First and foremost, if all you’re doing is releasing APIs, you can do whatever you want and developers won’t care.
For our latest “What I’ve Learned…” interview at OpenChannel, we sat down with Dr. Anant Jhingran, who leads Products for API Management at Google, helping enterprises build an API driven ecosystem and platform strategy. Prior he was CTO at Apigee (acquired by Google) and VP and CTO for the Information Management Division at IBM, including a co-chairmanship of IBM’s Cloud Computing Architecture Board. He earned a PhD in Computer Science from the University of California, Berkeley, holds over a dozen patents, has authored over 20 technical papers and serves on the board of Persistent Systems. A self professed data geek who has grown to love middleware, he sat down with us to share what he’s learned along the way…
OpenChannel: You’ve said “it’s easier to understand your business needs to operate like a software business than it is do anything about it”. Even though companies know they need to change, why do so many companies struggle to act?
Anant Jhingran: I follow the philosophy of “what got you here, won’t necessarily get you there”. Many enterprises lack competence in this new software driven world. Their competence may be in putting up hotels, setting up drug stores or doing pharmaceutical research. Whatever it is, it’s not in building awesome software products, connecting software products or leveraging software to build an ecosystem.
The challenge is that the software skill set required for future excellence isn’t the current skill set of most enterprises. It’s a dangerous gap that can result in serious issues.
Now, everybody kind of realizes this and says they’ve got to do something. Companies know they need to be more software driven, have APIs and embrace 3rd party developers, but going from realizing to actually doing it requires a real change in mindset.
OC: Then once a company decides to change, what framework do you use for building a platform strategy around APIs and 3rd party developers?
Anant: The fundamental questions is are you trying to build a platform on your own, or are you connecting with another platform and ecosystem? We call this either building a platform or connecting with other platforms.
Being a platform versus having a platform strategy are different. When building a platform, getting the supply side of developers is hard, getting the software right is hard and getting the demands story is hard. Basically, it ain’t easy.
While you might want to build your own platform, you can participate in other ecosystems too. There’s a wonderful quote “sometimes you’ll surrender some to get more”.
An example of a company that has done this successfully is Magazine Luiza, an Apigee customer. They built their own platform and have gone from supporting about 70,000 of their own skews to over 1.3 million skews by allowing 3rd parties to sell through them.
They also noticed that a lot of customers hang around the Facebook ecosystem and realized the value in having Magazine Luiza capabilities appear in Facebook.
They set up an affiliate model where you can set up a mini Magazine Luiza store within Facebook, sell to your friends and keep a portion of the revenue. Magazine Luiza, has effectively figured out that one size does not fit all.
Sometimes it’s good to be a platform where people come to you and you are the center of the universe. Other times you realize you can make your capabilities appear in other ecosystems. Great companies understand that both of those can be valuable.
Anant: Exactly, they are building their own ecosystem and also selling their app into other ecosystems. We tell customers about the pros and cons. They will keep a larger fraction of the total platform they create but it takes investment.
If you want to pursue a platform strategy, you have to go in with eyes wide open and be prepared to do three things right. Invest heavily, focus on your strengths, and don’t be late. You especially can’t be late to the market or network effects will kill you.
Another example is Walgreens, also an Apigee customer. Their core asset is their physical presence and network of walk in stores and they wanted to attract a digital audience to their physical presence. They started with photo printing capability because photo printing capabilities at Walgreens are way in the back. When you go and pick up your photos, it may only be $8.00 dollars but you end up buying more since you have to walk through the whole store.
They created an API around their photo printing capabilities which is very logical but then realized one more thing. People are not necessarily going to come to the Walgreens app to actually print photos. They’re going to use Shutterfly or whatever their favorite app is. Instead of trying to bring consumers to them all the time, they decided to leverage developers to connect the popular ecosystems of consumers to their supply of Walgreens APIs.
The most important thing for them is the connective tissue of developers. Developers who build awesome apps is ultimately what brings consumers to the Walgreens ecosystem.
OC: You said “you can’t be late” but can a company pursue a platform strategy too early? How do you know when it’s the right time to build a platform and developer ecosystem?
Anant: That’s a great point. Apigee customers are large enterprises where being too early is less of a risk. We’re trying to kind of shake them up to make sure they’re early enough.
That said, I’m not saying our customers can’t move quickly but more often they are very deliberate in what they do. Their deliberation is good but overthinking and waiting for strategic alignment across different organizations may actually make it too late to get into the market.
OC: What are the most successful tactics when building a platform strategy and developer ecosystem around an API?
Anant: First and foremost, if all you’re doing is releasing APIs, you can do whatever you want and developers won’t care. You must have quality APIs and that means two things.
- It has to offer something of value. Just because you think your APIs are good doesn’t mean they are. You have to make sure somebody cares about the value your providing and understand why they care.
- You need to change your focus from exposure to consumption. Companies bring up an interface and say “Here is a smorgasbord of capabilities we provide”, but developers are actually looking for easy consumption as opposed to a complete set of capabilities.
To help solve this, we tell our customers to devote a product manager to the API and treat their APIs as products as opposed to just an exposure for services. It’s very important to do this. We want to enable that type of product thinking in the context of APIs.
OC: How do you actually attract developers to those APIs?
Anant: Many people know that developers don’t like to be marketed to, but developers do like convenience.
When we as developers write code, we read blogs, look at peer reviews and competitors and hang around places like AWS, GCP and Firebase. If I’m looking for a Shipping API for example, the first thing I’ll do is search “Shipping API”.
I’m not a huge fan of hackathons. They are great and all but don’t really result in much value.
You’ve got to be present where developers are. Our customers buy AdWords around APIs so that they can be visible. They do promotions for their organic search and syndicate their APIs.
They also may need to solve the problem of procurement. Sometimes a company that is trying to sell its API service is not an approved vendor for a developer in a large enterprise to build against.
The company selling its API may not be known for software selling and instead known for selling a traditional products. They need to ensure that developers who are building an app are actually able to use the API.
With some small steps you may not have to fight a procurement team. For example, if the potential user of the API already pays one bill and settles through to AWS, GCP or whomever, you can create a relationship with them and work alongside the larger platforms.
The central idea is that don’t do overt marketing. We’ve gotten good at developer relations and we bring some of those best expertise to our customers who don’t necessarily understand developers as well as Google and Apigee.
Walgreens has been very successful. They hired amazing folks who write code, understand how developers feel and then go out and evangelize.
OC: What is the most important metric to track when building an ecosystem of developers around an API?
Anant: There are 3 because one size does not necessarily fit all.
The first is number of developers and number of developer apps. It’s the healthiest measure for a company. The number of developers, apps, and ultimately the number of end users of those apps is really important.
The second rather obviously one is API traffic. It’s the easiest to measure but only an “ok” proxy for value that’s being created with the API
The third is the revenue you’re driving through that API. What happens is that a lot of people confuse revenue through API as revenue for the API. In a lot of cases the API is just a channel to apply revenue to what a company already does at the core.
For revenue you have to be very clear on whether you are in the business of making money with your API. That’s why many of our customers, for example Pitney Bowes, track the revenue AND influence through their APIs.
I’m biased to the first and the third metric. Our customers will mix and match depending on their business, and sometimes get caught up in the “vanity metrics” as well 🙂
Pursue Your Platform Strategy!
Want more from platform strategy leaders at Shopify, Intuit and Box? Check out the “What I’ve Learned…” series.