(2020-02-17) Boyd Work Operating Systems No We Need Work Ecosystems

Stowe Boyd: Work Operating Systems? No, We Need Work Ecosystems. Machines are an inadequate metaphor for the future of work.

The term ‘business operating system’ (or ‘work operating system’) is getting used quite a lot these days. Most recently Monday.com, who I would consider a work management solution, declared that, no, it was, in fact, a work operating system, with their head of enterprise marketing, Oren Ezra, defining it in this way:
"A work operating system (Work OS) is a cloud-based software platform where teams build custom workflow apps. It allows teams to plan, run, and track processes, projects, and everyday work.""

While Monday.com’s management may believe it’s new release has moved the product beyond work management into a higher echelon of capability called work OS, others have used the term (or business operating system) for years. in How To Build A Business Operating System For Sustained Success, Reed Deshler defines things:
"....Perhaps most importantly, it establishes a cultural foundation so that everyone sees the organization and how it functions through the same operating lens. Taken together, the system is efficient and repeatable, and it is the catalyst for a self-sustaining way of working."

Deshler seems to be focused on industrial-age control, so that management’s objectives and strategy are translated into business operations in a predictable and efficient way. Monday is more focused on flexibility and less on control, but the two visions share a great deal in common.

In this post, I share a series of diagrams to make a case: I think that the concept of a work OS is already too limiting, here in 2020, as a means of expressing either the challenges that organizations are confronted by, or the organizational models that are emerging in the most successful and important companies, today.

The principal limitation of the work OS concept is that companies do not operate independently: they are increasingly connected to other organizations. The model of work OS is too inwardly focused, when the real leverage may come from the interactions across company boundaries. (membrane)

Making The Case

A company uses many apps, and these are increasingly in the cloud. For purposes of this story let’s just say that all apps mentioned here are cloud apps.

Apps are, in general, silos: monolithic programs managing their own data, with their own data schemata, and their own mechanisms of identity, security, and processes.

Because people would like to avoid copying information from one app to another, app makers have implemented app-to-app integrations, so a user of Slack can access information from Asana within a Slack channel. This is due to app vendors publishing APIs, and building such integrations.

The diagram above shows a related concept: LeftCo may want to share within the context provided by an app with RightCo, like shared channels on Slack or shared Google Docs. That doesn’t rely on APIs, but instead by an extension of the models of identity, security, and so on.

In recent years, because of the desire to share information across different apps technologies like Zapier and IFTTT have emerged, providing one-way integrations across apps.

We’ve more recently seen what I think of as ‘backplanes’ emerging, like Unito.io and Tray.io. These allow syncing — two-way mapping of data across applications — as well as extra-app processes, identity, and data handling. I am blurring the specifics for simplicity, but focus on the capability for backplanes to execute workflows that are not defined within the apps, but in the backplane.

Note also that Monday.com is saying that it wants to be a backplane: the place where data, identity, and processes are defined, and into which all the other apps — those that aren’t backplanes, themselves — can be connected. The tipping point for a work management tool to position itself as a backplane.

So this is a battlefield where the Mondays and Unitos of the world will be fighting: who is going to be the backplane?

This is where the story moves into where I think we are heading, or where platform-based solutions — like Amazon, or Haier — have already moved.

magine LeftCo is a management consulting company, and wants RightCo, a design firm, to assist in a marketing campaign. Since LeftCo has gone to great lengths to define a standard process for marketing campaigns and managing creative assets, they want to apply those processes. How to get RightCo integrated with that?

In a not-too-far-in-the-future world, those two companies might be using backplanes as work OSs. Perhaps they are both using a future version of Unito.io, and they could then create a cross-linkage: a backplane:backplane relationship between the two companies. Each company could run their own processes, manage their own data, run their favorite apps, and share only as much data as necessary to enable cross-company cooperation to take place.

On a different dimension of scale, LeftCo could be integrating with more providers of services and goods than just RightCo. It could be integrating with an entire constellation of dozens or hundreds of partners, following the same scenarios as before with RightCo.

But this has been skewed toward the central role that LeftCo played in the two company scenario. What was implicit is that LeftCo was really the BigCo of the pair, and it was making the rules of how things were supposed to work.

What if instead, BigCo (formerly LeftCo) published the rules of the road — in the form of standard contracts and processes — for how it would like to interact with partners.

For such a system to scale non-linearly (which BigCo wants), a technology platform that imposes governance on dozens or hundreds of ecosystem partners has to be developed.

This ecosystem evolution is more-or-less what Amazon and other platform players (like Haier, or Rent The Runway) are implementing, but they are building solutions designed solely for their specific needs, making it much more monolithic and less scalable than the ideal solutions I am handwaving at, here.

There are a number of other dimensions of innovation that could play a large role in the adoption of stacks like those envisioned here, such as AI-driven process optimization, the creation of open work marketplaces within and across companies, and the impacts of those on organizational shape and functioning. I hope to return to these themes in the future.

One last question: what is the role of management in a self-organizing and self-governed work ecosystem? Whatever the answer, it certainly won’t be what we consider management today.