What is a Positive Network Effect?

In order to understand the ways in which network effects help drive Apple’s sales, we must first define the term "positive network effect". In the book "Web 2.0: A strategy Guide", Amy Shuen describes positive network effects in the following way:

"Positive network effects increase the value of a good or service as more people use or adopt it. The simplest network effects are direct: increases in usage lead to direct increases in the value of the system. Telephone services is a great example of this. The more people available to call, the more valuable the system becomes."

In other words, users create value by joining the system. And as the system becomes more valuable, more users are motivated to join. Thus creating a positive feedback loop where adoption rates climb steeply.

Apple has done a good job of harnessing the power of network effects by creating products and services which make these positive feedback loops possible.

Type One: Direct Network Effects

Direct network effects occur within a single group of people, for example: end-users. Direct network effects have been created in Apple’s iOS environment by services such as iMessage, Facetime, and Game Center. Each of these services become more valuable as more users adopt them. iMessage in particular has the potential to drive future iPhone growth, much in the same way that RIM’s BlackBerry Messenger(BBM) helped drive sales of BlackBerry handsets in the mid-2000s. As more users own iMessage compatible devices (iPhones and iPads), the value of owning these devices grow.

Just as increasing numbers of users within a system adds value, decreasing numbers of users within a system removes value. Positive network effects can rapidly increase a system’s value, negative network effects can rapidly decrease that value. This isbad news for a company such as RIM which should expect to see it’s user-base erode at an increasingly rapid pace in markets where it directly competes with the iPhone. As BlackBerry users continue to see their BBM lists dwindle, the incentives to move to iOS increase.

Type Two: 2-Sided Network Effects

2-sided network effects occur between two distinct groups of people, for example: end-users and developers. An increase in the number of one group of people leads to increased value for the other group. Game consoles such as XBOX are a great example of 2-sided networks. As the number of people who own XBOXs rise, the value for XBOX game developers also rises because there are more potential customers to purchase their games. These two distinct groups, XBOX owners and game developers, come together to create positive network effects, increasing value for each other.

Apple harnesses the power of 2-sided network effects via their App store. The App Store brings together two distinct groups: iPhone/iPad owners and developers. As the number of iPhone/iPad owners increase, the value to developers also increases, leading to an increase of developers entering the App Store market. More developers leads to an increase in available apps which creates more value for iPhone/iPad users.

Type Three: Intra-Personal Network Effects

The final type of network effect is what I’ve termed intra-personal network effects. These are the network effects that occur within a person’s household when they buy into the Apple ecosystem.

As an individual, buying more Apple devices adds value to existing Apple devices because of the way data is shared between various devices. AirPlay and iCloud are the primary methods Apple uses to share data between devices. iCloud allows users to sync their data across multiple devices. For example, a document created on your Mac is automatically synced to your iPad and iPhone. Therefore, your iPhone becomes more valuable if you also own a Mac or iPad. In other words, the more Apple devices you own, the more valuable each of those devices become.

AirPlay allows Apple users to stream content such as videos, music and photos from their Mac to their iPad, iPhone, and Apple TV. This ability to share content across Apple devices further incentivizes people to buy into the Apple ecosystem.

Conclusion

With even tighter integration between iOS and OS X enabled by services such as iCloud, Apple has placed a big bet on the ability of network effects to drive future growth.

Taxonomy Problem:

The "My Documents" folder in Windows XP provides us with an important lesson in how a misguided informational taxonomy can cause major usability problems. As expected, documents (Word, PDF, Excel files, etc) are found within the "My Documents" folder. However, folders entitled "My Music", "My Pictures", "My Videos", and "My Downloads" are also found within the "My Documents" folder. This creates a parent-child relationship between "My Documents" and those sub-folders. This is problematic because videos, music, and pictures are not types of documents. Therefore, there should not be a parent/child relationship between those folders and "My Documents".

Collation Problem:

The second problem is related to collation. Collation is the organization of information (in this case folders and files) into a meaningful order. When viewing lists of folders on a computer the most common type of collation that people utilize is alphabetization. Because the word "My" is appended to the front of the folder names, users are unable to utilize alphabetization to find the folder they’re searching for. For example, the video folder wouldn’t be found listed with other folders that start with the letter "V", rather it would be found under the "M" section. From a user’s perspective, this is unexpected. Unexpected behaviour is very often the cause of usability problems.

The solutions:

The solution to both of these problems is simple. Placing the folders for documents, videos, downloads, and pictures on the same level will solve the taxonomy problem because it will remove the illogical parent-child relationship that existed between "My Documents" and its sub-folders.

Removing the word "My" from the front of the folder names will solve the collation problem because it will allow users to find folders based on the alphabetization of folder names. As you can see in the screenshot below, the software developers of Mac OS X have named its folders correctly.

Unfortunately, these content vacuums traditionally occurred shortly after the initial television broadcast, when consumer interest was highest. Obviously this wasn’t a problem before broadband internet. However, with the widespread adoption of broadband, consumers are now able to stream or download content at acceptable speeds which makes illegally distributing and consuming content much easier.

Monetizing the entire lifecycle

1: Television Broadcast Phase
This is the phase that we’re all familiar with. Networks broadcast their programs on TV which generates revenue through the sale of on-air advertisements or via subscriptions (e.g. HBO).

