DOUG SCHUMACHER

experience designer + writer

voicebotai

Your Brand Has a Personality Whether You Like it or Not. Here’s How to Define It.

By

Note: This article originally appeared in Voicebot.ai on April 4, 2020

Pay attention, mortal meatbot, and you might learn something about voice personality in this article. Or, maybe a better opening line would be, Are you interested in learning how to use voice personality to increase the effectiveness of your voice apps?

Or maybe not.

The point is, there are different ways to deliver a message. The style you choose will likely impact your message’s effectiveness. And that’s the point of a voice personality methodology. To hopefully nail the optimal voice for your brand.

YOU CANNOT NOT COMMUNICATE

The line above is from influential graphic designer David Carson. It’s a simple concept. Whatever font, imagery, or layout you use to represent your brand, it will make a statement about the identity and values of the company behind the message. Those who think they can hug the middle of the road attempting to avoid standing out can come across as bland and generic. And of course, bland and generic can be ways to describe a personality. Just not terms most brands are likely to aspire to.

The spoken word adheres to the same communication principle. Clifford Nass, a former Stanford professor of Communications who did extensive research on the impact of various voice personality traits, said that within seconds of a human hearing a voice, we begin to assign all sorts of personality traits to it. Whatever voice it is. However nondescript we may think it is. To demonstrate this, below are 2 human voice-over reads for the same script. Even with the bland, generic script used in those reads, listen to the two voices and note how your mind conjures different personas for each.

AUDIO SFX: S1VO1
Credit: Patrick Lagreid at Voices.com (LINK: https://www.voices.com/actors/plagreid)
AUDIO SFX: S1VO2
Credit: Debbie Feldman at Voices.com (LINK: https://www.voices.com/actors/debbiefeldman)

Of course, like any self-respecting professor at a research university like Stanford, Nass didn’t stop after a couple of experiments and call it a day. Through extensive lab environment testing, he developed the Media Equation theory[1]. It introduced the idea that humans assign personalities to machines we interact with similar to the way we do for humans we interact with. That’s highly significant for anyone building any type of technology interface. To demonstrate that concept, listen to the audio clips below.

AUDIO SFX: S2VO1
Credit: Sarah Raines at Voices.com (LINK: https://www.voices.com/actors/Sarah_Raines615)
AUDIO SFX: S2VO2
Credit: Scott William at Voices.com (LINK: https://www.voices.com/actors/SWilliam)
AUDIO SFX: S2VO3
Credit: Amazon Polly Synthetic Voice Joanna

In accord with the Media Equation theory, the persona behind the 1st and 3rd clips should feel more alike than the 2nd one. I believe they do.

A MODEL FOR PERSONALITY TYPES

In his book “Wired for Speech”, professor Nass provides detailed accounts of 20 studies on how voice personality impacts the user’s perception and behavior. He also points out the usefulness of a personality model to help in selecting the right voice characteristics and presents the Wiggins voice personality model as an effective tool. Wally Brill, head of conversation design, advocacy and education at Google, has given several good presentations on using the Wiggins Personality Model, as seen below.

The Wiggins model is simple. There are four quadrants, with the x and y-axis presenting polar traits. The x-axis contrasts distance with friendliness, while the y-axis goes from submissive to dominant. While voices don’t convey all personality characteristics like if a person is intuitive or spontaneous, there are a number of voice traits commonly associated with personality types. Nass focuses on four of them. Speech rate, pitch, variance in frequency, and volume. The Wiggins chart below plots how those voice traits influence personality.

Nass even plots where several common character types reside on this chart. A supervillain like Darth Vader? His deep, slow, stolid voice provides a mix of dominance and unfriendliness. A classic bad guy voice. Superheroes also need to project dominance while also having a friendlier tone. So while they may have the deepness of a villain’s voice, they combine that with a wider frequency range and greater speed, adding friendliness to the mix.

More secondary characters will likely have less dominance, i.e. be more submissive, and will vary along the friendliness and unfriendliness axis based on their desired perception of being on the good or the bad side of the story. This model may provide a blunt view of voice personalities, but it’s a good starting point for discussions around what voice personality would most likely fit a brand.

THE RIGHT PERSONALITY FOR YOUR BRAND

A logical next question is, How do you know which voice personality is right for your company? The answer to that will vary significantly by company, but a good place to start looking is the company’s brand strategy document. The brand strategy document defines the brand’s identity. What the brand stands for. And it culminates in a brand promise: A statement that sums up the brand’s identity.

The brand promise takes into account three things in particular.

  1. The target audience(s) for the brand’s products or services. Not only who they are, but what their needs are, and what they find most relevant about the brand.
  2. The competitive landscape. How the company creates brand distinction from its competitors.
  3. The brand’s capacity for authentically supporting the brand promise. In other words, do the brand’s values and capabilities realistically back up a given strategic position?

That brand promise and the backing research in those three areas should go a long way to outlining the type of voice that will likely resonate with your target audience. Here’s a brand positioning statement template from Hubspot that contains those elements.

For [your target market] who [target market need], [your brand name] provides [main benefit that differentiates your offering from competitors] because [reason why target market should believe your differentiation statement.]

And keep in mind that brand promises are aspirational. Something the brand feels they can be, or at least evolve into. If you don’t have an official brand strategy document, then any information on the above three areas of consideration should be helpful. The better you know your target audience, your competitors and your own brand, the better you’ll be able to determine where your ideal voice should be in terms of friendly vs unfriendly, and dominant vs submissive.

