Intelligent Automation Pitfalls You Need to Avoid When it Comes to Implementation

Now that RPA and intelligent automation are available to almost any business, here are the essentials when it comes to planning, resistance and pitfall avoidance.

1. Don’t go all in without a plan

Process automation, as with any other change to the business, should be planned correctly. While the demo or trial may have gone without a hitch, broader scale implementation requires leadership, departments and workers to be informed and on-side every step of the way.

The plan needs to cover every detail of the RPA from data flow to individual keystrokes performed by human operators, while expected results, down to every second and dollar saved, must be planned out across the life of the RPA.

To maximize the opportunity for success, the team needs to access internal and external intelligent automation expertise, provide training for new users and highlight what the technology does in plain terms to non-technical staff.

As part of the plan, ensure the team and appropriate leaders own and oversee the technology infrastructure. They need to build and develop close relationships with vendors/providers and implementation experts to ensure a level of expertise is available for all decisions, while creating the rules and decision process for choosing what is automated.

That plan needs leadership support at the highest levels and to be explained at ground level for the workers to understand the upcoming changes. It also needs to explain what workers will do with their new time and what new tasks they will perform, or if the aim is reduce headcount, then prepare them in advance.

Implementing any RPA across an enterprise requires highly detailed plans for operations, dedicated teams and resources. Return on investment and value goals, and reporting of key metrics are needed to ensure automation is done for the right reasons, provides enough value to the business and can be measured and monitored so success can be repeated and mistakes avoided in future efforts.

2. Avoid automating the wrong tasks

Having seen an RPA in action, many business leaders think, “great, let’s automate everything.” However, there are plenty of tasks that are too variable or unstructured, even for intelligent automation to gain control over. Even if a process can be automated, perhaps there are other aspects where manual oversight can add value further down the line, and there are some processes where the return on investment isn’t worth the effort.

3. Don’t rely on one department to do it all

The plan mentioned above requires the insight and agreement of many different areas within a company, so when performing your automation implementation, it’s essential to have people from those departments or divisions working as part of the team.

Leave it all to the IT team and they will likely miss some key business aspects. Leave it all to an operational unit and they will probably end up with data in a silo or creating additional processes that detract from the value of RPA.

Essentially, any successful project needs a broad set of input and support from across the business, otherwise, as with many other technology innovations that should work, it’s likely to fail.

4. Don’t rely on RPA from day one

Any business that dumps its existing tried-and-tested processes for a new RPA on day one is risking a major issue. Testing in isolation is not enough. Testing in the production environment may work fine for a short period, but there will come a time when data or workflow creates a previously unseen issue that can bring RPA grinding to a halt.

While an additional cost, keeping both the old and new systems running parallel for a short time allows for firmer validation of RPA. And the more RPA processes or elements of intelligent automation there are, the greater the risk of failure at one of the connecting points, which it may not have been possible to detect before they were all up and running.

Businesses should not expect instant results from large-scale deployments. RPA may take some time to show its benefits, usually within 2-3 months. If RPA fails to show results after that, then having the original systems in place, can provide backup while the goals and implementations are revisited.

Further reading:

• 100 Ways to Automate Your Enterprise
• Start Your Intelligent Automation Journey with RPA & Process Intelligence
• Enhancing Your RPA Implementation with Intelligent Information