The Gong Show

I have noticed a meme in the constant conversation about revenue models for web services recently. People are proposing a version of the “freemium” business model with the following twist: a product has slightly intrusive (but contextually relevant) ads baked in that users can remove by paying a small monthly fee.

I understand the intuition of the people proposing this revenue model. They know that consumers dislike ads, yet ads are a “necessary evil” in order to make free web services sustainable, or hopefully profitable. So, they’re trying to strike a compromise with consumers that appeases both free-zealots and anti-advertising-zealots.

But, this revenue model seems silly to me. Advertisers pay a premium in order to reach people in their specific demographic with disposable income. This idea of people paying to remove ads ensures that the audience for your ads are actually CHEAPER than the average internet audience. Why? Because the people in your audience with disposable income who are willing to pay for web services are the ones that will self-select out of your audience for your ads because they are willing to pay for your product. So, all that remains in the audience for your ads are people that are too cheap to pay for your service. That doesn’t sound like the audience that Disney, Coca Cola, or even your average direct response advertiser wants to reach.

Paying to remove advertising is an interesting thought, but it’s not fully baked at this point. The real sustainable solution is to create “paid content” that your audience doesn’t actually view as “ads.” For example, the last time I tried Adblock Pro, I noticed that it didn’t remove AdWords… in other words, the creator of the ad blacklist I used saw AdWords more as content than as advertising. That’s the real home run.

I launched TwitterSnooze informally late Monday of last week, so I thought I’d write a quick post with one week worth of stats. When I first launched the script, I thought I would wait a month before writing a post about usage, but I can see now that, based on the current usage rate, an additional 3 weeks of data will be fairly insignificant. Some of this data is a little inconsistent because I didn’t stick a Google Analytics tracking snippet on the site until a little bit into Tuesday; nonetheless, the data still shows the “flash-in-the-pan” effect nicely ;)

So, in the past week:

- 250 snoozes have been successfully entered.
- 200 unique people have been snooze.
- 202 unique people have hit snooze on someone else.
- 6,926 Visits
- 6,733 Unique Visitors
- 7,801 Pageviews

Here’s a graph of uniques throughout the week:

Here’s the top 20 sources of traffic (with corresponding performance metrics… I’m still amazed that StumbleUpon can drive significant traffic):

Here’s a breakdown of repeat visitors (clearly, I’m the guy in the 15-25 segment):

Here’s a leaderboard of who has been snoozed the most:

1st — scobleizer: 17 (Scoble, your trophy is in the mail)
2nd – techcrunch: 6
Tied 4th — djchuang: 4
Tied 4th — obamanews: 4
Tied 6th — gsnail: 3
Tied 6th — jasoncalacanis: 3

… so, the remaining 194 people have been snoozed less than 2 times.

The service was launched late night on April 28th, which you can see in this graph of Snoozes/Day:

Thanks to all the people that made TwitterSnooze the “Look! SHINY!!!” of Tuesday, April 29th :)

I had an itch to build something over the weekend, so I wrote a little Twitter toy script. I call it TwitterSnooze. It allows you to hit the snooze button on your Twitter friends. That means, you stop following them for a period of time, and then automatically re-follow them X days later.

Why use TwitterSnooze…?

• It’s a good tool to avoid a blast of tweets from a conference you are not attending… just snooze the conference goers for a few days.
• It’s a nice way to get back at someone for saying something stupid… give them the silent treatment ;)
• It’s a good way to ignore someone that just flooded your timeline for no good reason… but it was just a one-time offense and doesn’t merit permanent unfollowing.

Who shouldn’t use TwitterSnooze…?

• As Dave Winer points out, TwitterSnooze is not ideal because when a person is unsnoozed, Twitter will send them an email alerting that person that you are now following them again. This is an unfortunate side-effect of the only way I know to implement a “Snooze” feature (by unfollowing and then re-following a user) given the current API. If you don’t like your snoozers getting alert emails, then TwitterSnooze is not for you.
• Security Note: TwitterSnooze stores passwords on the DB. TwitterSnooze deletes all passwords once they are no longer needed, but if the idea of your password being stored on this server makes you squeamish, then TwitterSnooze is not for you.

TwitterSnooze is inspired by a Merlin Mann post.

Update: If you do use TwitterSnooze, I highly recommend sending this someecard along with it.

I’ve noticed an increase in the number of spammy companies, random websites, and unrealistically endowed women following me on Twitter. It’s a simple form of spam: by following me on Twitter, spammers know that Twitter will send me an email with a link to their Twitter user account, which will encourage me to click and figure out who they are.

A simple feature request will remedy this problem. If the followee/follower ratio on a Twitter account is greater than 15x (ie: if a user is following 151 people, but only 10 people follow that user), then an email message should NOT be generated when that account follows someone new. So, you can still create an account that follows 10,000 people, but only the couple multiples of 15 will be bugged by an email notification.

I like this solution (and think it makes good fodder for a blog post) because it uses the implicit usage data of the service as a filter to improve the service. It’s an example of how capturing as much data as possible about users’ actions can help shape and improve a web service through iterative design.

Are you asking this question? If so, then take two and a half minutes to watch this video:

Twitter in Plain English from leelefever on Vimeo.

Fred Wilson recently wrote about the declining power of firms. Fred argues that the internet has reduced the costs of transacting with a market so significantly that the nature of firms (as outline by Ronald Coase) must change as well. I totally agree with this argument on an intuitive level. I see this trend coming.

However, as firms break into smaller pieces (or as small companies disrupt and displace large firms) what will happen to the the power of brands?

