Policy of Truth

Conversational AI re-shapes the Wall Street earnings call

You had something to hide.
Should have hidden it, shouldn't you?
Now you're not satisfied
with what you're being put through.

It's just time to pay the price
for not listening to advice
and deciding in your youth
on the policy of truth.



The earnings call is a Wall Street staple.

Once a quarter, company executives grab the mic and share accomplishments and challenges with investors and media alike.

Earnings calls aren’t required by law, though a 2014 survey showed 97% of respondents use them to disseminate information on a regular basis.


Voice technology left a crater-sized impact on earnings calls just as recently as three weeks ago, when Cerence utilized cloned voices of CEO Sanjay Dhawan and CFO Mark Gallenberger to read the company’s prepared remarks.

The company did this to show off its impressive text-to-speech technology, which it accomplished.

But in doing this, the company also might have answered a question it didn’t even realize others were asking.


Legendary management consultant Peter Drucker made a habit of sitting outside popular retail stores within shopping malls every Saturday.

Sure, he had their earnings information and publicly-released financials to gauge the health of these companies.

But he wanted to get closer to the action. He wanted data that was harder to capture, and harder to quantify.



He would watch and see how many people would enter certain stores, and how many of those would exit with purchases, versus how many might leave empty-handed.

And he’d end the day ahead, in terms of real-time business intelligence, of the rest of his peers.

Drucker played an important role in essentially creating what is now known as alternative data, which is nontraditional data that may be difficult to collect, and even difficult to conceive of, but that yields invaluable insights into consumer behavior.


Fast forward to today’s budding age of AI, and the teachings of luminaries like Drucker loom large over how we decide to deploy these tools.

Consider Helios, a fascinating company I had the opportunity to speak with today in-depth.



They’re using razor-sharp AI within their Mercury tool and a Peter Drucker-like approach to surface valuable hidden insights within earnings calls.

The company presents this as next generation alpha:

The CIA and top government agencies have known for decades that tone of voice holds unparalleled abilities. We've leveraged ground-breaking neuroscience across decades of scientific research to harness this power of how people speak.

Now when a CEO paints a picture of all positive words, you can know the real story.



One of my favorite documentaries is The Smartest Guys In The Room, which details Enron’s criminal deception of investors and subsequent fall from grace.

The film, correspondingly, also details how many families that were invested heavily in Enron stock were decimated.

If the technology that companies like Helios are bringing to life can help save American families from a similar fate, then that’s a value that can’t be easily ignored.



If AI knows that the vocabulary, intonation, cadence, volume, and other attributes of a particular company’s earnings call accurately predicts a negative surprise two quarters from now, there’s a whole lot of incentive for that company to use synthetic voices to sidestep that potential problem entirely.

The question then becomes whether the market adjusts to the use of synthetic voices and penalizes their use so much that it then makes more economic sense to simply stick it out with the human executives as usual, despite what they may say and how they may say it.

Cerence’s use of voice technology on their earnings call is historic, and deservingly proud, and well worth the pause it evokes to think about what may come next.




Leading up to Project Voice 2021, the #1 event for voice tech and AI in America, Project Voice: Coast to Coast is bringing an in-person voice/AI briefing to small groups in 35 cities across the United States. Presented by Cerence.

If you want to get out of the house, we’d love to have you join us. Masks and social distancing will be enforced, and outdoor venues will be utilized.

San Jose, CA - January 11

San Francisco, CA - January 12

Los Angeles, CA - January 13

Phoenix, AZ - January 14

Miami, FL - January 18

Tampa, FL - January 19

Orlando, FL - January 20

Jacksonville, FL - January 21

Portland, OR - January 25

Seattle, WA - January 26

Spokane, WA - January 27

Boise, ID - January 28

Salt Lake City, UT - February 8

Denver, CO - February 9

Cheyenne, WY - February 10

Omaha, NE - February 11

Boston, MA - February 22

New York City, NY - February 23

Newark, NJ - February 24

Philadelphia, PA - February 25

Washington, DC - February 26

Minneapolis, MN - March 8

Milwaukee, WI - March 9

Chicago, IL - March 10

St. Louis, MO - March 11

Nashville, TN - March 12

Memphis, TN - March 15

Jackson, MS - March 16

New Orleans, LA - March 17

Dallas, TX - March 18

Houston, TX - March 19

Detroit, MI - March 29

Columbus, OH - March 30

Cleveland, OH - March 31

Cincinnati, OH - April 1

Project Voice 2021 - Chattanooga, TN - April 12-14


The Voice of Customer Service is a special two-day event THIS WEDNESDAY AND THURSDAY (!) all about how conversational AI is being deployed within modern customer service operations, including contact centers around the world.

Register here.


Project Voice is the #1 event for voice tech and AI in America.

Join us in person in Chattanooga, Tennessee, from April 12-14, or virtually for Project Voice Worldwide April 15-16.

Voice/AI in mobile apps, voice/AI in contact centers, and an industry deep dive into the hospitality sector will be particular areas of focus.

Register for Project Voice 2021 here or Project Voice Worldwide here.


This Week In Voice VIP’s Spotify playlist can be found here.