Facebook (FB 2.15%) is a company that is more concerned about driving long-term results to its business than about playing the game of pleasing Wall Street's short-term expectations. One testament to that belief is the company's Open Compute Project, which aims to significantly reduce the costs associated with building and running data centers by developing one of the most efficient computing infrastructures for the lowest possible cost.
Once this initiative is eventually scaled out across Facebook's entire network of data centers, its cost of revenue should begin to decline, gross profits should begin to expand, and Facebook shares today could be considered undervalued in hindsight.
Getting down to business
In 2011, when Facebook figured out a scalable way to build a data center for 24% less money and consume 38% less energy than other state-of-the art data centers, the company wasted no time in implementing these capacities. The Open Compute Project's designs have already been adopted in its Prineville, Ore., data center and in its recently opened Lulea, Sweden, data center. It will play a central role in its fourth owned data center in Altoona, Iowa, slated to go live next year.
Additionally, Facebook made the project's specifications and best practices freely available to the public in hopes that the open source community will drive new levels of innovation. While it may seem risky that any Facebook competitor can also adopt the open-compute approach to data centers, the move makes sense because it aligns with the company's purpose of making the world more open and connected. There's a good chance the community will return the favor by innovating in ways that by itself Facebook cannot.
Not only is the Open Compute Project a win for Internet services as a whole, it is also seen disrupting the conventional monolith server industry. In fact, it moved server giant Hewlett-Packard to become more energy conscious, prompting the company to introduce its Moonshot line of low-power servers.
A lesson on margins
A company's gross profit margin is really a measure of how well the organization can manage expenses and operations related to generating revenues. When tracked over a period of time, it could help investors gain insight as to whether management is effectively and efficiently utilizing its resources to generate sales.
Source: SEC filings.
For the foreseeable future, Facebook expects that as it continues building new data centers to support user growth, increased engagement, and new offerings, its cost of revenue will grow faster than sales, resulting in a decrease in gross profit margin. Ultimately, when Facebook has ample server capacity and the company begins to slow its data center construction rate, I expect gross profits to expand, largely because efficiency gains related to the Open Compute Project will be realized. CFO David Ebersman acknowledged during Facebook's second-quarter conference call that the Open Compute Project and others are "providing great returns for us and helping ensure we're able to invest our resources in a disciplined and efficient manner."
Walking the walk
The Open Compute Project demonstrates Facebook's long-term desire to manage its expenses and resources more efficiently. In other words, the company wants to see its gross profit margin grows over the long term. Given Facebook's track record of creating new business opportunities for itself, I tend to believe that management is walking the walk here and margins will ultimately improve over the next decade. In the meantime, Facebook will continue investing to set itself up for longer-term success in the form of margin expansion.
At the end of the day, this could be a surprise to investors who believe that Facebook's margins will compress over the next decade, based on the belief the company will likely need to invest more to keep growing profits. If enough investors are fooled (lower case f) by this, it could signal that Facebook shares are undervalued today.
