5 digital enterprise trends in the future of coaching networks (2/2)

Enterprise software is undergoing a digital transformation that dramatically changes what businesses can achieve with today's information technology. That much we know — but where's it all heading? As I explained in a previous post, leading VC Gordon Ritter believes the big new thing in enterprise software is coaching networks, which harness AI and collaboration in an iterative feedback loop to help people achieve better outcomes.

Coaching networks build on a number of other developments that have been gathering pace in the past few years. In fact, I feel as though this concept is the one missing piece of the puzzle that completes the picture of what comes next. It joins together five game-changing trends I've personally tracked for a while in the ongoing digital transformation of enterprise applications, and of the organizations that use them.

In this post I'll step through these five trends as I piece together how they underpin coaching networks, starting at the business level and then working down into the technology. This is an architecture stack that describes business organization and operations as much as the technology infrastructure that lies beneath them.

1. Frictionless enterprise and the move beyond paper

The first layer is all about organization and business process. As long ago as 2011, I set out the notion of frictionless enterprise as a way of thinking about how connected digital technology would transform business. I've since arrived at this definition of frictionless enterprise:

Frictionless enterprise is a business architecture that optimizes the use of connected digital technologies to strip out cost, delay and opacity when harnessing resources and delivering outcomes. Simply put, it erases the barriers that get in the way of getting things done.

Along the way, I came to realize that one of the central tenets of frictionless enterprise must always be the elimination of paper — not just the use of paper to transmit information within an enterprise, but also all of the systems and processes that grew up around it:

The traditional enterprise is ... structured around sequential processes designed to pass static documents internally from one functional department to another, carrying with them the information they need. Frictionless enterprise is structured around dynamic processes that connect digitally networked content, resources and participants, often criss-crossing organizational boundaries.

Implicit within this is the elimination of enterprise applications that, even today, are still built on a forms-and-database model of paper-based information capture and exchange. There's a good deal more to say about frictionless enterprise — including a lot of learnings about XaaS, aka everything-as-a-service — but for our purposes today, we must focus in on this core tenet of replacing all legacy vestiges of paper-based processes with something more suitable for the digital enterprise. With their emphasis on digital interaction with human behavior, coaching networks rise above that paper-based legacy.

2. In a connected digital world, collaboration rules

In parallel with my thinking about frictionless enterprise, I've long argued that collaboration will evolve to become one of the core application pillars of the digital enterprise. The tortuous evolution of digital collaboration tools — and how enterprises deploy them — has been a saga that I've followed closely.

It's not done yet. There's still a confusing plethora of collaboration solutions out there, from messaging and content sharing to workflow automation and team workspaces. My current stance is that enterprises are going to settle on a portfolio of tools or an ecosystem platform that anchors digital teamwork in a collaborative canvas:

A flexible, connected framework for participants to digitally share, organize, track and progress the work of a team.

Within a coaching network scenario, this becomes a focal point for applying machine intelligence to those interactions.

3. Conversational computing and headless apps

Alongside this ongoing evolution of digital collaboration, a new development in the past year or so has been the evolution of conversational computing:

Thanks to the rise of AI-powered voice interfaces and messaging chatbots, conversation is becoming the new frontier of how people interact with computing.

This has two separate effects that are both huge enablers for coaching networks.

First of all, it converges two previously separate domains — how people converse with each other, and how they interact with computers.

The computers therefore have to learn how to understand people, whether by listening to their voices, interpreting their messages, watching video, or collecting sensor data from around them. This adds a huge volume of new data that was never previously available digitally. At the same time, computers can use speech and messaging to guide people in the course of what they're already doing, instead of having them turn away to type at a keyboard or point and click on a screen.

In my previous post, I cited some portfolio companies of Ritter's Emergence Capital, but some of the principles of coaching networks he describes can also be glimpsed in work that software giant Oracle has been doing with its enterprise customers. Sri Ramanathan, Group VP of Mobile and AI Bots at Oracle, recently told me how the company is replacing mobile apps with chatbots that learn from your behavior and personalize their response:

A well-designed bot, he says, is 'one that over time you interact with less and less' — while at the same time, behind the scenes, it is progressively doing more and more for you.

