Internal Chatbots

One use case for chatbots that is often overlooked is the internal chatbot for employees within a company.

External chatbots are chatbots that are provided to a company’s customers. External chatbots normally cover two functions, customer support (providing customers with answers to general or specific questions) or customer enablement (assisting the customers in completing tasks).

Internal chatbots provide the company’s employees with similar services. Technical support bots are similar to customer support bots in that they allow employees to access support for technical issues. For example, technical support bots might be triggered when a user encounters an error in a system they are using. Just like for a customer support bot, a technical support bot can escalate the issue to a human if the bot is not capable of answering the specific question posed by the user.

Internal bots are also called chatbots for operations, and perform a function similar to customer enablement. They allow employees to complete tasks without knowing the details of the underlying systems they are using, essentially abstracting the underlying system. Not only do they not need to know how to log into the system in question and use it, they don’t need to know which system, or even systems, they are using.

In addition, these types of chatbots can perform loosely coupled workflow type functions by moving the task from one system or employee to another until it is completed.

The main rationale for using a chatbot is that that task is done infrequently and/or is ad hoc, and not too complex or specialized.

If a task is done infrequently it is likely that the user will forget how the task is done (which system to log into, what the log in details are, where to go etc), and therefore using a conversational UI will be much easier for them. For example, informing HR that you are going on holiday, or updating your expenses form might be tasks that the chatbot oversees.

If a task is not complicated or specialized and is done on an ad hoc basis, it may be more convenient for the user to use a chatbot. For example, you need to know what the revenues for a given customer were for last year and don’t want to have to log into the business intelligence system to get them.

Internal chatbots can be more powerful than external chatbots because employees will accept a bit more of a learning curve in using the chatbots and also because a large range of functions can be accommodated under a single log in.

In terms of the learning curve, this might mean that the employee is more willing to explore the chatbot to understand what services it supports and also explicitly train the chatbot to some extent to make it better for others.

In terms of supporting a wide range of functions, once a given chatbot is created and employees are using it, it is easy to leverage the user base and add another capability to the chatbot, even if it’s created by a different department.

Another advantage for internal chatbots is that developers and business authors of the chatbot have easy access to the user base of the chatbot and therefore the development costs of creating a high functioning chatbot can be reduced.

Of course, any chatbot that is created be it internal or external, needs to provide a good return on investment. Essentially this boils down to assessing the cost of creating the chatbot versus the impact on profits in terms of revenue generation or cost-saving.

The cost of building and running the bot is generally determined by the complexity of the bot and the ease of use and flexibility of the development framework used to build the bot. The cost of building the first bot will be more than the cost of subsequent bots, particularly if subsequent bots can reuse integrations and other common components developed for previous bots.

The extent to which business users can add useful functionality to bots also reduces the cost of developing bots because they typically understand the use case better than the developer.

For internal chatbots, the cost-saving may be measured in terms of employee time saved and related measures. It is also possible that internal chatbots can lead to increased sales by providing employees with timely customer information and cross-sell opportunities, and even by improving customer service for customers and making them more likely to be repeat buyers.

There is also the consideration of how many users will use the bots. It is obvious that the more users that use the bot, the more benefit in terms of profit impact that bot will have.

Like every chatbot project, however, the benefits and costs need to be assessed on a case by case basis.

It is also true that the costs of implementing internal bots will come down and the quality of the chatbots produced will improve as the company learns from its bot building activities both in terms of how to build bots and what use cases work best for the company in question.

The underlying natural language processing technology is improving extremely fast, so the benefits of using this technology will become more and more obvious as time progresses and those unfamiliar with the technology and how to use chatbots will be left behind.

Internal chatbots are a less well recognized use case for chatbots. These chatbots can have an outsized impact on employee productivity and ultimately profitability which is the reason that their use by companies is set to explode in the years to come.