It’s one thing to have launched a Conversational AI, but another entirely to launch to over 1M+ customer in an enterprise. Now imagine that 10 times over, and you have reached the level our interviewee has conquered and since surpassed.
Dennis Chan is an award-winning Enterprise Conversational AI expert. He has launched successful voice and text-based assistants across multiple industries, including banking, insurance, government and technology. For the last 17 years, Dennis has been involved with every aspect of a Conversational AI projects from the ground up, whether that be leading the project or building the technology himself. Currently, he is the Founder and Director of Pure Speech Technology, a consultancy for enterprises delving in Conversational AI projects.
We asked him for his 3 biggest lessons from launching so many Conversational solutions for enterprises. Here were his lessons.
1. Always define your Purpose and KPIs first
Too often, conversational AI projects start with an order from C-level, without a clear indication of what is to be achieved with the new technological endeavor. The project team is left to do their best with a hastily put together team and no subject matter experts on Conversational AI for enterprises in-house. In the scramble to produce an MVP without expert help and a rush to completion, it’s easy to oversee that the project proposal is missing a meaningful Purpose and KPIs to track milestones.
On the other hand, there are projects fueled from the bottom up, starting from self motivated engineers. They tinker around with chatbot and voice technology, attend a few conferences or courses and build the technology, but pitch the technology without real business value.
It’s key that the purpose of the project is meaningful. If end users do not benefit from the technology, naturally, metrics like adoption rates, usage and satisfaction will be far from expectations.
Related: Building an Enterprise Voice Assistant MVP: Wins, Losses and Lessons
2. Don’t focus on the technology
Work out your requirements, then select the best tool for the job. Depending on the purpose and requirements of the project, a straightforward conversation only needs a simple GUI drag and drop builder, but more complex user experiences may require a full development team.
When assessing Conversational AI solutions, project teams often take into account vendor restrictions, developer requirements, or are influenced by political agendas before the requirements for their own project. This is a grave mistake.
Too many projects become incredibly costly or difficult to build and maintain because other agendas were prioritized. These projects could’ve been 50% easier and less expensive had other Conversational AI technologies been considered.
Read: 6 Steps for a Winning Voice or Chatbot Channel Strategy
3. Where needed, get expert Conversational AI help
Launching enterprise-level conversational AI projects is particularly complex, requiring multidisciplinary expertise from integration engineering and data science to customer service and conversational UI/UX. It’s highly recommended to find a specialist for any disciplines with gaps, or at least an expert who has been across all stages of the Conversational AI lifecycle on advisory.
Equally important, ensure the right help is enlisted. There are many people who have launched Conversational AI bots (many of which aren’t live with real customers, or don’t make it past a proof of concept), but not many have at the enterprise level, and certainly not that many with cross-industry or cross-functional expertise. Select a subject matter expert or organisation who has succeeded in launching Conversational AI bots in enterprise, and ideally one in your industry.
Need a hand building, improving or scaling your Enterprise Conversational AI? Get a free 15 minute consultation with our chatbot and voice technology experts!
Leave a Reply