New open data standards on the Internet need to go through a series of stages of ecosystem development in order to become long standing parts of the overall infrastructure of the Internet. By ecosystem development, I mean the growth of a network of services that create the richness and vibrancy of what the protocol will eventually become.

Roughly, every mature protocol has an ecosystem of applications and services that:

* '''Produce''' the protocol, often indirectly from end user actions
* '''Consume''' the protocol and present it to end users
* '''Aggregate''' the protocol to provide higher level functions
* '''Extend''' the protocol with a new data protocol after it has become a StableBase.

And thus the cycle continues, up and up the ziggurat.

=== Produce ===

For a protocol like RichSiteSyndication (RSS) to grow, it first needs applications that produce RSS. For RSS, the first such applications were blogs. For XMPP, Jabber and GoogleTalk. While often you'll find EarlyAdopter""s hand produce instances of the protocol, for real adoption, end users require applications that emit the protocol by themselves.

Often producers are the same as consumers. For instance, email clients needed to both produce and consume email for the protocol to be of any value. When RadioUserland came up with RSS, they added RSS production and consumption into RadioUserland so their customers could connect their blogs into a DiaryCollective. 

It's not necessary for consumers to exist for a producer culture to form. MicroFormats and the SemanticWeb are examples of data protocols being liberally emitted without a large consumer ecosystem. Sometimes an EarlyAdopter producer culture will work for years before applications are found that connect with the general population, though by no means is this a guarantee (e.g. Atom).

=== Consume ===

The next set of applications to be built are consumers that operate on the data. For RSS, feed readers were created that allowed end users to keep up with all the blogs they were interested in. For [http://en.wikipedia.org/wiki/ICalendar iCalendar], applications like iCal, Google Calendar, Microsoft Outlook all consume the standard.

As mentioned above, often the first consumers are the same as the first producers. While this is frequently enough, a protocol really becomes explosive if new applications can be built that add more power for the end users. Feed readers are an excellent example. While RadioUserland had improved its own usefulness by evolving into a DiaryCollective, by opening the data for others, it encouraged others to create higher order functionality that they never could build themselves, such as services like DayPop and YahooPipes.

=== Aggregate ===

Higher order services like DayPop and YahooPipes get their power through DataAggregation. The chief advantage of an open data protocol is that it allows third party services to aggregate huge sets of data and then search over it, compute correlations, find patterns, filter it, and so on. This dramatically increases the power of the end user as they can rapidly tap into the entire market of producers that support the data protocol. In fact, aggregation creates the market, and whomever owns the aggregator can set the terms of the economy of the market as a whole.

One would think that there is no rational reason producer applications to create space for an aggregator that can squeeze them later. However, aggregators depend on producers. It's in their best interest to ensure there is a vibrant ecosystem producing data to aggregate. Moreover, by creating the mass market, aggregators dramatically increase demand for the producer applications, which is a win-win for everybody.

=== Extend ===

The final phase of the growth of a data protocol is extension. The mass market of the data protocol effectively creates a ''platform'' (cf. StableBase), and the shear richness of tools to produce, consume, discover, and operate on data creates a gravitational mass that end users will naturally want to continue using as much as possible. Therefore, the protocol provides a basis for extensions for new, possibly orthogonal features and functionality. For instance, you can embed other XML namespaces in the RSS feed. Yahoo! Search has even experimented with presenting data represented as MicroFormat""s and SemanticWeb data embedded in existing HTML pages.

Extensions go through the same growth phrase as any open data protocol, but they start with an entrenched set of applications that can either aid or retard the pace of growth depending on the amount of investment by the vendors in the space.

CategoryNetworkStandard