Next Wave: Speech Recognition Grows in Popularity


BLOG VIEW: It’s pretty safe to say that everyone has either heard of and/or used Siri, Alexa or Google Home technology to order laundry detergent, ask for directions or learn that the exact time of the last solar eclipse was “Aug. 21, 2017, starting at 10:16 a.m. Pacific and ending at 2:48 p.m. Eastern.”

That level of detail from a simple voice request is what many people have become used to receiving on a regular basis. Now, more people want to have that same level of access to information when it comes to customer service. This is precisely why more mortgage servicers will need to consider this kind of conversational speech recognition technology for their contact centers.

A conversational interface in an interactive voice response (IVR) system mimics human interaction. The caller may ask a question like, “Have you received my payment?” and the system will understand and respond appropriately. Virtually any borrower question may be interpreted by the system, eliciting a proper response. This type of conversational interaction allows the caller to immediately get to pertinent information without wading through a series of touch-tone menus.

As the industry moves further toward acceptance of natural language speech recognition, there are some key factors to keep in mind and to address if this technology is to be implemented successfully.

Embracing change

Let’s face it: Our industry has been bombarded with new regulations, new requirements, more new compliance issues, and increased pressure for a better customer experience than any other time in history. It can prove overwhelming. The light at the end of the tunnel is that everyone is aware that technology will help level the playing field and be an important part in remaining competitive and having a decent market share.


While some companies have shown an interest in implementing conversational speech recognition technology, the cost of developing and incorporating it into existing platforms can be sizeable. Not only will it require additional time, but there may also be a need for additional staff. While the up-front costs may increase budgets, the long-term benefits and customer satisfaction will make it worth the investment.


Companies will need to properly train contact center agents to effectively understand how to transition from a conversational system if the borrower wants to talk to them. There will always be a need to have a human being for these scenarios and for those borrowers who prefer interaction with real people. So, companies should not look at speech recognition-based systems as a way of eliminating or significantly downsizing contact centers. Therefore, having a new application in place will require that all staff are effectively trained.

There is an opportunity for mortgage servicers to show the growing number of tech-savvy borrowers that it is easy to do business with them. Servicers also need to show they understand how to incorporate the technology into existing systems without using all of the company’s technology budget. It can be done. It is just a matter of developing a plan, implementing that plan and following through with iterative improvements.

Barry Hays is co-founder and senior vice president of TeleVoice, a provider of contact center technology to the financial services industry.

(Do you have an opinion to share with MortgageOrb? Get in touch! Send an email to

Notify of
Inline Feedbacks
View all comments