Here’s the intro from this “USA Today” article:
Oscar-nominated actor Samuel L. Jackson is lending his iconic voice to Amazon’s Alexa – profanities and all. During Amazon’s event to unveil new products and services Wednesday, the online shopping giant announced that Jackson will be the first celebrity voice for its Alexa virtual assistant and was created using neural text-to-speech technology.
There will be both an explicit version and a clean version when the feature launches later this year. The Alexa “skill” will cost 99 cents as an introductory offer. After the introductory period, the price will be $4.99, according to the product page.
While Jackson is the first official celebrity voice for Alexa, in a 2018 Super Bowl commercial, celebrities Gordon Ramsay, Rebel Wilson, Cardi B and Anthony Hopkins were called on to fill in when Alexa lost hers. The commercial won USA TODAY’s 30th Ad Meter.
Playing music is huge for voice. Particularly in the car – the #1 type of use there. But it’s big outside of the car too. This cnet article provides some tips for music lovers – here’s a select few from that:
1. Change the “default” music to whatever music service you like to use the most. For me, I love Spotify. So I changed my default to that. There are other alternatives that you can select as your default including Pandora and iHeartRadio.
2. You can ask your voice assistant for similar music if you feel like exploring. For example, “Alexa, play music similar to Fat Freddy’s Drop.” This works for both similar songs & artists.
3. Creating playlists by voice only works if you’re using the default that Amazon provides – it’s own streaming music. But if you have playlists on your Spotify account, etc., you can ask for those by voice. You just can’t modify them by voice.
4. You can essentially set up a Sonos-like framework by using multiple voice assistants in your house, spread out in different rooms. Now that is a real cost-saver! [Per this note, you can also use Alexa with your Sonos system.]
5. You can have Alexa wake you up to music – “Alexa, wake me up to Santana.”
Since voice requires you to anticipate what will be asked by your customers when building a skill, what better way to figure that out than simply researching what customers have already been asking?
So as part of this research, ask those in your company in the front lines what customers ask. If you have people dedicated to customer service, their knowledge is invaluable for this exercise.
Better yet is if the company keeps records of their customer interactions – that can be an invaluable trove of data. But don’t solely rely on that. You want to talk to your customer service reps to understand their firsthand experiences. You need to learn the ways in which particular questions are asked – so you can anticipate all the different types of utterances that may be expressed – in addition to the types of questions asked.
Another valuable source of information about what customers are asking is looking at social media. Social media has become a preferred way for some people to seek customer service help. Look at the comments on Facebook and blog posts. Review your Twitter feed.
And of course, people email questions too. Or use chatbots. Review what is asked through those channels too.
A few weeks ago, I blogged about this free 48-page playbook by “360i” about what you should know about voice from a marketing perspective. To help instruct their clients in how to approach voice opportunities, 360i has built a “Voice Search Monitor” over the past year to study how the major voice assistants respond to various scenarios. They’ve witnessed changes over time with this monitor – and learned that, at least as of right now, Google knows a lot more than Alexa (5x more on average). Google Assistant prefers to draw from location-based data for retail queries, whereas Alexa relies on its own top-matching product recommendations.
This excerpt from 360i’s playbook shows the kinds of questions that 360i’s “Voice Search Monitor” is asking the major voice assistants to see how they tick:
– Did the assistant have an answer or not?•Was the answer good, bad, or incomplete?
– How do the assistants respond to commands vs. informational questions?
– Does performance and relevance differ by topic or industry vertical?
– Does the language used to express the same intent create different results?
– How does personalization and history impact the experience and results?
– How does this change over time for the same questions?
– Broc Romanek
The checklist below is derived from this video that captures a 40-minute presentation by Susan & Scot Westwater at this year’s “Lingo Fest” conference (check out their podcast called “Pragmatic Talk”):
1. Start small & evolve your skill over time
2. Do the Venn diagram to see where your business objectives intersect with your customer’s needs
3. How will they use it?
4. What language will they use?
5. Where will they use it?
6. Are there any limitations?
7. What will their emotional state be?
B. Design & Development
1. What is the ideal path?
2. Where is it necessary to include a confirmation?
3. How can you build context from the previous answer?
4. How do you handle an “incorrect” answer?
5. How do you handle an “out-of-bounds” request?
6. When do you call in a human (ie. refer them to an agent or a rep)?
7. How does the entire experience align with your brand’s personality?
1. Is the usage behavior what you expected?
2. What additional things is the audience trying to do?
3. Are there any new traps that were discovered?
4. Is there one time use or are people coming back?
5. Are there utterances that weren’t accounted for?
This video captures a 30-minute presentation by Jackson Carson – UX/UI Director at SnapEngage – at this year’s “Lingo Fest” conference. Here’s five interesting things that were discussed:
1. People buy stuff based on their “experience” with the company selling it. You can’t just compete on your product’s features anymore – you have to understand how your potential customers think about you.
2. When it comes to the “UX Maturity Pyramid,” most voice offerings are at the “useful” level. That’s why it’s smart to start with basic skills that meet a big need for your customers – so you can try to achieve the “delightful” level.
3. You can never do enough user research before you start building. Voice requires more u/x than other platforms – you need to do a lot more upfront. You need to understand exactly what your customers really want – and really do the research, don’t just anticipate that you know their needs because you’ve already been meeting their needs for a few decades already.
4. For design, consider the “Double Diamond” – and the “Design Thinking Process.” Make the creative process something that is disciplined that becomes a cultural practice within your company.
5. Boil down the “Design Sprint” from a five-day process into a single day process. Or even something accomplished over a long lunch. The steps include mind-mapping and “crazy eights” exercises.