Artificial intelligence (AI) and machine learning (ML) continue to advance, providing businesses with increasingly crucial competitive advantages. As organizations look to integrate AI into their digital transformation efforts, they will inevitably face objections. Here are some best-practice guidelines for addressing any issues and increasing the odds of a successful implementation.
AI and ML have been transported beyond the virtual to become a reality that features in many business applications ranging from natural language processing and chatbots to credit analysis and product recommendations.
There still exists confusion between various terminologies. Along with AI and ML, we also have robotic process automation (RPA) and intelligent automation. Before organizations can move toward implementation, they must first understand the core differences between them.
Clarifying Terminology
- RPA – The most basic form of automation. It typically uses a software ‘robot’ to automate tasks. This automation is limited to strict rule-based, repetitive actions that can span functions such as updating new customer details and populating reports in various formats. Given its simplicity, this is already a highly mature technology.
- AI and ML – The next evolutionary step after RPA. AI and ML are not constrained by strict rules and can “learn”. When companies feed these programmes sets of data (the more, the better), they can continuously tweak and improve their model – taking them past simple routine and closer to actual human-level decision making.
In financial services, many institutions are now using AI and ML for credit approval. A key advantage is that they can analyze unstructured (non-organized) data, such as social-media postings, and integrate them into their decision model.
- Intelligent Automation – An umbrella term used to describe the use of both RPA and AI solutions to transform business processes digitally. This is the most useful tool for actions that involve both structured and unstructured data.
Driving Increased Revenues and Cost Savings
Although this is an evolving technology, the bottom-line benefits of intelligent automation have already been seen. A McKinsey survey[1] found that 63% of companies that have already adopted AI reported a revenue increase in the specific business areas where it has been implemented. A further 44% noted that it has also reduced costs. When companies look at the broader process of digital transformation, intelligent automation must be given serious consideration.
Addressing Objections to Intelligent Automation
For companies seeking to reap the benefits of intelligent automation, they must first address the four common objections that will inevitably arise:
- Displacement
- Ease of Use
- Security Risks
- Ease of Implementation
In the following sections, we will look at each objection and explore how companies can mitigate and deal with them.
- Displacement – Will Intelligent Automation Replace Humans?
The first objection is one that usually jumps to most employees’ minds – will intelligent automation take people’s jobs and possibly affect morale? This is an understandable response that management must handle with care and empathy. Changes in the workforce are inevitable, and this should be acknowledged.
The key lies in helping employees to understand that intelligent automation exists to enhance productivity, not replace it. People will remain the decision makers. Companies also need to explain that smart automation will help remove the burden of repetitive tasks – freeing up employees’ time to perform higher-level, more value-added work. The answer, therefore, lies in open communication.
- Ease of Use – Will Intelligent Automation be Too Difficult to Operate?
Another concern is that intelligent automation tools will be too complicated for anyone but specialists to use. To address this issue, a business needs to integrate new technology into its process correctly and sensibly. Preparation is vital.
Companies need to have a clear plan based on input from all the relevant stakeholders. A dedicated center of excellence can also be established for planning, developing, testing, and supporting the various rollouts. While the team will naturally be IT-focused, it should also include members from across the organization. Such processes are necessary even if the company opts to employ technology partners for their intelligent automation efforts.
- Security Risks – Will Intelligent Automation Create Security Risks?
For algorithms to automate business processes, they should be granted a certain level of access rights. This can lead to concerns about potential security breaches – even more so if the software in question is sourced from an outside party.
The use of best practices can reduce such risks and companies must fully understand their own data first. This will help to establish a baseline. From there, a business can decide on the necessary level of access.
Cybersecurity risks will always exist, especially in today’s information age, but these should not be an excuse to halt progress. Instead, they are to be intelligently managed.
- Ease of Implementation – Will Implementation be Too Difficult?
There will always be those who think that the cost and effort of implementing intelligent automation are not worth it. McKinsey also observed that that 70% of digital transformation efforts fail[2], but data also shows that intelligent automation can drive real bottom-line benefits. Through careful planning and use of best practices mentioned above, companies can manage any risks.
There is also no need to go “all in” at the beginning, so firms should adopt a step-by-step approach. They can start with the already-mature RPA technology and review how the implementation went. By the time they move on to the more advanced AI algorithms, they should already have refined the necessary processes.
A Logical and Necessary Progression
The word AI can spark images of some far-off future. This is not the case. AI is already here, and it has become a powerful competitive advantage. As companies assess whether or not to use intelligent automation as part of their digital transformation strategy, it’s worth considering that the inevitable risks of waiting could far exceed the manageable risks of implementation.
[1] https://www.mckinsey.com/featured-insights/artificial-intelligence/global-ai-survey-ai-proves-its-worth-but-few-scale-impact
[2] https://www.mckinsey.com/~/media/McKinsey/Business%20Functions/Transformation/Our%20Insights/Why%20do%20most%20transformations%20fail%20A%20conversation%20with%20Harry%20Robinson/Why-do-most-transformations-fail-a-conversation-with-Harry-Robinson.pdf