This is a great example of the iterative feedback loop between human behavior and machine learning that gradually produces more effective behavior over time. The one ingredient from Ritter's model that's not explicitly present is the interplay in the iterative process between human creativity and machine learning.

The second effect of conversational computing is that the underlying applications become 'headless'. They no longer deliver their output in a predefined format. The impact on application architectures can be seen from the content management field, where similar factors are driving the rise of 'headless CMS':

[T]he bigger picture here is the separation of enterprise application and content infrastructure into three separate layers — resources, functions and experience.

It's in the experience layer of conversational computing and digital collaboration where coaching networks work their distinctive feedback loop between humans and machines. The two underlying layers of functions and resources, each accessed via APIs, are equally important in making this possible.

4. Loosely coupled APIs

Enterprise IT has been on an extremely long journey to break down once monolithic architectures into a much more flexible model where every function or resource can be called as an API. This was a core part of my own vision for cloud computing as early as 1999 — and it's taken nearly two decades to get to a point where notions such as functions-as-a-service and drag-and-drop data connections are now commonplace.

Over that time I've had numerous conversations with MuleSoft founder Ross Mason, and it's his company's 3-layer API model that I've adapted when I talk about engagement, functions and resources, although MuleSoft's layers are called experience, process and system:

At the foundation layer, system-level APIs communicate with enterprise applications, datastores and other IT resources. The next layer consists of process-level APIs, which add functionality that works with the system APIs. Finally there’s a layer of engagement APIs which provide the user experience. These three layers form the building blocks of what Mason calls the enterprise application network.

This is important because once the enterprise IT infrastructure has been broken down into these three sets of easily composable building blocks, it becomes possible to refactor the components in new ways — especially, Mason told me when we met last month, in the middle process or functional layer, where "you can actually refactor a lot."

I'd already noticed this effect last year in early demonstrations of conversational computing and headless applications by the likes of Apttus and Infor:

In all of these examples, the traditional bundle of functionality that makes up an enterprise application has been broken down into separate components that are then recombined in new ways to provide a different, more streamlined outcome that wasn’t possible without the new technology. This is a phenomenon known to economists as unbundling and rebundling and it’s invariably a harbinger of disruptive innovation in a given field as new patterns of consumption become possible.

All of this brings us back to the frictionless enterprise mantra of stripping out barriers to getting things done. The combination of conversational computing, headless applications and API-centric IT infrastructure enables refactoring of the enterprise applications landscape on a hitherto unimaginable scale. It's the big step that allows us to break free of the longstanding forms-plus-database model of traditional enterprise computing and introduce a new outcomes-focused paradigm that's digitally native from the ground up. We will still need a transaction store, but it will no longer be the design center of enterprise applications and may even end up based on some new distributed digital technology such as blockchain.

It just needs a catalyst to provide the spark for this transformation — and that's exactly what the addition of AI to the engagement layer provides in the emergence of coaching networks, by providing an engine to accelerate competitive advantage.

5. AI as a tool for augmentation

Like many of my diginomica colleagues, while I'm impressed by recent advances in machine learning and artificial intelligence, I'm also conscious of its limitations. I don't see AI replacing humans anytime soon — though we need to be wary of the unintended consequences of reckless programming.

The value of AI is as a tool that augments the capabilities of humans, and that's the beauty of the coaching networks model. It aims to harness the unique capabilities of both machines and humans and get them working in harmony to produce better outcomes. It's an elegant model in itself, but I'm equally impressed by how perfectly it builds on all these other trends I've been tracking over the past several years, cutting through all the old barriers to harness the raw potential of an organization's data, connections and people.

This has been a necessarily brief canter through quite a complex series of ideas but I hope it gives some sense of why I feel this notion of coaching networks is a concept whose time has come. It will take quite a while to realize the full extent of its potential. But in view of the way that it builds on what's gone before, I hope I've explained enough to show that it's not at all far-fetched to think of coaching networks as the launchpad for an entirely new generation of enterprise software. Expect to read much more on this topic in the months and years to come.