Here is my notes essay for Class 3 of CS183: Startup. Errors, omissions, and/or poor phrasing are my own. Credit for good substance and wording is Peter’s entirely. Please note that I actually missed this class (I was on my honeymoon!). Thanks to @erikpavia and @danrthompson for sending me…
Twitter has an obvious motivation to ask developers to stop making Twitter applications. It means more calls to their service which they’re paying for, and it takes users away from the official Twitter clients.
What made Twitter great was that the company listened to its users who had a better idea of what Twitter was than Twitter did. The @ mentions were invented by users and so were hashtags. Even Twitter search was made by a third party.
Even today, there are innovations going on in the community that Twitter couldn’t have predicted. For example, what about Tweetups? Building Tweetup functionality into an app is something that a 3rd party could do, but Twitter probably wouldn’t see it as an essential feature and worth building into the official client. It’s features that speak to the power users of Twitter that is going to keep users on a constant hunt for the next Twitter app. Small companies and developer groups can move much faster than a big company can and users are demanding features faster than these large companies can churn them out. Just look at Twitter for BlackBerry, it has taken them a long time to get to version 1.1 and it doesn’t even have multiple account support. They seem to be spending a lot of time on UI and small icon improvements, but they’re too scared to do anything truly innovative with the app. That’s what happens in a company of 15,000 people.
So balls to the idea of not creating more Twitter apps. We need more because the service is constantly evolving and big companies can’t keep up.
I’ve noticed a trend in business development and media buyers who want to advertise on my site lately. It’s an interesting tactic that I’ve become used to and have since developed a response for that may be of use to some other publishers. The tactic is pretty simple and it goes a little something like this:
1. Open the conversation up with a promise of a huge ad buy.
Example from email (copy and pasted):
We are interested in acquiring above-the-fold display inventory on the blackberrycool.com. We guarantee our publishers a 100% fill rate, and prepay our first time publisher partners. We are not limited by budget and can rapidly scale our spend to over $100,000/ month so long as your traffic performs for us.
2.Get specifics on how much it costs and reiterate how much money they’re willing to spend.
Example:
Once you’ve told me the rate at which I can buy the 336x280 (above-the-fold) ad spot, we can then come back to you with a set volume to begin with. Needless to say, should the traffic perform, we can rapidly increase our spend to over $100,000/ month.
3. Here is where the “trick” comes is. The media buyer now asks for a small sample buy in order to “test the water”. This is most likely all they’re actually looking to buy but they’ve used the previous emails to make you think the budget is much bigger than it really is.
Having said that, I’d like to add that we usually like to start off with a very small test before committing to a volume buy. The test is designed to understand your traffic quality and performance and for our system to learn and optimize accordingly.
To that effect, we’d like to buy about 200,000 impressions of your remnant traffic which can be spread evenly over a week’s time. Can you give me an indicative CPM at which we can make the buy at? We will pay the entire amount to you up front.
4. The media buyer will now start to haggle prices down to what is actually the budget of the ad buy. In this case, we started off with a conversation about spending upwards of $100,000 on the site, and finished with a $400 ad buy over 200,000 impressions.
We usually test for 100,000 – 200,000 impressions at $2 on most web properties. If you’d like then we can go ahead with 100,000 or 150,000 impressions as well. As I said, rates are only a question of time; the moment we see performance to justify the absolute CPM, we don’t mind paying you even higher. I’d also like to let you know that we’re looking only for long term opportunities – we can consume all your traffic from the US, Canada and UK.
In order to avoid these situations, just realize that you need to have a price in mind and anything under that price you simply won’t deal with. For me, when I calculate how much it costs me to implement and maintain an ad buy, I simply won’t do anything for less than $1,000. I’ve found this number is perfect for weeding out those who are serious and those who aren’t about buying ads.
According to Josh Elman, Twitter’s product lead, a user doesn’t become an active user until he/she follows at least 30 accounts and 1/3 of them follow back.
BusinessInsider is reporting that according to those stats, Twitter has roughly 21 million users that are following 32 accounts and being followed by at least 10 accounts.
But this doesn’t necessarily begin to explain the trouble that Twitter is facing with accounts. The biggest plague on Twitter is bot accounts who follow indescriminately. These accounts auto-follow, and through some tricky word play and marketing, grab enough users to fall into even what BusinessInsider is reporting is an active user.
I’d argue that there is an even greater subset of these 21 million accounts that shouldn’t be considered active users because they’re simply code and there is no human on the other end. According to this follow and follow-back ratio, @twBirthday is considered an “active account”.
Imagine what the total number of users would look like minus bots, spam, or people running multiple accounts. Twitter is probably a relatively small community.
Last I checked, Twitter had around 177M users but many of those are inactive accounts. Anyone who has tried to register multiple Twitter accounts knows that the username squatting is getting out of hand. Twitter announced a while ago that they would be considering canceling accounts with no activity over a 6 month period but they haven’t seemed to make good on that yet. Here is the quote from Twitter:
We encourage users to actively log in and use Twitter when they register an account. To keep your account active, be sure to log in and Tweet (i.e., post an update) within 6 months of your last update. Accounts may be permanently removed due to prolonged inactivity. Please use your account once you sign up!
