Having just come back from Mashery’s BAPI 2013 in San Francisco, there’s one word that keeps popping in my head Wow. I knew there were smart people in the world doing amazing things in the API space, and I knew to expect great things from this conference, but I was really blown away.
Most of the time at various conferences, there are select agenda items that are can’t miss, and some that you can sneak out to the lobby for some work time. At this year’s BAPI SFO, there wasn’t one presentation I wanted to miss. In fact, during some of the breakout sessions, I kind of wished I could clone myself like Jango Fett and sit in more than one at a time.
Since lucky enough to attend, I don’t want you to miss out on anything so I’ve compiled my “Best Of” summary below, in which I call out some of the key themes, or takeaways from I had during each talk. I think there are some common threads amongst the presenters, and I’ll do my best to call them out below.
The day began with opening remarks from Oren Michaels, CEO of Mashery. We share the same hairstyle, so I was happy to see him on stage proudly sporting the Bruce Willis look like I do.
What I took away from his talk was that in the fifth year of the conference, APIs are FINALLY starting to appear as a mainstream topic of how businesses can drive change, connectivity and transformation into their operations. He noted that there are more than 9,000 public APIs, and while it took over 3 years to hit the first 1,000 public APIs, the last 1,000 only took a matter of months. Adoption acceleration is happening. The tipping point, he said, was when APIs became not only read, but Read/Write.
Eric is an award-winning data visualization expert who focused his talk on finding the unmarked trail. Basically, he said to “let your data lead you to the questions, don’t take a question and turn it into data”
To illustrate, take a look at what visualizing one day worth of Nasdaq trades look like:
What questions would you ask having seen this data? If given the choice of reviewing a spreadsheet or this chart, which one gives you a better chance of catching change as it happens?
Finally, he championed the need to think about APIs differently. While they are machine language, what if we thought about APIs like novels? Could you say your API has a soul? Can it support new interactions or behaviors? What are the intentions? Is it interesting? Can I explore it? If I re-read it, will it seem as interesting as the first time around? Interesting angle to approach it….
These guys spoke at length about what it was like to run a startup, or multiple startups: challenges faced by new people-managers (removing obstacles is the chief goal), the distractions that can take away focus from actually building an innovative product, and key issues they faced as they scaled their business.
It was an interesting conversation that I found myself taking very few notes, opting instead to let myself absorb the conversation as it happened. I was impressed with the fortitude of these guys and the vision they each demonstrated as they described the passion they bring to their product on a daily basis.
This was an interesting presentation, given that Greenough categorized his group as a “rich startup” (i.e. backed by a big bank and funded with an according budget). The stakes are just as high, but the risk of catastrophe sinking the initiative is much lower.
He talked at length about Design Thinking:
Empathize – Define – Ideate – Prototype – Test
He encouraged the idea of hackathons, and suggested that as a first step, look internal for inspirations and work out the details before going external (if you do at all).
The other issue he touched on was the Golden Circle that placed “Why” at the center of every problem, followed by “How” and finally “What”. He encouraged us to start from the outside-in rather than inside-out.
His perspective: Inside-out = Conventional. Outside-in = Remarkable
His point: What is a tactical manifestation of "Why?"?
Finally, he took a moment to encourage everyone in the audience to watch this inspirational video by Simpon Sinek from aTED Talk 2010. I watched it, you should too... :)
This was an interesting presentation from the company that essentially killed Blockbuster.
What Jacobsen talked about was how they evolved their API strategy from focusing on support existing developers, partner integrations and enabling devices in 2010 to maximizing efficiencies of internal developers, optimizing systems for rapid innovation, ensuring reliability and building scalable systems in 2013.
He spoke to viewing APIs as a tactic, not a strategy, which I agree with, and how they re-assessed how their approach to bringing their product to market. He detailed and how this pivot has enabled them to focus more energy on building the right things for all rather than building hundreds of things for everyone.
Note: In 2013, Netflix supported 41 billions calls, with 2 billion of those coming daily, and over 1,000 devices. Given their shift to supporting their API rather than focusing on the hundreds of potential apps and devices, he noted that there were less than 1 million external calls made last year.
Much of this presentation focused on what Intuit was doing wrong before they got it right. In 2010, Tyson inherited the final stage of a two-year long implementation with their IT group and Oracle. Once implemented, the system was deemed inadequate and plans were laid to replace it. Yes, it made it uncomfortable for the people who designed and built the new system that was being replaced, but kudos to them for immediately acting once they realized the system wouldn’t give them what they need. Too many times people get defensive about the work, the investment, etc., leaving the end users to deal with a crippling tool.
Interestingly, Intuit uses NPS to rate the effectiveness of the API every quarter and YOY. For them, here’s the scale:
Proponents - Detractors = NPS. Their target score is +50
John Musser, the founder of ProgrammableWeb.com was up next. He spoke about the current landscape of API models compared to their state nearly a decade ago.
In 2005, there were essentially just four models:
Contrast that to 2013, where he calls out over 20 different models:
His main point was that API models are constantly changing. What you see in the picture above will look different in six months... maybe even three months!
GNIP is the largest provider of social data and social conversations. Among other platforms, they gather their data from Facebook, Twitter, Tumblr, WordPress, Instagram, YouTube, Disqus, Google+, Reddit, Intense Debate, bit.ly, Stocktwits, Flickr, Vimeo and Newsgator.
What Chris talked about can be summed up simply, and tied to the presentation by Eric Rodenbeck. The data is there, how do you see it? What would you do if you had access to it, and could react appropriately?
He shared an image of a reaction time grid, with Twitter as the “fastest” response platform, and YouTube and tumblr as the “deepest”. Mining data from such platforms can provide you with a quick read (fast), or some thoughtful comments to consider (deep).
This presentation may have been the one I found myself most interested in. Lynda was a fantastically energetic presenter, and had a tight message for the participants.
First, some stats:
They break developers into four simple groups:
Lynda also shared my favorite slide of the day… the hierarchy of developer needs. Here is the slide couresty of VisionMobile.com.
And finally, she walked through Twilio's strategy for reaching the developer audience, creating a relationship with them and earning their trust:
Dr. Weigand was one of the most engaging presenters of the day. He seemed to be a wicked smart guy who was able to provide some insight as to how Amazon views data, and what they do with it.
The short version: Amazon refines data to help you make better decisions.
He says their whole goal is to make it easy to: Contribute, Connect, Collaborate
The types of data Amazon looks at to make recommendations include:
He finished up with the observation that Amazon got credit for changing the way people think about shopping, Google got credit for changing the way people think about searching, and Facebook changed the way people think about connections.
So that’s a wrap, I hope you enjoyed the recap, and I’d encourage you to visit any of the upcoming BAPI conferences if this type of thinking gets your gears going!
Thanks for reading.
We're glad that you're excited about Best Buy's APIs. However, you're using a browser that's too old to view our new developer site. We encourage you to switch to a newer browser, preferably a recent version of Chrome, Firefox, or Internet Explorer.