Steve Wood on how APIs will power business apps in the cloud

Steve Wood is founder and CEO of ManyWho, a San Francisco-based startup focused on making it easy for business users and developers to quickly build business applications in the cloud. Steve’s previous startup, Informavores, was acquired by Salesforce.com. The technology now drives workflow aspects of the Force.com platform. While at Salesforce.com, Steve was VP of Product Management for many of the point-and-click tools. Steve also co-authored 2 books on the Force.com platform: Unleash the Power of Force.com and Executives Guide to Force.com.

API Economist: There are a lot of different cloud offerings out there. Why do we need yet another abstraction layer for automating business processes?

Steve Wood: That's an interesting question. With ManyWho, we're doing more than just automating business processes. We're thinking about business processes as applications. For example, Salesforce CRM represents a series of business processes in sales and in service and ExactTarget in marketing. Both represent business processes embedded within those applications. Ultimately, much of the reason people use applications or build software is to automate or improve a process inside their business.

Often, when people think about business process, they think all about middleware and moving data or events from one system to another system. When we think about business processes, we're thinking about applications that help enable the business to work more effectively. We see those business processes in everything from Microsoft Access and Excel to Force.com and to the apps that you purchase. The reality is that you can't buy an application for every single business process that you have in your company. And some of these processes are what make your company unique. They often span multiple systems and take the form of applications, including sales, customer service, HR, and customer applications.

Traditionally, enterprises solved those problems in one of two ways. They've either encoded the processes in a BPM [business process management] platform, which is a very heavy way of doing things, usually only reserved for the super heavyweight processes. Or they wrote them from scratch with custom code.

To answer your question, the reason why we need yet another layer is that traditional applications and systems of record are too rigid for many of the things business is crying out for: social, mobile, real-time. They’re also often architected as closed systems to developers whereas we’ve taking an API first approach. We view everything as being an API. The same APIs and point-and-click tools that our customers use are the same we use to build our platform. This modern interface gives us the ability to allow business analysts to automate business processes in a way that hasn’t been broadly achieved to date. We believe that if we allow the business stakeholder to build the core purpose and value of the application using point-and-click tooling and allow developers full API access to radiate that into new experiences, devices, and form factors, truly stunning things will happen.  The business needs to define the beating heart of the app – they know more about their business than the development team.  Everyone is innovating – the business and IT.

API Economist:  So what enterprise apps will you target first?

Steve Wood: We want to focus on providing deep integration, rich, and beautiful integrations, with Salesforce.com, SAP, Concur, Workday, and other widely adopted cloud and on-premises applications. We’ve started with Salesforce.com of course – and we leverage all of the Salesforce services rather than creating “yet another silo” in the enterprise landscape.  For example, we tie into the database, social network, identity and file storage of these platforms. We want to keep data where it is.

The goal is to really simplify the ability to build a business process, for example, that uses an account object in Salesforce, an order object in SAP, an employee object in Workday, and an expense object in Concur. When they're building that app, it should feel like it's a unified, beautiful experience where we have deeply integrated. That’s the goal. We're not treating those applications as just another endpoint. We're treating those applications really richly inside our platform, but in a standard way. We're providing a standard app building layer on top of those applications so we’re hoping our customers won’t even use the word “integration” – we’ll feel more native than native.

We think the customers will be very successful with this approach and don't need to call up Boomi, Cast Iron, Informatica, or SnapLogic in order to solve common day integration scenarios. If we don't have one of the connectors that they need, then we'll call in the integration specialists. We’ll integrate with the integrators rather than becoming a 5,000 connector integration company.

API Economist: How has your call center background helped you solve the integration problem?

Steve Wood: From an integration perspective, call centers can be a very challenging environment. Most vendors walk in to a call center with their shiny products and show these beautiful user experiences and their awesome user interfaces. Reality quickly settles as the requests to integrate with the 50 other applications in the call center commence.

The reality is that call center integration requires sticking in an ActiveX control here and injecting a new Citrix window there. It basically requires all sorts of stuff because call centers have all these applications that need to be available to the agent on their desktop.

What we found as a vendor going in to this environment is that it’s ugly. It's terrible architecture and old technology that has no simple hooks or APIs. It's all a bundle of really botched together technology that’s evolved over time.

This is a driving problem we are tackling. So, we treat the user interface (UI) as an API, so no more of this idea that UI is something that's dictated to you by the world of HTML. We don't dictate UI to our customer applications. We describe UI to them as object in an API, and if they want to render those objects in HTML5, great. If you want to render those objects as native applications, that's great, too. If you want to render those objects in Google Glass and show them to the user visually, great. Those are all user interactions that are going beyond the scope of just standard UI. Our approach is very similar to the revolution that happened in web content management – but with apps. We’re also borrowing many ideas from the consumer social networks and re-applying them.

Once we started doing that, we realized that we were solving a core problem, namely, how to manage applications in a mobile device form factor revolution. How does a CIO think about applications in a world like this? The only way you can think about it is your applications are just APIs. If you think that your application is HTML5, you're taking a way too limited view of the world. You have to think about everything as an API. We just took that approach across the board. Everything is an API. We think that will drive enormous customer success. Server-side UI is dead – that’s old-school thinking.

API Economist: Why does social (or feed-based updates) matter in workflow?

Steve Wood: First of all, workflow is the flow or progress of work done that is defined by a series of tasks within an organization. These series of tasks usually require some form of collaboration within the organization. This collaboration acts as the backbone and by its nature is social. How do you get work done by bringing people in and collaborating with them, and how do you get work done by bringing people together and bringing systems together? When you overlap the two, something really beautiful happens. You start realizing that you can actually see how your organization is working. You add a huge amount of flexibility to workflow by allowing users to collaborate and take control of process, but you also bring the benefits of automation, which the feed-based update doesn't give you. Social networks today give you the benefit of collaboration, but they don't give you the benefit of automation.

