Why Conversational AI Should Not Be An IT Project

Artificial intelligence (AI) has woven itself into the processes of many industries, businesses and teams, but who should own the investment or implementation, and who does it best serve?

Conversational AI, the use of automated messaging and voice assistants to create personalized interactions, can allow customers to sort out requests and problems, like inquiring about a package delivery or obtaining a mortgage quote. AI has taken transactional moments and pushed them beyond just messaging, transforming them into full-blown experiences with tone, sentiment and emoji comprehension, as well as the ability to include app-like features, such as carousels, maps, surveys, scheduling capabilities and much more.

On paper, this seems like something every company would want. But despite companies having the ability to access and adopt technology with increasingly personalized messaging and a trove of accessible data, deployment of conversational AI often fails.

Why? Because implementing technology like AI-driven solutions has largely fallen on the shoulders of IT teams. But this is a mistake.

There is a better way to incorporate conversational experiences into your company. In my experience, to increase the likelihood of overall conversational AI success, businesses should do the following:

1. Empower customer experience executives

Many failed attempts are a result of IT teams being mandated to carry out these solutions, eventually bringing it back to the build-versus-buy debate. Ultimately, power should be transferred to those with an eye for user experience (UX) and customer engagement, as IT has become overly obsessed with the natural language understanding (NLU) and integration components of conversational AI. Instead, businesses should focus on tying customer experience back to solving business problems.

How can conversational AI be leveraged to improve the experience? By emphasizing personalization, tone and quality content, backed by accessible data and insights.

2. Design AI by incorporating business goals

Often, businesses turn to vendors without the necessary skills and expertise. Third-party providers must have specific managed services in order to provide concrete value, including design thinking, user experience and content expertise; otherwise, implementations will run in circles around business problems rather than solving them.

Vendors, meanwhile, tend to be too focused on self-service products and technicalities as opposed to the bigger business picture that enterprises aim to solve. And many large system integrators have made the mistake of assigning only IT resources to clients’ conversational AI projects. As a result, many fail to produce return on investment (ROI) because the focus shifts to technical issues.

When in doubt, turn toward business objectives, goals and problems, even on AI initiatives like designing a conversational customer experience.

3. Earn experience via high-stakes use cases

Many conversational AI proofs of concept or pilots have not had high enough stakes to push deployments to produce high-quality outputs. For example, small, internal use cases for employees who have a single-task scope are not likely to inspire enterprise stakeholders to suddenly think bigger or explore the art of the possible.

This approach ultimately inhibits creativity and can slow a company’s path toward deploying conversational AI to its maximum potential: Consumer-facing solutions deployed at scale that achieve real ROI via increased customer satisfaction and reduced costs.

Over the next year or two, I believe businesses will still see mixed results in conversational AI deployments. The winners will be those that think about the problem differently.

To succeed in implementing conversational AI, measures have to be taken to ensure customer experience does not fall by the wayside. Instead, UX needs to be the driving force of experience design, with an emphasis on personalization, tone and quality content.

Conversational AI should not be just an IT project. Empower those with customer empathy to lead deployments, and surround them with experts in business analysis, customer journey mapping and design. Only then will businesses find ways to make conversational AI a true competitive advantage.

This article was originally posted in Forbes.