What is Business Process Automation
Business Process Automation (BPA), sometimes Business Automation, describes the partial or complete execution of a business process by software or other technology without human intervention. Process steps are automatically triggered by upstream process steps, and decision logic is built into the process.
Business process automation is a central part of Digital Transformation.
Next to eliminating human error, it also has the advantage of speed, facilitating the handling of bigger amounts of data and being often more cost-efficient than manually executed processes.
On the downside, automation decreases flexibility and requires budget and expertise to implement.
EFFECTIVENESS FIRST, EFFICIENCY SECOND
The trend and the sense of urgency to automate processes is very strong. Sometimes, the decision to automate even feels like a reflex of an entrepreneur aiming to improve processes or scale their business.
We strongly advise, though, to first put in the time and effort to engage in Business Process Management (BPM). The goal is to make sure that the processes in your company are the right processes and do the right things (effectiveness). Only then, using automation to increase speed or output and reduce cost makes sense.
Once you are sure that your business only has the right processes (that generate value for the customer or are necessary as a supporting function), the next step is to examine which processes to automate.
What can and should be automated
Not all processes are equally suitable for automation. Intelligent automation starts with a well thought out decision on which processes to automate.
From an economic perspective, a process that generates low value for the customer but has high effort (like data-entry of travel expenses or other paperwork) is the right candidate for automation.
For most repetitive tasks, automation will be an interesting option. If a process occurs only rarely in an unchanged form, the efforts of setting up automation will offset the gains through increased efficiency.
If you create automation for a process, that process should not change too much over time. When picking possible automation candidates, ask yourself: When have I last touched or altered this process. If that was recently, maybe even multiple times, it is probably not a good idea to automate it. Try to find the correct version of the process first (effectiveness, again) before you automate.
Processes that need to handle large amounts of data or repetition are good candidates for automation. Think about marketing automation, where your interaction with an email triggers the next email.
A process that has a lot of exceptions is not an ideal candidate for automation. Look for processes where 99% or more cases are handled in the same logic.
PREDICTABLE INPUT AND OUTPUT
The input parameters of a process should be stable, and they should be mapped to outputs that don’t change much. If they’re not, an automated process will not be able to handle them.
Processes that involve complex but logical decision making can be automated. However, automation (at least the readily available tools) is in many cases not yet able to make decisions that include soft, ethical or unquantifiable factors.
You will have processes that require human intervention. These are processes that would produce worse outcomes if automated fully. Either, the human component is expected by stakeholders of the process (think about customer support), or human interactions or judgment are the very core of the process, like in strategy development or consulting.
You will be able, though to automate process steps around human interaction. An excellent example for this is workflow automation, where you automate the routing of information and process steps between human decision-makers. Or think about chat bots that prepare the chat with the right customer representative.
Again, one of the most significant advantages of automation is the predictable, provable compliance of a process with requirements.
What Technology is available
There are different approaches to automate processes. Not all of them are a good choice for smaller businesses (yet).
ROBOTIC PROCESS AUTOMATION
Robotic Process Automation describes an automation technology by which software robots carry out tasks on local computers, in local networks or on online servers. This possibility of a local approach is one of the benefits of RPA: The ability to run on on-site servers without internet connection makes RPA the choice for larger corporations who run local software and servers and don’t want sensitive data to leave company networks.
Also, RPA brings certain flexibility in terms of software involved in the automation, as it is able to work with any kind of software.
RPA in recent years has become much more approachable for smaller businesses. Implementation and technology cost has lowered, last but not least because the big technology providers offer free community editions of their software and the number of qualified implementation consultants and programmers is growing.
ARTIFICIAL INTELLIGENCE AND MACHINE LEARNING
AI and ML are critical parts of the automation development. Disciplines of AI such as Optical Character Recognition or Natural Language Processing open up the spectrum of possible input factors for automation. Machine Learning o,n the other hand h,as improved machine-made decisions based on large amounts of data.
For small businesses, the best way to benefit from these technologies is to use up-to-date software, be it related to automation or for other purposes.
CLOUD-BASED PROCESS AUTOMATION
The easiest way for cloud-based process automation is to use integrations to other apps that have been prebuilt by the software developer. They identify and cover the most common use-cases that can occur between two apps.
If there are no integrations that cover your use case, cloud-based process automation software like Zapier or Connectwise will be the best choice for smaller businesses.
It requires all data and software involved in the process to be online as well. The setup of these automation tools is, in most cases, no-code, using user interfaces that lead the user through the process. Where this is great because it’s easier to use and set up, it requires the possibility to connect to apps to be pre-built.
In a case where there is no integration or cloud-based automation, using the API of the software you use to have custom automation developed might still be a way to go. But you will require a developer for implementation and maintenance.
How does Implementation work
Let’s focus on the implementation of cloud-based digital process automation.
In the first step, you need to map out your processes as detailed as possible. Try to think about possible exceptions or edge cases, and don’t forget to involve the business users, i.e. the people that actually execute the process.
Secondly, get someone to help you build out the automation in the cloud software. You might start doing it yourself for simpler processes, but it will require quite some time to go through the learning curve and arrive at a stable and working solution for more complex automation. You can either hire someone to take care of your automation, or you bring in an external expert. If you do the latter, make sure you receive the documentation you need to maintain the automation. Also, the software firms usually provide a wealth of best practices.
Thirdly, test the automation and run it in parallel to your existing processes for a while to be confident that it works.
In the last step, implement the new solution. This requires the training of your employees as well as change management for more complex automation projects. Your team might feel that automation is taking away some of their tasks, so make sure you find the right storyline and communication (ask HR or a human resources consultant for help if necessary).
Maintenance and Responsibility
Once you have the automation up and running, you never have to worry about this particular process anymore. Errm, no wait, that’s not how it works.
The rule remains the same: No process without a process owner. Even if a process is fully automated, you will have to assign responsibility for it to someone.
Additionally, you want to make sure you have someone in your company or at a trusted partner that has the capability to technically maintain the automation.
Use the transparency that the data-driven automation gives you to improve the automation further and to measure if it actually brings you the value you hoped for. This will be valuable information for your next automation processes.
If you have questions about how automation can help your business, get in touch.