Working Out the Value of Chatbots and Voicebots

Reviewed and updated 22 October 2026. Original publication date retained.

Define what better would look like

Start with the work your team and customers are trying to complete. A useful improvement might be easier access to information, a clearer next step or less repeated administration. Describe the current service and where it struggles. This gives the business case a foundation beyond the number of conversations a tool can handle.

Use your own baseline

Review representative enquiries, volumes, handling effort and the outcomes people need. Identify which tasks are suitable for self-service and which need a person. If the baseline is incomplete, collect enough evidence to support the next decision rather than building an elaborate forecast from assumptions.

Include the work around the technology

Consider source preparation, conversation design, integration, testing, training and ongoing review alongside subscriptions or usage charges. A service may release capacity while still requiring human checking and support. Explain how the organisation intends to use any time released, rather than presenting every minute as an immediate saving.

Use a pilot to test the important assumptions

Choose measures that reflect the journey, such as useful answer quality, completed tasks, appropriate handovers and unresolved exceptions. Review whether customers and colleagues can use the service successfully. A trial should help the team decide whether to continue, adjust the scope or improve the foundations first.

Connect value to a delivery decision

A practical business case records the expected benefits, costs, assumptions, owners and review points. It should support a decision that can be revisited as evidence improves. Explore our AI business case and prioritisation, chat and knowledge and voice and telephony pages for the work involved.