One of the paradoxes that emerged during lockdown is that people working from home did more, but often found they achieved less. It's been a long-established fact that much of the time people spend at work is frittered away searching for information or organizing their day — being away from the office added to that burden, while making it harder for managers to co-ordinate their teams.
If distributed working is here to stay, enterprises need to do a far better job of remotely organizing the work that needs doing. But few of the tools they've turned to are able to give them enough of an overview to help them do this. Instead, teams have adopted a mish-mash of video conferencing, messaging, content sharing and task management tools that together provide a temporary workaround for their current situation. This has left them still in the early stages of what diginomica has mapped as the maturity model of digital teamwork, with no considered strategy for going fully digital.
Two very important things start to happen in those enterprises that do start to move up through the levels of this maturity model. First of all, they use the tools more effectively to automate much of the routine of work, freeing people's time to focus their talents on the activities where they can really make a difference. Secondly, the tools collect data about all the work that's going on, which can then be analyzed using machine learning technologies to discover the most successful teamwork patterns, and ultimately start to guide people to collaborate even more effectively.
Only a handful of work management vendors currently address this bigger picture in enterprise digital teamwork. One of the most established is Workfront, which has been developing its own maturity model focusing on the automation of work.
Speaking to industry analysts last week, Darin Patterson, who is Director, Product Management at Workfront, set out its five levels of maturity. One characteristic that it shares with diginomica's digital teamwork model is the importance of collecting and analyzing data about work. Thus the first three stages focus on automating work processes, and once there's enough data in the system, then the focus moves to automating the analysis and interpretation of that data, says Patterson.
In order for us to achieve those really exciting, highly sophisticated additional elements of automation, it'll be important for us to lay that foundation with those process-driven pieces, so that we can collect the right data at the right time and of course, enable us to deliver on those really exciting data-driven automations well into the future.
Here are the five layers.
1. Automating systems
The first step in the Workfront maturity model is the automation of the work management system itself. This level encompasses processes such as auto provisioning users and putting them in the right teams, groups and workflows to get started effectively. It also includes the automation of processes to collect or calculate data, such as validating timesheets, mapping actions against business criteria, or totting up how much time and budget is available to complete a task. A key objective here is to make it as effortless as possible to get the data into the system without people having to stop to update a record.
At one large ad agency, the Workfront system is used to do complex validation of every user's timesheet. For example, if a user enters in too many hours on a particular day, or enters non-holiday hours on a holiday, the system automatically notifies the user and helps them rectify it without any intervention from their team. This helps to speed timesheet processing and therefore revenue recognition for those hours.
2. Automating workflow
The next step is automating processes that move the work from one person or stage to the next. This helps to ensure that the right handoffs take place, whether between people or systems. For example, this might mean automatically invoking a template on receipt of a request to initiate a new project or task, which then organizes any associated content and data and notifies the relevant people to take it forward. This stage often involves integration across multiple systems or applications to ensure that the work flows through the process without requiring manual intervention to move it across system boundaries. Patterson explains:
It's about notifying the right people at the right time that work is ready. And it's about moving data from one system to another to ensure that those workflows can go seamlessly across your organisation, regardless of which systems you use.
At a media company, the Workfront system notifies the finance team of any changes to forecast publication dates. This provides all the necessary details about what changed and why, so that they can update capitalization and amortisation schedules based on the new dates.
3. Automating knowledge work
With those two elements of automation in place, the next step is to begin automating the work itself — letting machines take on some of the more predictable and repeatable elements of what a knowledge worker does. For example, in a creative marketing organization, this might mean automatically pre-populating the content and data needed to get started on the task in hand. In a development organization, it could extend to automatically compiling some of the routine code a developer will need. The point is to free up the individual's time and energy to get on with the parts of the job where they can make most impact, says Patterson.
When I'm going to work on a particular issue, I can just work on the knowledge-added elements of that quickly and effectively.
At a software vendor, the Workfront system processes all requests going into the global marketing organization. Users fill out a form stating what they need, and the system automatically identifies not only the right project template to use for each request, but also who needs to get engaged, with appropriate routing and approvals. The vendor has called this "headless project management," says Patterson, because the automation sets up projects without tying up project managers:
They're no longer focused on having project managers come in and set up all the right tasks and set up all the dependencies. Instead [they] have those processes automated, so they can spend the real time focused on making sure that they can produce the right results.
4. Automating insights
Once all of these various types of automation are in place, there's enough data available to take the next step up — applying data science and machine smarts to automatically surface insights to decision makers. In Workfront's model, examples range from automatically analyzing newly created content to determine whether it adheres to brand guidelines, to intelligently identifying and prescribing more successful work patterns, or alerting management when the metrics indicate that certain teams are about to have too much workload.
5. Automating decisioning
The final step is to let the machines take some decisions automatically. For example, if the algorithm detects that certain teams or individuals are overloaded, the system could automatically change work assignments to even out the load. Or in a marketing campaigns example, automatically select the most effective marketing material for a mailing without involving a human in taking that decision.
Workfront's maturity model is more than a theoretical construct. Some of its customers have been with the company more than a decade and its experience is that they typically start out in just one department but usage gradually spreads out over multiple years. Some have now reached a point where they are tracking work across the entire organization and have gathered significant quantities of data that helps them analyze and refine how their teams perform.
This is a long way from the ad hoc digital teamwork that most companies have suddenly embarked upon having temporarily abandoned their offices. But it provides a foretaste of the roadmap that may be ahead for all enterprises as they look to improve the effectiveness of their remote workforces.