- Subscribe to RSS Feed
- Mark as New
- Mark as Read
- Bookmark
- Subscribe
- Email to a Friend
- Printer Friendly Page
- Report Inappropriate Content
- Handling Business Process Changes as Part of RPA Maintenance
- Incorporating Changes to Data Types to Your Automated Process
- Addressing Integration Changes in Your Automated Business Processes
- Ongoing Maintenance Is Key
- 4 Keys to Effective Process Maintenance
A lot of the things needed to keep our digital workers functioning well have obvious parallels with what you need to do with your human colleagues to equip them to do their jobs well, with your car to keep it running or with your house to keep it standing.
In this blog, we’re exploring “The Art of Maintenance” applied to your automated processes and how to keep them in top shape to deliver consistent impact.
So, let’s get started! 🚀
Handling Business Process Changes as Part of RPA Maintenance
Occasionally (or not, that really depends on your organization), you’ll need to evolve processes because you must take new regulatory constraints or a new initiative to improve customer experience.
Before automation, if a business process changed, you would’ve had to train your staff and follow the typical principles of change management to take the change into account.
Let’s rewind a few years back. When you used to change a business process prior to introducing automation, you’d typically:
- Plan the change
- Communicate the change
- Involve employees in the change process,
- Provide training and support,
- Measure progress and success
Now, it’s not that different with your team of digital workers and automated processes. They also need to be retrained and explained what’s effectively changing (although they don’t need to be “involved” in the change process since they won’t rebel).
Instead of enabling the team who performs the process through a training session or e-learning, have a conversation with the process owner and a process analyst (and eventually your developer) to:
- Understand the change,
- Scope the change’s impact on the process,
- Translate those into changes to the automation,
That takes us back to the importance of stakeholder engagement in the automation process.
Incorporating Changes to Data Types to Your Automated Process
Just because your overall business process doesn’t change doesn’t mean some of its components won’t.
Your automated processes ingest and process data; all of them do. So, what happens if a data type changes somewhere in the process? Probably something like filling your car with the wrong type of fuel — it breaks.
It could be a CSV file ingestion, API calls or an Interact form fill populating a work queue for a digital worker to act on. If a field that used to be a string data type becomes a number or vice-versa, you’ve got a problem. And tens of thousands of potential transactions suddenly fail.
We all know how our fellow human colleagues would react to this kind of change ; and digital workers would react in a slightly different but equally unambiguous way, which will materialize through a noticeably increasing exception rate.
There’s no miracle solution to handling this type of change. All that’s needed is change control and check gates:
- Understand when they’re changing and why
- Establish proactive “roadbooks” to ensure the transition between data types is seamless
- Introduce data validation rules
To address the risk of input changes, you can act proactively (and of course reactively).:
Addressing Integration Changes in Your Automated Business Processes
Just like all your processes ingest and process data, they also all interact with target applications – they “integrate” with them through VBOs that tell digital workers how to interact with them.
That’s probably the most common “challenge” in the automation community — when the automation team wasn’t aware of a change, and then exception rates spike up.
The application has changed, and the digital worker no longer knows how to interact with the application. That could have been a request from the process owner, operational stakeholders or your business application management team.
Once again, just like you’d most likely have to enable your human colleagues on a new interface after an application upgrade, you should do the same with your digital workers. We can probably all agree that perhaps a digital worker will be more sensitive to small changes than a human in applications that change frequently.
In essence, there’s a change and it needs to be notified. That’s standard. And not something that any process owner can expect to drop from their requirements list as they implement automation.
On this front, nothing will beat a conversation (regular, not just one-offs!) with your stakeholders (process owners or application owners). It’s just as simple as saying, “Your staff needs to be retrained when an application changes, and so do digital workers.”
Clarify That “Unannounced Application Changes” are NOT a Benefit of Automation:
If you’re just starting a new business area with automation, make it clear that application change notifications can’t be dropped from the BAU to-do list (that’s not one of the benefits of automation, so make it clear).
Make yourself heard in application CABs
Ask to be part of any relevant application CABs or make sure that application owners have a representative in your automation CAB. If that wasn’t done from the beginning, you’ll want to make sure to embed your own process adjustment SLAs into the wider process for application patching and upgrades.
Work with IT to ensure software update policies work for you
For anything that has to do with overly frequent updates (such as browsers), work with your IT team to help design the application upgrade policy so you have time to apprehend and build those into your processes. It can be tricky to handle browser updates every other week, and your business can implement or adjust their policies.
Unrelated to process aspects: We’re also introducing Smart Vision in SS&C Blue Prism 7.3, which will help you build more resilient application models and help your digital workers solve issues themselves when an application changes. This means platform maintenance activities will also help you out with process maintenance activities. If you missed it, check out our previous blog on automation platform maintenance.
Ongoing Maintenance Is Key
Let’s remember two of the “maintenance mindset” principles we discussed in the first blog of this series:
- It’s not because it works and delivers value that it will continue to do so.,
- It’s not because an automation works that it is (still) efficient.
The best way to address this is to establish a formal process re-certification cadence with your process owners from the outset , as you start scoping a new automation.
It’ll help you understand (and refresh on):
- Current process goals
- Current process success (that you should keep an eye on (and hopefully your process owner wouldn’t want for this cadence to raise dissatisfaction)
- Process ambitions / roadmap (if any)
- Opportunities for improvement
The cadence depends on the scope and criticality of the process, but we recommend leaving no longer than 12 months between two reviews , and spreading the review dates for your process portfolio throughout the year.
The Pareto principle is a good gauge for how much time you might spend on process maintenance and re-certification in a sprint versus new automations: 80% on new automations and 20% on keeping the lights on is a good starting point.
4 Keys to Effective Process Maintenance
If you had only four things to remember about process maintenance, we’d recommend they be the following:
- Stakeholder engagement and proactive conversations are the most important part of maintenance for automated processes.
- Changes to the process itself, as well as the data driving that process and anything that process is using as an integration should be taken into consideration in process maintenance.
- Digital workers need to be equipped to handle any changes that might come their way, just like human workers: the complete disappearance of “training” isn’t a benefit of automation.
- Proactive process maintenance is critical to the efficacy of automated processes – implementing fixes whenever exception rates go up isn’t sustainable, but keeping an eye on exception rates will help you pick up anything that’s unfortunately fallen through.
And why not introduce an anthem for your automation maintenance operations?
“'Cause your platform don't work and if it don't work, then it’s not platform of yours.”