HOW YOU COMPARE WITH COMPETITORS

Making our brand personality distinct from competitors requires that we understand our competitor’s brand personalities. And indeed, while we probably won’t have their brand strategy documents, we can easily observe competitor’s communications. From advertising campaigns to logo designs to color palettes, the assets used in brand communications are a reflection of the values and positioning in their brand promise.

To illustrate this, let’s use an example of two brands from the financial industry: Vanguard and eTrade. When casually observed, financial industry brands can all sound the same. However, with some fairly quick observations of the brand’s communications, we see differing approaches to presenting themselves in the marketplace. 

Let’s start with Vanguard. The company is an icon within the investing community. Indeed, it’s founder, the late John Bogle, was a respected industry giant known for his thriftiness and efficient trading system. Company founders often drive the personality of the brand. Think Ford. Ralph Lauren. Um, Trump. Accordingly, the Vanguard logo font is a conservative serif font. The vintage-looking ship illustration certainly isn’t trying to rock the boat, if you will. Even the color palette is simple and about as unexciting as red can be. 

Now consider eTrade, in contrast. The brand font? Bold sans serif uppercase. The logo? Two colliding arrows. A strong sense of motion and energy. And the color palette. Purple and chartreuse. If you’ve ever seen eTrade TV commercials, you may know them for their sometimes brash and outrageous sense of humor. Vanguard TV commercials? You probably can’t think of any.

While two brands within an industry may pursue different target audiences and may stake out different personalities within their industry, each brand personality needs to be authentic to the company behind it. Considering our example brands, it would be hard to imagine Vanguard using a color palette like eTrade’s. Likewise, eTrade using anything like the Vanguard ship illustration for its icon would seem odd. So at least on the surface, the two company’s brand personalities feel appropriate. Going back to the Wiggins model, you can see how these two brands plot differently in terms of friendliness and dominance. Other brands can be added until you have a complete sense of where your brand stands within its competitive landscape.

WHAT TO EXPECT

Hopefully, this article has provided the reason, value and methodology for giving deep consideration to the type of voice, be it human or synthetic, that will represent your brand. Keep in mind, a process like this won’t define the entirety of your brand personality. But it gives you a structured view for comparing your brand to competitors and provides anchor points for discussion and further analysis.

Chainsaw Product Highlights Risks for Brands Rushing into Voice Industry

By

Note: This article first appeared in Voicebot.ai on April 1, 2020

chainsaw image

A dubitable voice product that garnered early VC investment is being taken off the market. The investigative journalism arm of Voicbot.ai has taken a closer look at this remarkable story.

There were doubtful murmurs when the concept of a voice-enabled chainsaw was first revealed back in 2017. But sometimes, the rising tide of a fast-moving industry lifts all barriers to good judgment.

CHAT N’ CUT

The “Chat n’ Cut”, as it was named, had humble beginnings. According to the ‘origins’ story on the brand’s now defunct website, the two founders, both advertising copywriters, had apparently polished off a case of Lumberjack Stout beer after work one Friday night and started brainstorming voice product ideas. In their inebriated state, the Escheresque connection between chainsaws and the voice industry became a direct path to startup stardom.

Sources close to the founders suggest that upon waking the day after their brainstorming binge, no hangover of any magnitude could keep the two from pursuing their newfound dream. Perhaps most surprising, they secured $1 million in funding with little more than a conversation flow diagram. Wil Gamble, CEO of seed funding firm Gamble & Shock, explained their logic for investing in the company:

“So many legacy industries were digitizing. We figured if the taxi and office rental businesses could be appified, well, it seemed like the chainsaw industry was ripe for the picking. Add to that the rapidly emerging voice space and, well, what’s not to love?”

And, while hindsight is 20/20, it seems there were a number of questionable decisions made along the way. The initial target market was the lumber industry. Early field tests were conducted with a logging firm in Veneta, Oregon, and a problem quickly surfaced. While it was easy to turn the chainsaw on by voice command, after it was started, the chainsaw motor noise overwhelmed the mic, and no additional user commands could be picked up.

“It was just weird,” said a field crew boss at the firm who asked to remain anonymous. “Usually, when the ‘jacks are out at work, all you hear is the buzz of the saws and falling trees. But this time, all you could hear was the saw, and then a bellowing voice of one of the ‘jacks screaming “Hey Chat ‘n Cut”, trying to shout commands above the chainsaw noise.”

CHAIN OF FOOLS

Field reviews were understandably bad. But what might have seemed a clear and obvious alarm was misinterpreted by the brand’s market research consultancy. Instead of viewing the research as an indication the Chat ‘n Cut was not a professional grade product, they instead took it as an indicator the product should be repackaged and sold to the consumer market.

We got in touch with the former director of marketing for Chat ‘n Cut to get a sense of how that strategy played out. Apparently, things didn’t go well. Under a blanket of anonymity, she noted that while the OKRs (Objectives and Key Results) were initially focused on metrics like CPS (Cost Per Sale) and NPS (Net Promoter Score), the company soon transitioned to the CPMWM metric (Customers Per Thousand who were Wounded or Maimed).

Voicebot’s investigative reporting team was able to source a key document that indicates management should never have sent this product to the market. The sample dialog flow had plenty of branching leaving us to wonder if there was an actual “happy path” that Chat ‘n Cuts entry-level conversation designers could agree on. It was also heavily edited to add new error states associated with unintelligible speech.

The voice industry remains promising, and we present this story in complete objectivity. While the personal injury attorneys and media feast on the carrion of this derailed initiative, keep in mind the lessons learned from our ancestral early-stage technology ventures. Just because you can doesn’t mean you should.