Setting a job description for your chatbot may seem like a strange piece of advice. Analogous, maybe, to set KPIs for your telephone system or reporting channels for your anti-virus software.

But as weird as it may sound, taking the time to set a firm list of expectations, tasks, and even reporting channels is a perfect way to ensure that your AI bot is meeting your business needs.

Why do you need a job description for your AI chatbot?

Introducing an AI-enabled chatbot to a company often presents challenges. Embracing a new intelligent form of application is definitely a step into the unknown for many businesses. Therefore, if it is not carefully planned, a chatbot can cause more disruption than help, or simply fail to achieve its potential.

The challenges of setting up a chatbot fall into two main categories:

• AI the silver bullet: One of the key misnomers around business AI is that it is capable of doing and improving everything and anything. While there is no denying that AI is a powerful tool, likely to change the world, we are still in the first stages of learning how to utilise it. Having inflated expectations of how an AI chatbot may help your business will result in its applications being too broad and ultimately failing to achieve a useful business function.
• Knowing where to start: Implementing an AI chatbot intended to ‘improve customer satisfaction’ is a big ask. Where do you start with such an extensive objective and how can you even define when it is achieved? The more specific your chatbot application is, the more successful your teams can be at implementing, monitoring and improving it. Having an undefined role for your bot will likely result in undefined results.

To avoid both of the above situations we advise a restructuring of the whole discussion. Rather than seeing your chatbot as a general machine application that will help everyone, much like the introduction of a phone line, instead treat your chatbot as you would a new recruit.

As with any prospective hire it is important to carefully consider what the person will be adding to the business. This will likely include day-to-day responsibilities as well as weekly, monthly or even yearly KPIs. Clear report lines are also essential. These enable team members and stakeholders to identify who is responsible for any issues and provides stated accountability.

For example, a chatbot responsible for fielding customer service calls may have a KPI of resolving three cases a day and hitting a customer satisfaction score of eight out of ten on follow up surveys. If it fails to achieve this, its assigned line manager is responsible for investigating why. This clearly defines what the chat bot is responsible for on a daily basis and means that a business can identify what it is contributing.

However, this approach might not suit all chat automation projects. Simple voice search bots, and other utilitarian implementations, can usually get by without too much thought to their “job function”. If you are building an Intelligent Assistant with a broad set of duties though, then this is definitely a good idea.

The Alfie Bot example

A fantastic business use case of a chatbot with a clear job description is the Alfie bot, launched by AGL Energy in 2017.

The creators of Alfie focused its functionality around a concentrated task area. Their aim was to make this specific area as successful as possible, instead of having broader objectives that could have been less easily managed and achieved.

For this reason, Alfie was given the very specific task of handling all account cancellations. Its success was monitored in relation to how well it achieved this result and it’s progress was monitored by the cancellation manager.

By making a clear job description for Alfie the team at AGL were able to easily communicate the value and function of their chatbot to stakeholders and also accurately calculate how much it contributed to cost savings.

How can you make the right chatbot job description?

Creating a clear job description for your chatbot should take into account a number of factors. Firstly, it is important to consider what focus makes the most sense from an ROI process – how will it save you time and/or money?

Next, how does this focus fit into the skill set of chatbots? Bear in mind that AI is best at completing repetitive, definable tasks using well-defined processes. Will this be suitable from a consumer perspective? Are your customers comfortable receiving this form of service from a chatbot, or would they prefer a screen or human interaction?

Carefully thinking through what task niches chatbots can best fill within your business structure avoids confusion at every level and harnesses AI’s power effectively. Remember that once a definable task area has been achieved you can continue to grow your AI chatbots functionality, provided it follows the same thought processes. Starting with a small goal will help your company learn where the true value of chatbots lies and how best to utilise it.