As a PLM (Product Lifecycle Management) consultant, I can’t help but notice that organizations who revamp their PLM landscape often face a brick wall when it comes to user acceptance. All of the promised benefits, all the glowing forecasts of improved KPIs—none of it can materialize unless a business transformation happens along with the IT transformation. PLM users have to adapt to the new system and processes, or there will be no improvements.
A chatbot is a computer program that conducts a conversation with a user through speech or text. According to Wikipedia, such programs can be designed to convincingly simulate how a human would behave as a conversational partner, such as the recently launched Google Assistant.
I believe that chatbot technology can be deployed to help overcome the obstacle of user adoption in PLM implementations. The basic premise is that the implementation team would code the chatbot with the rules and help features necessary to provide conversational assistance to users in real time.
What’s Wrong With the Way We Train New PLM Users Now?
The traditional approach to user training is for implementation teams to write training manuals and lead training sessions. Unfortunately, implementation teams are rarely aware of all the reasons behind existing business processes, which can lead to gaps in their understanding of which features of the new system warrant the most focus. Training sessions instead predominantly focus on how to use the application, rather than how to complete a process. As a result, users might not follow what they are trained on, and opt to delay correcting their misunderstandings until the new system goes live. These are risky decisions and lead to multiple issues:
- Users who are unable to figure out how to perform a specific task
- New functionality and ways of working being perceived as defects and routed to IT support. The large number of support tickets leads to a misconception that the application is unstable, while also causing redundant efforts by IT to close support tickets that represent gaps in user knowledge.
- Developing manual workarounds to avoid using the system
- Missing out on the advantages that the new platform provides
- Additional follow-on investments in user training
This cascading effect of poor user knowledge leads the implementation team to make multiple efforts to address the perceived shortcomings, which in turn delays the rollout of the PLM system to other areas of the business.
How Can PLM Specialists Address These User Training Issues?
One way to solve the problem is to provide continuous support to the users as part of their daily job execution. This is typically done by assigning a set of identified individuals to be trained as experts who can then locally help users. This solution can work for small rollouts, but does not scale well when the user base is very large.
By contrast, a chatbot actually gets better as the number of users increases. It is a self-learning artificially intelligent program whose response quality improves over time by understanding individual as well as group usage behaviour. This way it predicts users' needs, acts autonomously on user's behalf and builds trust with learned results. Chatbots can be integrated as virtual assistants within applications or used as part of a messaging platform.
A Use Case for Chatbots in a PLM Implementation
Chatbots can be deployed in a new PLM rollout to aid users with clarification of steps, provide useful links and quickly explain key topics. Chatbots can also monitor user actions and suggest the appropriate next steps. They can bridge the gap between formal training and the day to day independent work of a user.
Consider that most business and PLM transformation implementations introduce rule-based systems to ensure data quality and that business processes are followed. These rules often run into hundreds of different scenarios. For users who are not up to speed on these rules, they can run into “show stopper” situations where, until they find out the correct path forward, their work is stopped. A chatbot can be an efficient solution in this case, helping the user get past his show stopper situation. In a more advanced chatbot implementation, the user could be cautioned against ever going down the path that caused them to run into the problem in the first place.
Beyond providing help to use the existing system in its current configuration, chatbots can also help teams to gather feedback that can lead to improvements in the system design. It is rare for the PLM implementation team to get all aspects of the rollout perfect on the first try. There are bugs to fix and process changes to configure. Many teams use social platforms or discussion forums to gather this feedback from users. If users were instead presented with a chatbot interface, they can more easily provide feedback that the implementation team can analyse, which can lead to faster improvements to the system.
Developing a Chatbot for a PLM Implementation
Development of a chatbot solution is not technically very complex. The key capabilities that any solution should target are:
i. Artificial intelligence,
ii. The ability to converse in natural language and
iii. A self-learning ability that improves the chatbot performance over time by understanding individual as well as group usage behaviour.
According to TechWorld , some good platforms for chatbot development are Engati, Microsoft’s Bot Platform, ChatScript, Pandorabots, Rebot.me. Some of these, such as like Engati, claim that you can build, manage, integrate, train, analyse and publish a chatbot using its software with no major coding effort.
In summary, the success of any PLM transformation is driven by the business usage of the new solution, which in turn is a function of how well the users adapt to it. Classical organization change management techniques can often fall short, especially on very large transformations. A chatbot can be a cost-effective alternative to help users meet their business transformation goals.