There's a lot of information (and some hype) out there about what process automation can do for your business. How do you separate fact from fiction? In this post, we'll review some of the most common truths, myths and misconceptions business owners have when they start to consider automating their own processes.
First things first; before we explore truths and myths, let's define process automation:
Process automation (also known as Robotic Process Automation, or RPA) lets us use digital technology such as software to accomplish almost any task, workflow or activity we'd otherwise have to do manually. Think of your favorite accounting, marketing automation or batch file processing software.
There are several practical uses for process automation in our daily lives, including:
- Expense tracking
- Data analysis and reports
- Complex calculations and projections
- Flagging transactions for human review or intervention
- Onboarding new employees and customers
Do a search for process automation and you'll quickly find it's a hot topic. But how do you separate truth from myth? That can be difficult, so let's take a closer look at some popular statements to see if they hold up.
1. Process automation can be rolled out as soon as the automated process is as good as a manual one.
We should maintain high standards when deciding which processes to automate, since processes we choose to automate are evaluated for reliability on a different scale versus manual methods.
Think of a process that needs a lot of oversight, but that works 95% of the time. Exceptions must be handled as they come up. While 95% might sound like a great success rate, this means that someone will have to intervene or make a correction once for every 20 times the process is run. If you approach the process manually you can intervene fairly easily, but if you fully automate the process, you'd need to use a triage system to fix issues or deal with exceptional cases. Suddenly, 95% is a bad success rate if there isn't a system in place to handle anything that goes wrong (and, Murphy's Law does apply to automation: Anything that can go wrong, will go wrong).
For a classic case of an automated process that's improved over time, we can look to handwriting recognition.
People didn't call this task machine learning back then, but many postal and delivery services began to use optical character recognition technologies to automatically recognize letters and numbers in postal codes, then sort mail into postcode bins. This way, mail could be routed efficiently. (Note: The Modified National Institute of Standards and Technology database - MNIST - holds a large number of handwritten digits used to train image processing systems. It's also popular for training and testing in machine learning.)
When the codes didn't match the digits in the database, they were sent to a worker to be processed manually. At first, accuracy was poor and many pieces had to be manually sorted - a labor-intensive task some teams outsourced to a central office by scanning the mail and sending it electronically to workers.
As time has passed, optical character recognition has improved. Incorporating better machine learning has yielded better results to the point that it needs less human intervention. But this happened over many years, as advancements were made and systems improved.
Some think the much-cited cost savings that result from automating processes happen because automation replaces the need for human workers. This simplification can prove dangerous, even outright false, depending on the process that's being automated.
Automation isn't free and it doesn't happen in isolation - there are often direct or indirect costs involved.
To say there is a lot of work involved in even partially automating a process is an understatement. Sometimes, you can get 90% of the way there, but to bridge the 10% gap to full automation may prove extremely expensive, perhaps cost-prohibitive. We suggest assessing these factors on a case-by-case basis, then making an educated judgement call on whether automating 90% of a process is good enough.
For example, consider Amazon shipping. If you're a small manufacturer that makes packing pellets and boxes, it might make more sense for you to employ workers to load them manually onto a truck than to use robots and use Amazon Prime to fill your orders. Rather than automating small parts of a process, you may opt to stick with manual labour or invest that revenue in your product.
When considering whether to partially or fully automate a process, remember that automation isn't free and it doesn't happen in isolation - there are often direct or indirect costs involved in doing it, and it's imperative to assess how much value it will bring, as well as the short-term and long-term impacts to your business overall.
Process automation has many practical applications across a wide range of industries, including marketing, finance, manufacturing, retail and others. But this doesn't mean that all processes should be automated, or that the job should be entrusted to just anyone.
If processes are low volume, require less than three people, no deadline, don't link other systems and have no need for transparency, automation is a waste of time. It could even be a wasted investment, as it would take a long time to see any ROI on these workflow items.
Choosing the wrong process to automate (or doing it incorrectly) could have dire consequences for your business and even lead to serious setbacks in productivity, operations and other areas. To avoid that, it's worth it to consult someone with demonstrated and specialized expertise in process automation. They should act as your partner, helping you navigate your way through what can be a long and confusing process, and be able to plan the best approach according to your specific requirements. Leveraging our years of experience, we can help you find efficiencies in critical areas such as internal communication, marketing, customer service or other areas, then develop and implement a precise strategy that works for you.
Automated processes will still often need human supervision. It's important to factor this in when deciding which, if any, processes to automate in your business.
This is another pervasive and stubborn myth that's often repeated by people without a nuanced understanding of the purposes or advantages of process automation. It's important to note that automation done right can absolutely solve practical problems and bring value in the form of savings on labor costs, integration of legacy systems, more efficient customer service (which allows your staff to focus on building relationships and generating revenue), and more.
But if oversights happen in the ever-important planning or implementation stages, automating your processes can actually have the opposite effect, and cost you valuable time and money, or lead to embarrassing public mishaps.
If you automate a process that you consider 100% reliable, and decrease human oversight and supervision as a result, this can go very wrong. An unfortunate consequence of automating a process is that it's possible to forget to check in on it to ensure everything is working as planned.
Take marketing as an example. Many businesses automate their social media and email marketing, but we often hear about embarrassing mistakes, such as someone forgetting to cancel scheduled messages if a horrific attack or tragedy happens in the news. Or, sometimes we receive emails with a blank [NAME] or other field.
The old adage that it takes years to build a reputation but minutes to destroy it applies here.
In addition, while improved efficiency is an essential reason to automate a process, it shouldn't be the only reason you do it.
Make a list of your most mind-numbing, repetitive and error-prone tasks that everyone dreads taking on. We bet you're thinking of data entry, but you can also add expense tracking, complex calculations and projections, onboarding new employees and customers and more to your list.
Beware of "always" and "never" statements when it comes to automation. Since so much will depend on the process you're considering automating, your industry, services and requirements, organizational structure and culture, and your overarching business strategy, it's impossible to say that you'll always see a direct cost benefit to automation. One thing people often forget to consider is that automated processes cost resources to maintain.
As we alluded to earlier, you'll likely need someone to "babysit" that process, whether you automate it or outsource (keep in mind: deciding whether to outsource entails an entirely separate list of considerations - a topic we'll explore in a future post).
Deciding whether process automation will work for you involves making a number of critical decisions that will have immediate and long-term impacts on your organization. Here are some questions to consider before moving ahead:
What am I automating?
Know the process you're automating inside and out. Define and document it.
Why am I automating it?
Without knowing why you're automating a process and what your goals are, it will be next to impossible to track ROI and discover whether it was a success.
How will automation impact our departments and employees?
How will responsibilities and division of labor change after you've automated the process? Who will oversee the newly automated process? What value will automation bring to your workforce, and what new challenges or opportunities await once the process is automated?
What costs are involved in automating this process?
Which direct and indirect costs do you need to consider when weighing whether there's a cost-benefit to automation? Don't forget: automated processes need to be maintained.
Images in this article courtesy of StartupStockPhotos from Pixabay