A part of Darwin’s property management arm is managing a never-ending list of tasks. From service requests to leasing, every part of Darwin’s services depends on tasks that one employee would assign to another. Sending emails and delegating tasks got in the way of fixing a leaky roof.
Some quick data analysis revealed that we created approximately 400 combined tasks a week.
To better understand the root of our problems, my team chatted with associates and managers in different areas of the business such as leasing, resident experience, and vendor dispatch. Their biggest pain points were:
- Task creation takes a long time.
- Users do not understand where they fit into the larger process.
- Information within a task is not always complete or trustworthy.
- Users do not know who's performing what tasks and when.
Task creation isn’t quick or easy. Our field team must know when and how they can access the home. Our internal team needs background information and prior communications from a resident to determine if we can renew their lease.Incomplete information is costly, time-consuming, and can lead to repeat trips into the field or the renewal of a bad tenant.
Working in a startup in an evolving industry means that process is ever-changing as the business establishes its footing.
Our Ops leaders desperately want to standardize our processes but are unable to experiment with new methods without the proper digital tools created by the Product and Engineering teams.
Our Product and Engineering teams want to create a great digital tool for our Ops counterpart but are unable to without a clearly defined process.
Does Product force Ops leaders to commit to a subpar process OR does Ops force Product to waste Eng effort to create temporary tools?
On a high level, workflow creation is simple. We configure a trigger and then fill the workflow with the desired actions. The complexity came from what role we wanted our new workflow tool to play.
If the Ops team could create an end-to-end process (E2E) all within a single diagram, we can create a complex and powerful to automate everything. We imagined a tool that could place the entirety of leasing flow every market, owner, and situation into one large workflow.
After our initial designs, we gathered two major insights:
Ops is unable to map out complex E2E workflows and account for all major use cases. Ops leads were unable to create an E2E diagram of a particular domain or process because it varied by state, owner, and type of resident. There were so many exceptions. Ops could not agree on a process diagram. Any workflows that were implemented were simple and had few exceptions.
A complex tool, but powerful tool may not empower Ops to do their best work. The advanced logic and philosophy behind workflow creation was not something anyone wanted to spend time learning and doing. We found the complexity overwhelmed users They often deferred to engineers to build these workflows, which defeated our goal to empower Ops to act independently.
Changes made to the automation tool focused on making workflow logic simple- aimed at automating simple, straightforward processes within a larger business domain. For example, we would create a workflow for a set of tasks if a home is listed in Atlanta and a different workflow if the home is listed in New York. Keeping our workflows simple allowed us to use Catalyst as the arms and legs of our broader software, Origin. Small workflows:
1. Make our tool easy to use and understand
2.Reduce the damage of any accidental changes
3.Provide structure while maintaining flexibility
While our V0 contained processes hard-coded by some engineers, we saved significant amounts of time for Ops managers. The Ops team was extremely excited- eager to use the tool, provide feedback, and save time. Now that their work was automatically delegated, some managers were able to spend up to 50% of their day triaging fires and real emergencies.