2: On-line Broadcast Phase
This new phase will occupy the space between the original television broadcast and the DVD sales phase. Piracy activity has traditionally been highest here. Due to the fact the content wasn’t legally available, pirates moved to fill this content vacuum.

Networks must continue to move into this space and create revenue by streaming their content via the web (YouTube, Hulu, etc.) or selling content via programs like iTunes. Revenue during this phase can be ad based, pay-per-view based, pay-per-download based, or subscription based. It’s important to remember that the latter three options require a low-barrier ‘one-click’ payment method similar to the iTunes’s music store’s process in order to facilitate transactions.

Obviously this will not eliminate piracy. However, these activities will reduce piracy and monetize a phase of the product lifecycle that was previously ignored.

3: DVD sales Phase
DVD sales should begin shortly after the television program’s season is complete. It is important to note that any free on-line broadcasts of the same content (eg. on YouTube or Hulu) should be stopped. Free online content will cannibalize DVD sales if the ‘on-line broadcast’ phase and the ‘DVD sales’ phase are allowed to overlap.

Conclusion

Television networks are experts at monetizing the ‘television broadcast’ and ‘DVD sales’ phases of the content lifecycle. Embracing technology such as video streaming and developing in-house or partnered (i.e. YouTube) delivery services will allow customers to legally consume content. That, coupled with a streamlined content purchasing system similar to iTunes or an ad based revenue model, will allow television networks to monetize a previously untapped revenue source.

Update (February 2012): The Oatmeal posted a cartoon which illustrates the problem with content vacuums perfectly.

Just a quick observation…

Back in the old days, computer processing was centralized in large mainframe computers. "Dumb" terminals with very little processing power were used to access the mainframe and display data to the end user. As time passed, the mainframe/terminal configuration was eventually replaced by PCs that were able to handle complex processing.

With the push towards mobile computing, history is now repeating itself. Google Chrome OS, which is designed for low-powered netbooks, is based on centralizing processing and data storage in the cloud. Removing the burden of processing from netbooks will allow these computers to run low-power, energy efficient CPUs. Additionally, storing data in the cloud, rather than on the netbook itself, allows these computers to keep a small form factor which aids in mobility.

Example:

<div id="sideColumn">
...
...
...
</div><!-- #sideColumn ends -->

In 2007… [Google] began offering 800-GOOG-411, a free directory assistance service that interprets spoken requests. It allowed Google to collect the voices of millions of people so it could get better at recognizing spoken English.

A year later, Google released a search-by-voice system that was as good as those that took other companies years to build.

Goldman Sachs reduced sales estimates from 3.5 million units in 2010 to a mere 1 million, and only 2 million in 2011 if they launch a second phone, do a lot more marketing themselves, and decide to sell it in retail stores too.

UPDATE: 74 days after its launch, 135,000 Nexus One devices have been sold. By comparison, after an initial 74 days the Motorola Droid sold 1.05 million and the iPhone sold 1.0 million.

Here are the steps for creating your own burndown chart:

1. Capture user stories.
2. Assign high-level effort estimates to the user stories.
3. Organize and prioratize the user stories into iterations.
4. Break each user story up into tasks.
5. Create your your burndown chart.

1. Capture user stories

First off, you’ll have to work with the stakeholders to capture the business requirements of the project in the form of user stories. User stories are a great way of capturing requirements because they allow the stakeholders to express their requirements in non-technical business language. An example of a user story would be:

The ability for customers to purchase widgets on-line using their credit card.

Work with your stakeholders to prioritize each user story. Label each user story in sequence using numbers.

2. Assign high-level effort estimates to the user stories

Work with your developers to assign high-level estimates against each user story.

3. Organize and prioritize the user stories into iterations

Organize your user stories into iterations. Our iterations will be 30 days in duration (about one calendar month). In this example, we’ll aim to have approximately 40 days of work within each iteration.

4. Break each user story up into tasks

Now it’s time to work with your developers to break each user story up into tasks and apply effort estimates to each task. It’s most likely that your effort estimates at the task level will not match the effort estimates at the user story level. This is okay, the user story estimates were used to roughly divide the work into iterations. These new task-level estimates are more accurate.

5. Create the burndown chart

Now add up the effort for all the tasks in your iteration. Also, take a look at a calendar and add up all the working days you have during the duration of your iteration. Recall that our iteration is 30 days, however not all of those days are working days due to weekends and holidays.

At the end up this step you’ll have two numbers:

Work (effort): 43 days

Time: 20 working days

Now it’s time to start creating your burndown chart. On the Y-axis we’ll place our work remaining in days. On the X-axis we’ll place our time remaining in days. Each day we’ll track our progress by placing a dot in the appropriate place. Once a task is complete, we can remove it from our task list and reduce the amount of work remaining.

You’ll notice we’re not tracking a % complete. A task is either complete or not complete. This eliminates the problem of tasks getting stuck at 90% complete forever.

Meet with your team each morning to update the burndown chart.

A burndown chart will quickly show whether or not you’re on track or behind schedule. If you’re plotting above the line you’re behind schedule. If you’re plotting below the line you’re ahead of schedule. This allows you to accurately report project status and allows you to identify potential schedule problems quickly.