A common challenge for businesses of all sizes is the necessity to do more with less; to deliver greater benefit with less investment. Understandably, there is often a focus on ensuring that the business operates as efficiently as possible. Eliminating waste and re-engineering business processes so they are quicker and slicker is certainly one way of saving money whilst also continuing to deliver value to the end customer.
A logical extension of business process optimisation is the introduction of technology to automate (or semi-automate) business processes. The word “automation” often brings in images of robots in a car assembly line, but there is a parallel in just about every industry—whether multi-national or mid-size. It’s often associated with growing and successful mid-size organisations that are looking to scale up or expand their operations. In the service industry, automation often relates to the provision, processing or delivery of information rather than a physical product. For example, a call centre may provide responses to their most regular query using Interactive Voice Response (IVR) systems and recorded messages, rather than keeping the customer on hold. A bank may automate its retail loan applications process, rather than relying on manual underwriting.
However, automation is not without its pitfalls. Before any organisation considers automation, it should undertake three key activities:
- Analyse and diagnose the existing process
- Simplify it
- Collect data and benchmark the existing process
Analyse, diagnose and simplify the existing process
Before any process is automated, it’s extremely important that it is optimised and simplified. Exceptions and errors must be considered and prevented wherever possible—if you automate a faulty manual process, you’ll end up with a faulty (and possibly less flexible) automated process. Building upon our manufacturing example: If you ran a car manufacturing plant, and 50% of your output was faulty, the last thing you’d do is install robots to undertake the same process. All you’d end up with is more faulty cars!
The same is true of information processing. If you can’t get your invoices out on time, don’t assume that a new accounts package will be a silver bullet. Work to understand the existing process—where it works and where it doesn’t work—before automating. You may need to collect and analyse new data to achieve this, or you may need to utilise additional analytic capabilities to help you understand the data you already have.
It’s also important to simplify the process as much as possible. Often manual processes evolve over time; ask whether each step is actually necessary? What can be eliminated completely? Can the process be simplified and made more efficient without the need to automate?
Benchmark the existing process
Before automating, make sure you have enough data about how the existing process performs (in terms of performance, efficiency and accuracy) and ensure that you project the likely benefits that a new system will bring. Consider how often the process needs to change—find out whether this flexibility is possible once it’s automated.
Once you have this benchmark, you’ll be able to accurately measure the success of the automated solution that you have implemented. This data, combined with your team’s expertise, will drive insight into how further efficiency savings can be made.
Organisations often rush to automate their troublesome processes. It is much smarter to take a step back, analyse, diagnose, simplify and benchmark before assessing whether automation is appropriate. Ensure the data is available for you to understand current performance and demand, so that you can objectively measure success.
This post was written as part of the IBM for Midsize Business program, which provides midsize businesses with the tools, expertise and solutions they need to become engines of a smarter planet