders what will happen next? How will this turn out?
  • "Gap theory" of curiosity--we are compelled to fill knowledge gaps. Shift thinking from "what information do I need to convey?" to "what questions do I want my audience to ask?"

Sunday, June 17, 2007

Made to Stick - Unexpected

2 questions: How do I get people's attention? How do I keep people's attention?

Make your message unexpected.

Surprise gets our attention. Interest keeps our attention.

We're hard-wired so that surprise jolts us to attention. Our mental models have failed to predict what just happened. Surprise can make us think: Why was I surprised? What's the underlying cause? What are some other possibilities? How do I avoid surprises in the future? Surprises make us want to find the answer, to resolve the question of why were were surprised.

Some caveats--the surprise can't be a gimmick. It has to reinforce our core message, otherwise it just turns people off.
Second point: if it's just a perplexing surprise, it might end up just being frustrating to the consumer. It's got to be a surprise where the consumer gets some insight. "To be surprising, an event can't be predictable. Surprise is the opposite of predictability. But, to be satisfying, the surprise must be 'post-dictable.' The twist makes sense after you think about it, but it's not something you would have seen coming."
If you want your ideas to be stickier, break people's mental models in an unexpected way, but allow them to fix it so that it makes sense afterwards. Easiest way to do that is to have your surprise be in the service of your core message.
Here's a suggested process from the book:
1. Determine what your core message is.
2. Figure out what's counterintuitive--what are the unexpected implications of your core message? Why isn't it already happening naturally?
3. Communicate your message in a way that breaks the audience's "guessing machine" along the critical, counterintuitive dimension.

Made to Stick - Simplicity (schemas)

How do you make something complex into a simple idea that can be communicated? One technique is to leverage schemas, or mental models, that people already have in their heads.

How do you describe a "pomelo" to someone? You could go into a detailed description of how it's a large citrus fruit with a yellow, thick rind, that tastes anywhere from spicy-sweet to tangy and tart, depending on ripeness. Or you could just say it's a supersized grapefruit with a soft and thick rind. In the second example, you're leveraging the schema that we all have of a grapefruit. It's so much easier for us to grasp this new concept. We already have associations with a grapefruit, and now this concept of pomelo will inherit these associations.

People that pitch movies in Hollywood often use schemas to communicate their ideas quickly and effectively. "Speed" was pitched as "Die Hard on a bus." "13 going on 30" was pitched as "Big for girls." "Alien" was pitched as "Jaws on a spaceship."

Wednesday, June 13, 2007

Persuasive online copywriting techniques

Saw 2 interesting vids on YouTube today:

They recommended 3 techniques for making your online copywriting more persuasive:


  • Engage the right-brain, then go after the rational left-brain
  • Writing in meter (“melts in your mouth, not in your hand”; “Winston tastes good like a cigarette should”)
  • Invent words that the reader intuitively understand the meaning of
  • Use meter
  • Use verbs


  • Replace common, predictable phrases with unexpected, colorful ones


  • Choose a revealing angle—put your reader on the scene (e.g. Everest ad for Rolex)
  • Select your details carefully—include only what’s interesting, and barely that
  • Put the “known” underwater—never tell the reader something he already knows or can figure out himself

Clear use for these when you're writing copy as sales collateral for your website, but can possibly also be applied to regular copywriting on a consumer app website.

Tuesday, June 12, 2007

Test with at most 5 users

There's an interesting post from back in 2000 on Jacob Nielsen's site here. He makes a good case for why you don't need to do usability testing with any more than 5 users. This graph is illustrative, but shows how the learning from usability testing has diminishing returns.

I spoke with John Briggs, the first product manager at Yahoo, and he had mentioned that you only need to be 70% right before you ship (and iterate in the field to get to 100%). He also mentioned that you can learn everything you need to learn to get to 70% by talking to 3 users. So 2 independent expert sources are saying almost the same thing here.