One reason they probably aren’t in a rush to make good on this is that companies in the social space are often valued on a per user basis. If Twitter made good on this policy, they would assumingly have to start publishing a decline in users, as a large number of their inactive accounts begin to free up. This would look pretty bad for Twitter on paper.
Twitter probably sees account creation as something not crucial to the end user experience, but it really is. If you’re a non-trademarked individual looking to register, Twitter currently doesn’t let you grab the username, but rather suggests:
If a username you would like is being used by an account that seems inactive, you should consider selecting an available variation for your use on Twitter. In general, adding numbers, underscores, or abbreviations can help you come up with a great available username.
Anyone who is truly interested in personal brand knows that numbers and underscores look terrible and don’t work well for people trying to remember your username off the top of their head.
But it’s personal brands that are driving Twitter. A good statistic backing this up is that Twitter derives 50% of its Tweets from 20,000 elite users. Personal brand is what’s pushing the acitivity on this platform and freeing up accounts is going to foster a better brand for those who are looking to be prolific users on Twitter.
Twitter may think that user account creation, as long as it’s on the rise, is a top priority, and total account creation is the ultimate goal. But have they considered the number of potentially huge contributors to the platform who have shied away or would do better if they had access to the right username?
A billion users is great, but it doesn’t mean anything if most of them are asleep at the mouse.
Pretzil just got even more social. Now, you can connect your Twitter and Facebook accounts to Pretzil and broadcast your activity and awards to your favorite social networks. We’re adding a number of special awards based on whether you connect your accounts and we’re hoping this will further drive adoption. Stay tuned because we’ve got some huge news coming up soon.
As you probably know, the NYT has implemented a paywall scheme where you get 20 articles per month from their site for free, and then after that you must pay. The catch, is that if you follow a link from another site or social outlet such as Twitter, you can read the article and it doesn’t count towards your 20. In response to this, many people have set up websites and Twitter accounts that merely link to the NYT to give readers a free version of the newspaper.
To try and combat this, the NYT is asking Twitter to shut down accounts that are linking to articles in an attempt to give readers a free subscription. Not only that, but they’re going after people who have started websites and slapping them with lawsuits.
The whole situation reminds me of the the border between Mexico and the US. Conservatives will have you believe that building a fence will stop illegals from entering the US. This mentality is summed up in McCain’s TV ad “Complete the Danged Fence”. But the idea that a fence can keep millions of starving people who are search of money and food for their families is ridiculous. Just look at these girls who scaled the impenetrable fence in less than 18 seconds.
So just like the US has to realize that a bullshit fence is a bandaid solution to a bigger and more complex issue, so too does the NYT have to realize that their system of dropping cookies and slapping lawsuits on people isn’t going to solve their problems.
What the NYT has to do is completely rethink their technological approach to the paywall and figure out how to offer some of their content for free, and charging for additional use. There are a million freemium content models out there. Choose a good one and emulate it.
Gamification is the idea that everyone can be motivated by basic game mechanics and HighScore House uses this to motivate family members in a household to get chores and tasks completed. From the site:
You and your children decide on a set of tasks, like walking the dog or mowing the lawn, and rewards, like an extra hour of TV time or a trip to Grandma’s. Then as your children complete the tasks, they earn points that they can redeem for the rewards.
If I had a family I would love to test out this site with my kids. It would also be great to set up a leaderboard on a tv just in front of the entrance so as everyone walks into the house, they’re reminded of whose a top points earner.
Open Data Pilot Project
The Open Data Pilot is part of the Government of Canada’s commitment to open government, which is being pursued along three streams: open data, open information and open dialogue, and aims to drive innovation and economic opportunities for all Canadians.
The Open Data Pilot seeks to improve the ability of the public to find, download and use Government of Canada data. You are invited to search the catalogue, download datasets and explore the possibilities of Open Data.
The Canadian Government has committed to Open Data by launching an official site where they’ve already published 3 sets of data including:
Canada - Total entries of foreign workers by province or territory and urban area - English version
Meteorological Service of Canada (MSC) - City Page Weather
Meteorological Service of Canada (MSC) - Weather element forecasts
Hopefully the Canadian government doesn’t stop here and works hard to make as many sets of data available as possible. Weather is great and the public needs more sets of data that they can use to build useful services for Canadians. Points of Interest and maps should be a top priority for this data.
Awesome first steps, Canada.
I’ve been checking out the Great Canadian Appathon and it doesn’t seem so great. The event is set up like most hackathons or hackdays, except it comes with some pretty lame terms and conditions.
First off, the event is for Windows 7 only. A hack event should let you develop for whatever platform you feel most comfortable coding for because it will let you create the best app in the end. Microsoft has been throwing around a lot of money to get developers to code for Windows 7, and this looks like another way they’re trying to influence the dev community. Rather than force developers to code for your platform, it would make far more sense to provide devices, support and incentives for them to try it out, rather than stagnate an entire hack event.
Another important caveat for the event is that the participants will not own their code at the end. Instead, the code will be “co-owned” by XMG Studios out of Toronto who are also the judges.
A big incentive for people to come out to these hack events is that it motivates people to create products that they’ll late own and potentially turn into a business. Why should XMG Studios be allowed to own this source code? Just because they held the event? In the end, it just seems like way for them to get free work done by poor students.
The Great Canadian Appathon seems pretty contrived when compared to hack events where developers can choose whatever platform they want, and own their source code at the end of the day.
