20 Apr, 2020· read

How to decide what to automate in your business

The top seven criteria to use for identifying automation projects destined for success

We’re constantly being told in the press about the benefits of automation, how it helps businesses improve their efficiency and profitability, and how everyone else is doing it and reaping the rewards.

It’s become such a hot topic that the problem for many businesses now isn’t understanding the ‘why’ when it comes to automation, but the ‘how’. According to a recent study by Gartner, over a quarter of businesses say that the major barrier to adoption is knowing what they should automate.

There are some basic principles when it comes to identifying tasks and processes suitable for automation, so we’ve pulled together a list to help you identify the best uses for it in your business.

Women And Man On Sofa@2X


There is a common misconception that automation is a new technology and should only be trusted to undertake the simplest of tasks. In fact, artificial intelligence has been used quietly for complex tasks by businesses across most sectors since the 1980s and is a surprisingly mature, tried and tested technology.

Its rising popularity over the last five years has enabled software companies to evolve the technology even further to improve the scope of what automation is capable of. As such, it is important not to limit yourself to what level of responsibility you think it can be trusted with, as you may vastly underestimate its capabilities.

Instead, ask yourself where your business experiences backlogs or delays, or which tasks struggle to get completed on time because of a lack of staff on hand to do the work, or which teams are constantly recruiting in order to cope with demand.  Then consider within those situations what ‘thinking’ or ‘actioning’ is required as part of the process and could theoretically be delegated.

Intelligent or cognitive automation is used for complex, thinking tasks, such as analysis, decision-making, communication, recognition and/or learning.

Robotic process automation (RPA) is used for simpler, computer-based tasks that require ‘actioning’. RPA uses software robots for things like mimicking the movements and clicks of a human’s computer mouse (e.g. logging into different applications, moving text, submitting forms, and/or sending emails).

If a task requires a combination of thinking and actioning, RPA and cognitive automation can be used together to complement each other – e.g. completing a thinking task and then carrying out an action based on the result.


To be a good candidate for automation, a task must have a clear, defined process. The process doesn’t necessarily have to be simple, but it does have to have a clear objective, and boundaries dictating where the task should begin and end.

It is important to appreciate that not every automated task is linear, so although the automated process may vary depending on the relevant input given, the objective and boundaries will not.

One example of this in action is the use of intelligent automation for tasks like technical troubleshooting, whereby call centre staff are given questions and prompts by the software to help them diagnose a customer’s issue, and then resolve it. There may be over 100 different customer issues that the call centre staff have to try to diagnose and resolve, but the objective is always the same – to solve the customer’s technical problem. Similarly, the boundaries do not change, as the technical staff would not try to use it to deal with billing or product upgrade enquiries.


When assessing the potential of a process or task for automation, always ask yourself if it is considered important within your business? Automation requires initial and ongoing time and financial commitments for set-up, maintenance and licensing. If the task or business process you want to automate isn’t considered to be business-critical or doesn’t offer a way of significantly increasing profitability, it is unlikely to get the internal support required to get it off the ground.

When automating for the first time, many businesses are tempted to ‘dip their toe’ and automate something inconsequential that feels low risk. However, this approach is rarely successful, simply because not enough people care about the project to secure the necessary budget and resources to support it and make it a success.


The main aim of automation is to reduce the amount of time that people collectively spend on doing a task. So, to be a good candidate for automation, an activity must be labour-intensive. The more time people spend on the task in question, the more likely it will be that the investment in automating it will be worthwhile – especially if those people are highly paid.

Many businesses find it beneficial to qualify this by calculating the total number of hours spent on the task and the average cost of those man-hours per annum. That way, they can see how long it would take to achieve a return on investment if automated. Some intelligent automation software providers and consultancies have online calculators available on their websites to help calculate this figure.


Anyone who has ever had to undertake a repetitive task for a long period of time will know how boring it can become. When humans become bored they struggle to concentrate and mistakes become more likely. In some business situations, the repercussions of making a minor miscalculation can have devastating consequences.

It is a good idea to look for elements of a job role that are repetitive and where accuracy is critical, and investigate whether this could be automated.

Automation enables humans to delegate the most repetitive elements of their jobs, rather than their whole job. This reduces the likelihood of human error and allows staff to spend their time on more rewarding and useful work. By reducing the capacity for error, an organisation’s service delivery and customer service are inevitably enhanced too, improving the business’s reputation.

One example of this in action is the increasing number of law firms that are using a combination of intelligent automation and RPA to process incoming emails – i.e. identifying what the emails relate to and moving them into the relevant client folders, forwarding or replying to them. This reduces the amount of time solicitors or attorneys are spending managing their inbox.

A common error related to repetitive tasks can be the failure to accurately follow procedure. Tasks that require certain checks or procedures to be followed can have this built into an automation process to ensure that the necessary requirements are met as a matter of course.


For automation to be cost-effective, it needs to be applied to tasks that are done several times per day. The more times humans spend doing it, the greater the return on investment will be from automating the task.

Look for activities within the business that often experience backlogs or spikes in demand, as these are often good candidates for automation.


Whilst time-sensitivity isn’t necessarily an important part of the mix when it comes to identifying processes to automate, the ability to deliver something faster and at scale, can be a game-changer for some businesses. This could be to meet industry guidelines or internal best practice for turnaround, to maintain customer satisfaction, or to keep ahead of the competition.

Want to find out more?

Get in touch