Made to Stick - Simplicity

The 1st chapter in "Made to Stick" by Chip Heath and Dan Heath talks about one of the key elements of making a message stick: simplicity.

Simplicity: find the core of the idea.

Some of the ideas mentioned in this chapter:

Commander's Intent (CI): In the military, it's "a crisp, plain-talk statement that appears at the top of every order, specifying the plan's goal, and the desired end-state of an operation."

Finding the core=stripping an idea down to its most critical essence. You not only have to weed out the superfluous and the tangential, but also the things that are really important, but just aren't the most important idea. The value of the CI is its singularity.

In journalism, reporters know not to bury the lead. The first sentence of the article, called the lead, contains the most essential elements of the story. After the lead, information is presented in decreasing order of importance. This is known as the "inverted pyramid" structure. This is great for readers. No matter what the reader's attention span--whether just reading the lead or the entire article--the inverted pyramid maximizes the information gleaned. Finding the core--writing the lead--creates forced prioritization. If you can only send one line before the line is cut, what do you send?

Great simple ideas have an elegance and a utility that makes them function a lot like proverbs. Short sentences (compact) drawn from long experiences (core).

Saturday, June 2, 2007

Different content models on the Internet

There are 3 major content models on the Internet:
1. Pure-play content
2. Pure-play platform/community
3. Hybrid

On one extreme, you have the pure-play content sites. These include websites such as "," "," etc. These are destination sites that own proprietary content--the content is the main reason why people visit the site. Product functionality is secondary to the content. The website has enough functionality to meet the minimum acceptable standards, but is not known for product innovation.

On the other extreme, you have the pure-play platform/community sites. These are applications that enable users to find content in some way. For example, Google (pure-play platform) lets you search the web to find the content that you're looking for. Stumbleupon lets you discover new content through its community rating of websites and videos. Digg also lets you discover (news, video) content by its community of users that scours the web and then filters the most interesting stuff for you. These sites are all about their platforms and their communities. They don't typically own any proprietary content, but they piggyback on the good content that's already available out there on the web.

In the middle are websites that are in between pure-play content and platform/community sites. Youtube, MySpace and Flickr sit here. These sites own rich proprietary content (typically in the form of UGC), but also innovated to create compelling new product features. Their products, in addition to their content, are an integral part of the consumer experience. They built significant communities around their content.

Many of the pure-play platform/community sites also own content, but it's typically in the form of metadata (comments, votes, etc.) rather than rich, original content.

Of course, very few websites are on the extreme of pure-play content or pure-play platform/community. But it's interesting to use as a framework to think about how sites differ along these extremes.

Aside from content models, there are other models for the Internet, most notably e-commerce/marketplaces (e.g. EBay, Amazon, PayPal, Craigslist), communication (Hotmail, Gmail, AIM), infrastructure/tools (Photobucket, Rapidshare, Mapquest). It's possible that some sites (e.g. Yahoo) combine several different models under the "portal" umbrella.

If you look at the top 30 US sites on Alexa, you can make the following observations:
  • No standalone communication businesses (all absorbed by portals, part of social networks); communication businesses are among the highest-trafficked properties on the web, though not necessarily as monetizable as search; as such, they are "loss-leaders" for the portals/social networks
  • No pure-play content sites in the top 15 (CNN is the highest at #16)
  • No infrastructure/tool sites in the top 15 (Megaupload is highest at #19)
  • Of the top 10, 4 are pure-play platform/communities (Yahoo, Google, MSN, Windows Live); 4 are hybrids (MySpace, YouTube, Facebook, Wikipedia); and 2 are e-commerce/ marketplaces (Ebay, Craigslist)

The results are not really surprising. The sites that are the most trafficked are the ones that leverage the power of the Internet. It is interesting, though, that none of the top 15 sites are pure-play content sites.

The type of content sites that do make it into the top 30: news (CNN, NYTimes), weather (, movies (IMDB)