We realized that social isn't just about collaborating in a feed. It's also about real-time and about real-time sharing, not just in terms of interacting with each other, but actually everything should be real-time and contextual. It becomes like a Google Doc where everybody who's collaborating on the document can see updates in real-time. But it’s more than just a doc, because there’s progression.  It’s not a standing object, it’s moving, evolving and adapting – it’s about reaching for and achieving goals as a team.

Also, there's no more concept of having to save a record in a database, and then everybody talks about the record after it's saved, and then somebody has to be the owner of editing it. No, everything is in real-time, across all devices, and form factors. The cloud is keeping the context of the app synchronized across all subscribers. It’s like when you read a book on Amazon’s Kindle – if you open the book on another device, it remembers the page you left off.  I mean, why wouldn’t it?

This is the next generation of business applications: social, mobile, real-time. Not only are they hugely enabling for the business, they’re expressive, natural and fun to use.

API Economist: Business process management solutions always had the dream of allowing non-technical business analysts to easily create workflow and automation without the need of involving IT. It seems like that's never really come true. Are we ever going to get there?

Steve Wood: We're banking on it, and we're going to solve that problem. The evolution of technology in this space has been about creating layers of abstraction that eliminate complexity. We want business users to be able to create their apps. It is simply not sustainable for IT to have to code every app from scratch. It's insane. I think when people look back over the next 20 years from now and say, "That was crazy. Whenever they wanted to build an app, they had to ask a developer to code it for them. Isn't that just nuts? How oppressive!"

What often held back the business was that a lot of the processes required integration. Too many of the integrators took the view of providing the lowest common denominator for their integration solution. And when you choose the lowest common denominator you wind up getting get the worst possible thing.  It’s the difference between the impact of J2ME on the mobile industry versus the impact of the iPhone.  One said: allow code to run on all devices, regardless of how bad, the other designed for the best device on the market.

API Economist: One concept that really seems to capture what’s happening with mobile devices and the app economy is the idea of “micro-moments” that represent the series of tiny, moment-by-moment interactions that happen between a mobile user and the app they are using. For example I find myself pulling my iPhone out about 18 times a day to perform a very specific task that usually lasts no more that 30-45 seconds per session. How do you think this idea of micro moments is going to impact this idea of simplifying workflow?

Steve Wood: Yes – it’s a huge shift. It's certainly affecting the way applications are designed, and it's certainly affecting the way applications are provided to users. The people who are really most affected by micro moments is actually the poor IT team, because micro moment applications have created a huge complexity for the CIO, which is, "If I'm building lots of tiny applications, I still have to manage applications with some form of IT rigor, which means I have a change management process for every app we deliver. I used to have 50 apps. I now have 500, because they're all miniature apps as opposed to one big one."

If you're a CIO thinking, "How do I manage my portfolio? My business is screaming out for these little micro-moment apps to help them in field sales, field service, to help customers submit claims for insurance, to help my employees work more effectively. They need to be connected to existing systems, they need to be social, they need to be responsive, and real-time. How am I going to do this?!”

The answer is there hasn’t been a good answer. Today's platforms were not designed with this explosion and this turmoil in mind. They were designed when one of the biggest delivery problems a CIO had was, "Is the screen 1024x768 or bigger?" Ha ha! Because they didn't have to worry about touch, they didn't have to worry about form factors. They knew it probably was going to be for a browser and that was it.

We see this as a big opportunity and workflow will be one of the key areas where these micro-moments need to be re-imagined for the mobile device.  Micro-moments equal micro processes – purpose driven, task driven, automated.

API Economist:  Now time for our favorite question here at the API Economist, Steve. What's your favorite device or devices, and what are some of your favorite mobile apps?

Steve Wood: Favorite devices are actually interesting. I think one of my favorite devices is actually one that I don't own, but I'm about to. I really like the Samsung Galaxy Note. I think it's quite nice. I love the screen size and I love what they've done with Android. I like the Galaxy Note because it's sort of a hybrid of a tablet and phone. I think I'm getting a little tired of Apple. I think they're not showing themselves to be as innovative as they once were. Microsoft has done some really neat things. I'm not entirely convinced by their OS, but nobody could argue that it's not different. It's certainly not an Apple clone, and I think we should celebrate Microsoft for having taken a bold move to try to actually show real innovation, whether they got it right or wrong. The market will decide.

Speaking of Microsoft, I actually think the Surface tablet is kind of neat. It’s a classic v1 (version 1) product and I think they've got some work to do before it'll really be anything close to the stability and usefulness that you're seeing with iPads.

In terms of apps, the one I love the most is Uber. That was probably the app that, for me, was the most useful, because it just took an existing industry and it made it customer focused. It said, "Hey, cab drivers, the customer is now in the driving seat. The customer can actually see what's going on and be much more relaxed about what's happening." I even had a bad situation where we were out in San Francisco, and I spent three hours trying to find a taxi. Now, with Uber, I can sit in the bar, enjoy another drink, and I can see the cab coming. It’s actually a great example of a well-designed micro-moment app.

I also really like Spotify. It's just a great, great service, just to be able to stream music off the cloud. Anything I want to listen to is just there. Because of Spotify, I find myself listening to much more music of different genres and styles that I didn't do so much before. I’m rediscovering my love of music.

API Economist: Steve, thank you for your time!

Steve Wood: It’s been a pleasure!