The power of brands was recently reinforced for me by a friend of mine from Stanford, Brendan O’Conner. Brendan ran an empirical test on search engine relevance. He compared the relevance of search results from Ask, Google, Yahoo, and Live by running real-world, common queries on all the engines and had subjects rate the relevance of the top five results. His methodology is spelled out in painstaking detail, so I won’t elaborate on it, but it definitely passed my sniff test. It’s important to note that branding was stripped from the results returned.

Brendan found that Google, Live, and Yahoo all performed equally well (within a margin of error) and only Ask had inferior search results relevance. Yet, this outcome is counter-intuitive, because so often you hear people say anecdotally that Google provides the best search results (thus justifying Google’s overwhelming lead in search share). So, why is Google so over-hyped? I would argue it’s because of the value of the Google brand.

Brand value is still very important online. Brand value is the reason why I only trust PayPal will my bank account data (instead of… Neteller, EasyCharge, AlertPay, etc…). I know that the PayPal’s (and eBay’s) brand, reputation, and market cap is on the line when I use PayPal. They have far more to lose if they screw up a transaction or expose my private data, and so I (perhaps naively) trust they will handle my account information with the utmost care.

Similarly, I think Google search results are the best because I associate their brand with great usability and user experience design. By contrast, I think of Live or Yahoo as a roach motels… sites designed with stickiness in mind that make me less efficient. That’s why I use Google every time.

So, returning to my original point, what happens to brands when firms fraction due to the lower transactional costs of marketplaces? Will high-value brands hold a firm together that might otherwise break apart? Or will brands become more niche (in other words, will I trust a greater number of brands to service a wider array of services)? I think brand value is a cost of engaging in a marketplace that will give large firms (with high brand equity) an edge over smaller companies.

Google announced a game-changing hosted platform for application development last night. Our initial enthusiasm at Union Square Ventures about this platform is captured by Albert over on our blog. I’m trying to tease out initial limitations in order to wrap my brain around the implications of this offering. Any insights would be much appreciated.

To get the ball rolling, the top comment on Hacker News is quite insightful. I haven’t verified all these limitations with my own research, but here’s a repost of “nickb’s” comment:

Lots of limitations:

- Python-only for now but the complete architecture is language neutral and they will add other language support (will rely on feedback on which language support to add next). Before they add a language, they need to ‘harden’ it (i.e. remove some features from it).

- Django is the only API that’s currently supported, can upload other framework(s) but you’re on your own (which will be an issue since there will be some issues with other frameworks due to other limitations)

- you cannot write to the filesystem – due to distributed nature of the system, you have no idea where the file will end up.

- you cannot open sockets! you can only use the limited API that they provide (URL & mail sending API). Forget about Twisted :(

- no threads (Google says they provide scalability in other ways)

- limits on how long an app request can run – forget about uploading large files for now.

- admin console contains version source deployment client (svn or git?) which means Google will have easy access to your source code. If you’re competing with Google, beware… Google potentially has access to your source code (and data of course) so you will have to trust them and their legal agreement! Definitely some conflict of interest is possible.

Update: Here’s the relevant info from the TOS:

By submitting, posting or displaying the Content on or through the Service you give Google a worldwide, royalty-free, and non-exclusive license to reproduce, adapt, modify, translate, publish, publicly perform, publicly display and distribute such Content for the sole purpose of enabling Google to provide you with the Service in accordance with its privacy policy.

Seems to me that there’s a lot of magic happening behind the scenes but you get simplicity for that.

Some of these limitations seem crippling, particularly the ToS.

Also, how do you create server logs if you can’t write to the filesystem (FS)? Or, how do you store any flat files at all (such as users’ uploaded content like images, audio, and video)? The limitation about “not being able to write to the filesystem” must be exaggerated… I’m guessing that users can write to some kind of file handling interface that places and retrieves flat files across Google’s distributed FS architecture… In other words, the FS is abstracted, but it’s still accessible. I’d love to get confirmation on this speculation.

What about the fundamental decision to create an insular platform that cannot be used a la carte, which is different from Amazon’s decision to provide a services architecture that can be used a la carte with other technology from other environments? It seems like Google favors reliability and scalability over flexibility, but is this choice necessary? If all a user wants is a scalable and reliable FS, and the user is willing to take on the liabilities of managing his own database, why is Google unwilling to address his needs?

How about SLAs? Without SLAs is anyone willing to build a business on top of Google’s platform? What incentive does Google have to provide SLAs that have some teeth? Amazon’s SLAs turned out to be pretty wimpy because they had no reason to create significant liabilities.

Other open questions?

While the tone of this piece may come off as bearish, that’s not my attitude at all. I’m interested in limitations because I want to figure out open opportunities and likely points of future expansion. I’m eagerly awaiting my invite, and I look forward to hosting my code on Google’s infrastructure.

This is a crime against all that is cardinal-colored. Mike Montgomery, legendary ex-Stanford basketball coach, just signed on to coach Cal! Seriously? How is this not some belated April Fools joke?

The Trent vs Monty games over the next few years are going to be epic… better than the Lute vs Monty games in the early part of this decade. I can’t wait to watch Stanford dismantle the turncoat.

What’s up with the way airfare is marketed? I clicked on a link at Orbitz that listed an American Airlines flight from NYC to London for $218. I knew the fare was too good to be true, but the result was pretty ridiculous. The airfare is $218 and after “taxes and fees” the flight costs $582. In other words, the “taxes and fees” are 168% of the base airfare. And it’s not just Orbitz… this seems to happen everywhere. Airfare marketing is so broken.