What are the mistakes that financial service providers could make in the automation projects? How should we address them in a better way? This article will present those mistakes and propose possible solutions for reference, based on the automation in financial service providers.
1. Heavily dependent upon a single technology
Partners and technicians claim that a specific technology like RPA, NLP, low-code, no-code could help save money for the company. But each coin has two sizes.
Once they buy and deploy RPA, they may rush to promote it among wider audiences. However, CIOs often do that focusing on a business process, rather than taking the result-oriented approach to realize automation with more tools.
For Hyper Automation, these technologies should be combined to create the real value. So before choosing the automation tools, you'd better determine your business goals, and your methods to realize automation based on an architecture. You should focus on business implementation, rather than the use of technology alone.
Build the tools with of several technologies to provide more functions to reconstruct the business.
2. Claiming that automation can be realized without IT support
Low-code and RPA are well known for their simplicity use. Many enterprises falsely think that they could implement them with IT support.
As a result, IT usually does not get involved from the scripting to the implementation of automation, but is required to maintain the operations of all automation tools.
In reality, most business users do not have systemic knowledge about handling user data and records, so they are more likely to make mistakes in using the data.
While utilizing a fast, agile approach in developing processes is the key of a successful automation project, we should never ignore the inputs from IT and technical experts anyway.
Set up a delivery center where everyone should not only have the business knowledge about the banking and insurance services, but also the analytical and technical skills as well as the IT management experience in the areas. Furthermore, they should have a strong desire to automate processes and strive for innovation.
3. Not all departments affected get involved in the automation implementation process
Automation in nature will have big impacts on some departments. Some employees' daily jobs may be changed, which may create resistance among them. So, HR and all managers at various departments should take actions to encourage them to adopt it, and remove any barriers, in order to prevent projects from being stagnant. Also the IT should provide support to keep the progress in sync.
Engage someone with management skills in the departments affected.
4. Think automation never fails
Featuring quick deployment, automation could be a good long-term solution if any business with system integration in an enterprise fails to meet its business and financial goals. However automation may sometimes extend the life of some redundant functions and hide inefficiencies with cost savings.
Evaluate the pros and cons of automation projects regarding system replacement and integration strategies, and balance short-term gains with long-term sustainability.
5. No enough time spent on testing
Automation technology is not fault-tolerant. Generally, it can only work when the algorithms and rules are absolutely correct. Once it fails, the business will be damaged. So, testing must run through the entire process.
Businesses that fail to spend a lot of time upstream on testing for automated tools will quickly find such issue in production, which would affect customers, employees, and business processes altogether.
Solution: Define multiple process paths that can be taken, conduct the fullest user testing on them, and monitor data while running these automated tools.
6. Waste time on business processes that are too complex and not ready for automation
Despite detailed steps defined to identify processes at the front end, there would still be cases where the wrong automated process is selected.
This may be caused by several factors. For example, the process was not documented in detail or fully understood, the processes are not contiguous, or too many variable factors are out of control in the decision-making stage.
Don't fall into the money-saving trap. More time and energy would be wasted if such processes are not stopped immediately.
Set out some rules and metrics to help decide which processes and tasks not ready for automation need to be dropped. For example, decide based on the number of steps in the process, the number of integrations required and the transparency of existing processes, etc.
7. Treat automation as an exercise to replicate processes
A common mistake in deploying process automation tools is replicating processes as-is, rather than re-constructing them to improve business benefits.
There are some fresh new automation scenarios where the original process can be redesigned from ground up or use a totally new process based on the business goals or directions to achieve the expected benefits for the enterprise.
Solution: Leverage the four-step approach adopted by GARTNER in the automation decision-making phase to come up with a well-defined automation method, to achieve the expected business goals and automation goals.
8. Ignore corporate culture and its impact on employees
In many discussions, people always overlook the impact of automation on employees. However, employees' responses to automation projects are key to ensuring they can be completed successfully.
So sufficient emphasis should be put on the internal culture shifts in an enterprise to help spark ideas to re-construct processes at work, while driving a belief that all tasks suitable for automation can be automated.
Create the solid plan based on Gartner's research paper on cultural resistance to help the company break cultural obstacles in implementing automation.
9. Use wrong metrics to measure results
Businesses have moved beyond pure cost savings to measure the success of automation changes, realizing that they need to incorporate more meaningful factors. For example, an automation proposal aimed at increasing customer engagement should be measured by whether it can elevate customer satisfaction, and save customers' labor costs, and whether it can help improve the company's reputation.
Set some metrics first, and make sure you meet the basic criteria for the current metrics, such as SLAs (Service Level Agreements), customer satisfaction scores or error rates, etc. before making a proposal.
10. No enough attention paid to post-automation tasks
Unlike other projects, automation projects require continuous and extensive IT support after implementation. Ongoing assessment, monitoring, and regular quality checks after deployment should be in place to ensure that robots are scripted correctly and can be run as expected, while avoiding major mistakes, and viruses or technical issues preventing businesses from working well.
Establish some post-automation processes to ensure that the automation tools can be continuously monitored by operation teams, while helping businesses recover from disasters.