"To a man with a hammer, everything looks like a nail." — Mark Twain.
Identifying which processes are worth automating should be easy, but in reality, it is not. There are thousands of processes that might work for you and just as many that might not.
With that in mind, it's important to realise we're not here to walk around the company with a 'figurative' process automation hammer. We're here to find which automated process will lead to the most significant ROI in cost-saving, operational efficiency and increased output.
The trick is to find processes that, once automated, can free up more time and resources to do higher impact work that requires a 'human' touch. For example, automating sales and marketing reporting so marketers can spend more time coming up with better creative strategies would be a good use case for implementing process automation.
So, to answer the question "How to identify the right processes to automate?" we first need to gather some intel… Step one.
What processes can we automate?
*A study by McKinsey found that current technology can fully automate less than 5% of jobs. However, around 30% of tasks within 60% of total jobs can be automated. So, we should be looking for process automation to make roles more effective and maximise our employees' productivity, more than we should be looking for process automation to replace people.
Using process automation will lead to more capacity with the same resources or will enable us to use fewer employees to achieve the same performance.
With that in mind, we're typically looking for processes that have the following characteristics
- Rules-based. Processes that can follow "if this, then that" type of logic.
- High volumes and/or repetitions with low exceptions.
- Input data that is electronically readable and follows patterns or is structured.
That being said, there are technologies, such as Decision Management tools that can deal with more complex business rules than your “if / then”, and in the same vein Document Understanding technology can take non-digital, unstructured inputs and transform them into structured inputs, ready to be ingested by task or workflow automation solutions. Our advice is to start small, and build up to these more complex automations
Step two. Does it make sense to automate?
We're back to our hammer analogy here. Just because you can automate something doesn't mean you should. For example, sending automated messages to customers just because you can, doesn't mean it's always best. Blanket or automated messages on irrelevant promotions risks lowering brand value and customer satisfaction more than increasing it. Although less efficient, it's better to have someone pick up the phone.
On the other hand, automating onboarding and making it easy for customers to get the information they need would be a great use case for process automation. As we know, user experience matters.
Step three. What is the readiness of the process to automate, and do you have buy-in?
Here we're looking at the readiness of the people and the technology they're already using. As obvious as it sounds, we often overlook how much easier it is to successfully implement automation into a process that is already using a lot of technology by people who want to automate.
Some departments/job roles will be more open to automation and more willing to make it work than others. Often departments that already use many technology tools, such as accounting or marketing, will buy-in more quickly than those who could be worried about job security, like customer service reps. And this will significantly increase the likelihood of success.
Think about processes performed by technicians that are entirely computer-based over something that is a digital-physical mix performed by people who wouldn't consider themselves "tech-savvy". E.g. automating data input for report building over physical goods packing processes from eCommerce purchases.
And finally, on the 'human' side of things, having the buy-in from key stakeholders and senior management teams is essential. But remember to get the employees affected by the automation onboard too; this can be critical to the project's success.
As for the technology readiness, here we're looking at how integrated the different systems already are. It's easier to automate processes across systems that already speak to each other than those that don't.
Step four. What's the ease of implementation?
Ease of implementation is the final part to consider when weighing up candidate processes to automate. You might have a process that, when automated, would yield much larger returns than others. However, if this is your first process automation project, we urge you to consider ease of implementation as a significant determining factor.
The Global Intelligent Automation Market Report found as high as 30–50% of initial process automation projects fail due to poor choice of processes to automate. Most often, this is because companies chose to automate processes that are too complex or have too many exception paths.
The "Rule of Five" provided by Forrester is worth paying attention to here:
1. No more than five decisions
2. No more than five applications
3. No more than 500 clicks
Another sensible approach to automating is to chunk up the work based on complexity and its relative benefit. For example, by delivering an MVP solution that say, automates 70-80% of your process throughput, will provide much quicker returns of investment. Using this approach, you can then tackle any remaining scope in future iterations.
The upside of process automation is obvious, more output with less effort and expense. Through 2021 and beyond, process automation will play a vital role in any business to stay ahead. It's hard to compete when your competitor can charge less, have larger margins with better service, just because they use process automation in the right areas…
If you'd like to discuss your potential process automation project with an expert, we'd be happy to offer you a free initial consultation. We'll help you to determine the viability of the project you have in mind or help you to identify which processes would be best to automate.