Yesterday, Facebook released a roadshow video ahead of its IPO. The video highlights Facebooks’ opportunity to win more advertising dollars, and also points out the company’s role as a social platform. I’ve been thinking a lot about Facebook lately, so wanted to use this opportunity to jot down some notes.
First, Facebook has the potential to win a bigger chunk of the global advertising market. According to FB, brands spend about $600B on advertising per year. Global advertising spend is a slow-growth business; according to ZenithOptimedia ad spend will grow 4.8% from 2011 to 2012, which is roughly in-line with global GDP growth. Since the overall pie isn’t growing, FB needs to win ad spend from other areas, like television, radio and print. Will they be able do this? Hell yes. People are spending an increasing amount of time on Facebook, at the expense of other media. (Remember, the hours in day are static so time spent on FB equates to less time spent elsewhere.) And what is everyone doing during this time? They are watching videos, reading articles and listening to music. So without even factoring in that FB campaigns are extremely effective, FB will have no problem growing its advertising revenue.
The second opportunity is best described with a riddle: What is the only thing on the web that is unique and ubiquitous? Give up? It is your Facebook page. If you’re like me, you have three active email accounts and two phone numbers. But you only have one Facebook page. Thus, developers building social products have no choice but to integrate with Facebook.
I found this out the hard way. When we first designed Spottah, which is currently being reviewed by Apple, we decided not to use Facebook Connect. We thought integrating with Facebook would diminish our value proposition of sharing photos among only close friends and family. So we decided to connect by email or phone; after all, close friends would know your email address. We pushed a beta. It was a disaster. No one knew which email address their friend signed up with. For example, I’m registered using my @spottah.com account but everyone was trying to connect with my @gmail.com account. We looked at our sharing process and realized we needed a purely unique identifier. It became immediately obvious that Facebook has a monopoly on unique identifiers and that we would have to integrate Connect.
To my knowledge, Facebook has no plans to directly monetize Connect and Open Graph. The goal is to create an ecosystem in which independent developers create great apps that increase the value of both companies. While this makes sense, I could also see a world in which Facebook charges a small amount for Connect and Open Graph. As a developer, I would pay for this. When we started Spottah, I did not think twice about whipping out my credit card to buy computing power from Amazon, analytics from Mixpanel, and smarter email from MailChimp. But all of these are worthless without customers, which Facebook provides. Yet, customers acquired from Facebook are free.
The idea of Facebook charging developers is controversial. Many would point to Microsoft as a parallel example. Microsoft never charged its independent developers. A strong supply of programs designed for Windows meant more sales of Windows. This in turn increased the demand for PCs, which Microsoft capitalized on by grabbing the largest chunk of value chain.
Microsoft and Facebook are not parallel examples, however. Microsoft had greater control. For Microsoft, the better the ecosystem the greater the sales of PCs. Plain and simple. Facebook, on the other hand, does not have as great of control. Instagram is the perfect example. Facebook enabled Instagram’s rapid growth by improving connections and spreading the word via Newsfeed postings. In the end, Instagram became so compelling that people were visiting Instagram before Facebook, and also spending more time on the app. Realizing this, Facebook bought Instagram, proving that Instagram captured the lion’s share of the value.
Instagram will not be the last company to capture a greater portion of the value than Facebook. The key for Facebook will be designing a structure in which they foster independent development but are also able to capture their share of the value created.
Google, rather quietly, released a new Doc product called Fusion Tables. Tables is a database tool, which means that Google is now matched one-for-one with the Microsoft Office Suite. Like all Doc products, Tables is intuitive, collaborative and integrated with the web, especially location-based data. Tables is not nearly as robust as Access. But most people don't need that much power, which is why Tables is a great tool for anyone with the slightest data analysis needs.
Tables is also focused on being a data repository. In other words, if you have a dataset that you want to share with the world, simply make it "public" on Tables. And vice versa; if you need a dataset, peruse the public tables located here. Right now there are about 100 sets, including Goals at the 2010 World Cup, Coffee Production, and Homicides in Colombia.
Testing Tables: Uploading and Analyzing One Data Set
The best feature of Tables is the ability to overlay information onto a map. To familiarize myself with this functionality, I needed a rich data set that was not too cumbersome. The Bureau of Labor and Statistics is a natural, so my first test is an analysis of firm size and employment by each state.
First, I had to create the table. This could not be any easier. To create a new table you are given three options: From this computer (i.e. an Excel or Access file), From a Google Spreadsheet, or Empty Table. Since I downloaded the Excel dataset from the BLS, I chose the first option: From this computer. The next step is to confirm that the columns are correctly formatted. Amazingly, Google correctly recognized every column. For example, the first column was State, which I thought would have been tagged as text but was correctly recognized as location.
The next step is attributing the data. This is key for sharing tables with the public. If you plan on publishing the table, proper attribution is vital, or else it is of little value to the research community.
With the file uploaded and the data attributed, it was time for fun: analysis. My first stop was the Intensity Map, which is located in the Visualize drop-down. At first, the data was skewed because it included employment by firm size for all fifty states as well as the United States as a whole. Also, the data was noisy because it included every firm size, ie 0-4 employees, 5-9 employees, etc. Therefore, I needed filters, which were extremely easy to add; simply click the "options" link. Here is a screen shot of the Intensity map showing the number of companies with over 500 employees.
Merging Two Tables: Good, as long as you do not need a calculation
I ran into my first problem / bug while merging two tables together and then performing a calculation. Basically, I wanted to understand the data on a per capita basis, so I created a second table with state-by-state population going back to 2004. (I made it public here). Merging the two tables was simple. In the original table, I simply clicked on “Merge” and was presented a two question “wizard.” The first question was “which columns do you want to match.” This is equivalent to dragging the cursor between two tables in Access. In this instance, I chose to link STATE data together, so that I would see employment by state along with population. The second question is a no-brainer: which table do you want merged.
Next, I wanted calculate the Enterprises per Capita by dividing the population by the number of firms. So I went to Edit, Add Formula and wrote in my command. It didn’t work. For some reason, I could not get a merged table to perform a calculation. My work-a-round was exporting the newly merged table to Google Docs Spreadsheet, and then importing it back in as one table. When I did this, I was able to perform the calculation. This is not an acceptable solution because the table is no longer dynamic. If anyone out there knows what I’m doing wrong, please shoot me a note or write a comment below.
Conclusion: A Key Tool for Any Analyst
Aside from my inability to calculate, Google Tables is a great resource. My intensity map of firm size is just the tip of the iceberg. Check out these amazing examples.
Lastly, Tables has the potential be transformative. Humans advance on data. Right now, there is a vast expanse of data, but it is silo'd across many systems and written in different formats. A central, crowd-sourced repository of data will improve the quality of analysis, all while reducing the time wasted finding, verifying and cleansing.
I'm the Founder and CEO of Peak Support. This blog is my take on early-stage companies and innovation. Every so often, there may be a post about culture, networking, family -- you name it. After all, what is a blog if it isn't a tad bit unstructured.