Craig Wright, Sr. Director, Infor CX
Retailers are continually looking for ways to reduce barriers to simple, consistent, and excellent buying experiences. But traditional shopping done via voice is adding complexity that can be very frustrating for retailers if they are focused on the wrong things. It’s similar to shopping via text messages—it works well in some situations and but can be very difficult in others.
For example, consider the necessary interactions to purchase a bar of soap.
Customer: Buy soap.
Voice-activated shopping assistant: There are 41 options for soap. Shall I read them?
Customer: No. Buy the most popular soap.
Voice-activated shopping assistant: The most popular soap is Dove. What size do you want?
Would you like to buy a Dove three-pack of soap? Do you want regular scent?
Voice-activated shopping assistant: Do you want next day shipping?
And so on…As you can see, the key to making conversational commerce work is understanding context. The best use cases for purchases done by voice are ones where CRM can bridge the context gap and fill in the details.
“Buy soap” would be interpreted as buying the same soap as the last purchase. “Buy a new water filter” actually means buy a new water filter that fits the customer’s refrigerator at the best price available. The CRM system knows purchase history and preferences so that the right buying decisions are made. This provides the context that simplifies interactions.
If the customer decides that they no longer like that brand of soap it is much easier to say, “I don’t like that soap” and be confident that this preference is saved to the CRM. Next time the customer says “buy soap,” the voice-activated shopping assistant says, “Shall we try a new brand?”
The data collected by CRM makes the entire interaction “smarter” by remembering preferences. Many preferences can be learned and applied over time or inferred from demographics, psychographics, and situational context.
Along this line of thinking, hyper-personalized voice recognition becomes a necessity to restrict or enable as needed. A recent newscast that played on televisions across America led to newscaster, reporting on a story of a little girl ordering a dollhouse via voice-activated shopping, himself triggering the voice-activated shopping devices in viewers’ homes to inadvertently place the same order.
To instill public trust in voice-activated technology as well as to implement proper security measures, there will have to be the ability to recognize the difference between voices with complete certainty. For example, only the parents voice will be recognised as authorised to purchase soap in the household. While convenience is the ultimate benefit to consumers, as well as to the brands marketing and selling to them, trust and security are critical to the widespread adoption and success of the technology.
While there may still be some hurdles to overcome in widespread deployment and adoption of this technology, the exciting combination of CRM, machine learning, and voice response is clearly a game changer and poised to bring conversational commerce to the forefront of retail.